82556 Financial Inclusion GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Financial Inclusion © 2014 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 16 15 14 13 This work is a product of the staff of The World Bank with external contributions. Note that The World Bank does not necessarily own each component of the content included in the work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denomina- tions, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http:// creativecommons.org/licenses/by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2014. Global Financial Development Report 2014: Financial Inclusion. Washington, DC: World Bank. doi:10.1596/978-0-8213-9985-9. License: Cre- ative Commons Attribution CC BY 3.0 Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@world bank.org. ISBN (paper): 978-0-8213-9985-9 ISBN (electronic): 978-0-8213-9990-3 ISSN 2304-957X DOI: 10.1596/978-0-8213-9985-9 The report reflects information available up to September 30, 2013. Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Abbreviations and Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 Financial Inclusion: Importance, Key Facts, and Drivers . . . . . . . . . . . . . . . . . . . . . . . 15 2 Financial Inclusion for Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 3 Financial Inclusion for Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Statistical Appendixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 A Basic Data on Financial System Characteristics, 2009–11 . . . . . . . . . . . . . . . . . . . . . 152 B Key Aspects of Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 C Islamic Banking and Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 v vi CONTENTS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOXES O.1 Main Messages of This Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 O.2 The Views of Practitioners on Financial Inclusion: The Global Financial Barometer . .4 O.3 Navigating This Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 1.1 What Makes Finance Different? Moral Hazard and Adverse Selection . . . . . . . . . . .17 1.2 Overview of Global Data Sources on Financial Inclusion . . . . . . . . . . . . . . . . . . . . .19 1.3 The Gender Gap in Use of Financial Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 1.4 Islamic Finance and Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 1.5 Three Tales of Overborrowing: Bosnia and Herzegovina, India, and the United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 2.1 Remittances, Technology, and Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . .55 2.2 Correspondent Banking and Financial Inclusion in Brazil . . . . . . . . . . . . . . . . . . . . .59 2.3 The Credit Market Consequences of Improved Personal Identification . . . . . . . . . . .62 2.4 Insurance: Designing Appropriate Products for Risk Management . . . . . . . . . . . . . .71 2.5 Behavior Change through Mass Media: A South Africa Example . . . . . . . . . . . . . . .86 2.6 Case Study: New Financial Disclosure Requirements in Mexico . . . . . . . . . . . . . . . .90 2.7 Monopoly Rents, Bank Concentration, and Private Credit Reporting . . . . . . . . . . . .97 2.8 Exiting the Debt Trap: Can Borrower Bailouts Restore Access to Finance? . . . . . . .99 3.1 Financial Inclusion of Informal Firms: Cross-Country Evidence . . . . . . . . . . . . . . .108 3.2 Returns to Capital in Microenterprises: Evidence from a Field Experiment . . . . . . .110 3.3 The Effect of Financial Inclusion on Business Survival, the Labor Market, and Earnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114 3.4 Financing SMEs in Africa: Competition, Innovation, and Governments . . . . . . . . .119 3.5 Collateral Registries Can Spur the Access of Firms to Finance . . . . . . . . . . . . . . . . .124 3.6 Case Study: Factoring in Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127 3.7 Case Study: Angel Investment in the Middle East and North Africa . . . . . . . . . . . .135 3.8 Case Study: Nigeria’s YouWiN! Business Plan Competition . . . . . . . . . . . . . . . . . .137 FIGURES O.1 Use of Bank Accounts and Self-Reported Barriers to Use . . . . . . . . . . . . . . . . . . . . . . .2 BO.2.1 Views on Effective Financial Inclusion Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 O.2 Correlates of Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 O.3 Effect of Collateral Registry Reforms on Access to Finance . . . . . . . . . . . . . . . . . . . . .7 O.4 Fingerprinting and Repayment Rates, Malawi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 O.5 Individuals Who Work as Informal Business Owners in Municipalities with and without Banco Azteca over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 O.6 Effects of Entertainment Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 1.1 Use of and Access to Financial Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 B1.1.1 Financial Exclusion in Market Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 CONTENTS vii 1.2 Trends in Number of Accounts, Commercial Banks, 2004–11 . . . . . . . . . . . . . . . . .20 1.3 Provider-Side and User-Side Data on Financial Inclusion . . . . . . . . . . . . . . . . . . . . . .21 1.4 Selected Methods of Payment, 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 1.5 Reasons for Loans Reported by Borrowers, Developing Economies . . . . . . . . . . . . .25 1.6 The Use of Accounts and Loans by Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 1.7 Finance Is a Major Constraint among Firms, Especially Small Firms . . . . . . . . . . . . .27 1.8 Sources of External Financing for Fixed Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 1.9 Business Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28 1.10 Formal and Informal Firms with Accounts and Loans . . . . . . . . . . . . . . . . . . . . . . . .29 1.11 Reasons for Not Applying for a Loan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 1.12 Financial Inclusion vs. Depth, Efficiency, and Stability (Financial Institutions) . . . . .32 1.13 Correlates of Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 1.14 Reported Reasons for Not Having a Bank Account . . . . . . . . . . . . . . . . . . . . . . . . . .34 B1.4.1 Islamic Banking, Religiosity, and Access of Firms to Financial Services . . . . . . . . . . .38 1.15 Ratio of Cooperatives, State Specialized Financial Institutions, and Microfinance Institution Branches to Commercial Bank Branches . . . . . . . . . . . . . . . . . . . . . . . . .39 1.16 Correlation between Income Inequality and Inequality in the Use of Financial Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42 2.1 Mobile Phones per 100 People, by Country Income Group, 1990–2011 . . . . . . . . . .54 B2.1.1 Remittances and Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 B2.2.1 Voyager III: Bradesco’s Correspondent Bank in the Amazon . . . . . . . . . . . . . . . . . . .60 B2.3.1 Fingerprinting in Malawi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62 2.2 Mobile Phone Penetration and Mobile Payments . . . . . . . . . . . . . . . . . . . . . . . . . . .65 2.3 Share of Adults with an Account in a Formal Financial Institution . . . . . . . . . . . . . .66 B2.4.1 Growth in Livestock and Weather Microinsurance, India . . . . . . . . . . . . . . . . . . . . .71 B2.4.2 Payouts Relative to Premiums, Rainfall and Livestock Insurance, India . . . . . . . . . .72 2.4 Equity Bank’s Effect on Financial Inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 2.5 Borrowing for Food and Other Essentials, by Level of Education . . . . . . . . . . . . . . .78 2.6 Survey Results on Financial Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79 2.7 Effects of Secondary-School Financial Education, Brazil . . . . . . . . . . . . . . . . . . . . . .83 B2.5.1 Scandal! Cast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 B2.5.2 Effects of Entertainment Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87 2.8 Consumer Protection Regulations and Enforcement Actions . . . . . . . . . . . . . . . . . . .91 2.9 Evolution of Business Conduct, by Financial Market Depth . . . . . . . . . . . . . . . . . . .93 2.10 Credit Information Sharing and Per Capita Income . . . . . . . . . . . . . . . . . . . . . . . . . .95 B2.7.1 Credit Bureaus and Registries Are Less Likely if Banks Are Powerful . . . . . . . . . . . .97 B2.8.1 Bailouts and Moral Hazard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .100 3.1 Percentage of Micro, Very Small, Small, and Medium Firms . . . . . . . . . . . . . . . . .107 3.2 Biggest Obstacles Affecting the Operations of Informal Firms . . . . . . . . . . . . . . . .107 B3.1.1 Use of Finance by Informal Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109 viii CONTENTS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 B3.2.1 Estimated Returns to Capital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110 B3.3.1 Individuals Who Work as Informal Business Owners in Municipalities with and without Banco Azteca over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114 3.3 Employment Shares of SMEs vs. Large Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 3.4 Voluntary vs. Involuntary Exclusion from Loan Applications, SMEs . . . . . . . . . . .117 3.5 Voluntary vs. Involuntary Exclusion from Loan Applications, Large Firms . . . . . . .117 3.6 The Depth of Credit Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 B3.4.1 Financing Small and Medium Enterprises in Africa . . . . . . . . . . . . . . . . . . . . . . . . .119 B3.5.1 Effect of Collateral Registry Reforms on Access to Finance . . . . . . . . . . . . . . . . . . .124 B3.6.1 Actors and Links in the Financing Scheme, Peru . . . . . . . . . . . . . . . . . . . . . . . . . . .127 3.7 Average Loan Term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 3.8 Financing Patterns by Firm Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133 3.9 The Estimated Effect of Financing Sources on Innovation . . . . . . . . . . . . . . . . . . . .136 B3.8.1 Business Sectors of YouWiN! Winners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 3.10 Number of Countries with Joint Titling of Major Assets for Married Couples . . . .139 3.11 Adults with an Account Used for Business Purposes . . . . . . . . . . . . . . . . . . . . . . . .142 MAPS O.1 Adults Using a Bank Account in a Typical Month . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 1.1 Adults with an Account at a Formal Financial Institution . . . . . . . . . . . . . . . . . . . . .22 1.2 Origination of New Formal Loans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 1.3 Access by Firms to Securities Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30 1.4 Geography Matters: Example of Subnational Data on Financial Inclusion . . . . . . . .31 2.1 Mobile Phones per 100 People, 2011. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 A.1 Depth—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .159 A.2 Access—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160 A.3 Efficiency—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 A.4 Stability—Financial Institutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 A.5 Depth—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 A.6 Access—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164 A.7 Efficiency—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165 A.8 Stability—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 TABLES BO.2.1 Selected Results of the 2012–13 Financial Development Barometer . . . . . . . . . . . . . .4 B1.4.1 OIC Member Countries and the Rest of the World . . . . . . . . . . . . . . . . . . . . . . . . . .36 B1.4.2 Islamic Banking, Religiosity, and Household Access to Financial Services . . . . . . . . .37 B1.4.3 Islamic Banking, Religiosity, and Firm Access to Financial Services . . . . . . . . . . . . . .37 B2.2.1 Correspondent Banking, Brazil, December 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 CONTENTS ix 2.1 Financial Knowledge around the World. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 2.2 Effects of Financial Literacy Interventions and Monetary Incentives, Indonesia . . . .81 B3.1.1 Snapshot of Informal Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 . . . . . . .152 A.1.1 Depth—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .159 A.1.2 Access—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160 A.1.3 Efficiency—Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 A.1.4 Stability—Financial Institutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 A.1.5 Depth—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 A.1.6 Access—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164 A.1.7 Efficiency—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165 A.1.8 Stability—Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 B.1 Countries and Their Level of Financial Inclusion, 2011 . . . . . . . . . . . . . . . . . . . . . .167 C.1 OIC Member Countries, Account Penetration Rates, and Islamic Financial Institutions, 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .174 Foreword T he second Global Financial Develop- ment Report seeks to contribute to the evolving debate on financial inclusion. It fol- Indeed, half of the world’s adult population lacks a bank account. Many of the world’s poor would benefit from financial services but lows the inaugural 2013 Global Financial cannot access them due to market failures or Development Report, which re-examined inadequate public policies. the state’s role in finance following the global This Global Financial Development fi nancial crisis. Both reports seek to avoid Report contributes new data and research simplistic views, and instead take a nuanced that helps fill some of the gaps in knowledge approach to financial sector policy based on a about financial inclusion. It also draws on synthesis of new evidence. existing insights and experience to contribute Financial inclusion has moved up the to the policy discussion on this critical devel- global reform agenda and become a topic of opment issue. great interest for policy makers, regulators, The new evidence demonstrates that finan- researchers, market practitioners, and other cial inclusion can significantly reduce pov- stakeholders. For the World Bank Group, erty and boost shared prosperity, but under- financial inclusion represents a core topic, scores that efforts to foster inclusion must be given its implications for reducing poverty well designed. For example, creating bank and boosting shared prosperity. accounts that end up lying dormant has little The increased emphasis on financial inclu- impact, and policies that promote credit for sion reflects a growing realization of its all at any cost can actually exacerbate finan- potentially transformative power to accelerate cial and economic instability. This year’s development gains. Inclusive financial systems report offers practical, evidence-based advice provide individuals and firms with greater on policies that maximize the welfare benefits access to resources to meet their financial of financial inclusion. It also builds on the needs, such as saving for retirement, invest- benchmarking of financial institutions and ing in education, capitalizing on business markets fi rst introduced in the 2013 Global opportunities, and confronting shocks. Real- Financial Development Report. A rich array world financial systems are far from inclusive. of new fi nancial sector data, made publicly GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 xi xii FOREWORD GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 available through the World Bank Group’s including governments, international fi nan- Open Data Agenda, also accompany the new cial institutions, nongovernmental organiza- report. tions, think tanks, academics, private sector Following in the footsteps of its prede- participants, donors, and the wider develop- cessor, this year’s installment of the Global ment community. Financial Development Report represents one component of a broader initiative to enhance the stability and inclusiveness of Jim Yong Kim the global fi nancial system. We hope that it President proves useful to a wide range of stakeholders, The World Bank Group Acknowledgments G lobal Financial Development Report 2014 reflects the efforts of a broad and diverse group of experts, both inside and Kaushik Basu, Chief Economist and Senior Vice President, and Mahmoud Mohieldin, Special Envoy of the World Bank President, outside the World Bank Group. The report provided overall guidance. The authors has been cosponsored by the World Bank, the received invaluable advice from members International Finance Corporation (IFC), and of the World Bank’s Financial Development the Multilateral Investment Guarantee Agency Council, the Financial and Private Sector (MIGA). It reflects inputs from a wide range of Development Council, and the Financial units, including the Development Economics Inclusion and Infrastructure Practice. Vice Presidency, Financial and Private Sector External advisers included Meghana Ayy- Development Vice Presidency, all the regional agari (Associate Professor of International vice presidencies, the Poverty Reduction and Business, George Washington University), Economic Management Network, and Exter- Thorsten Beck (Professor of Economics nal and Corporate Relations Publishing and and Chairman of the European Banking Cen- Knowledge, as well as inputs from staff at the ter, Tilburg University, Netherlands), Ross Consultative Group to Assist the Poor. Levine (Willis H. Booth Professor in Bank- Aslı Demirgüç-Kunt was the project’s ing and Finance, University of Cali fornia ˇ director. Martin Cihák led the core team, Berkeley), Jonathan Morduch (Professor of which included Miriam Bruhn, Subika Farazi, Public Policy and Economics, New York Martin Kanz, Maria Soledad Martínez Pería, University Wagner Graduate School of Public Margaret Miller, Amin Mohseni-Cheraghlou, Service, and Managing Director, Financial and Claudia Ruiz Ortega. Other key contribu- Access Initiative), Klaus Schaeck (Professor tors included Leora Klapper (chapter 1), Doro- of Empirical Banking, Bangor University, the Singer (box 1.3), Christian Eigen-Zucchi United Kingdom), Robert Townsend (Eliza- (box 2.1), Gunhild Berg and Michael Fuchs beth and James Killian Professor of Econom- (box 3.4), Rogelio Marchetti (box 3.6), Sam ics, Massachusetts Institute of Technology), Raymond and Oltac Unsal (box 3.7), and and Christopher Woodruff (Professor of David McKenzie (box 3.8). Economics, University of Warwick). Aart GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 xiii xiv ACKNOWLEDGMENTS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Kraay reviewed the draft for consistency and North Africa Region); Idah Pswarayi-Riddi- quality multiple times. hough, Sujata Lamba, Shamsuddin Ahmad, Peer Stein and Bikki Randhawa have been Thyra Riley, Aruna Aysha Das, Henry Baga- the key contacts at IFC. Gaiv Tata, Douglas zonzya, Kiran Afzal, and Niraj Verma (all Pearce, and Massimo Cirasino have been South Asia Region); Jorge Familiar Calde- the interlocutors at the Financial Inclusion ron, Maria Cristina Uehara, and Sivan Tamir and Infrastructure Practice. Ravi Vish has (all Corporate Secretariat); Cyril Muller been the key contact at MIGA. Detailed (External and Corporate Relations); Mukesh comments on individual chapters have been Chawla, Robert Palacios, Olav Christensen, received from Ardic Alper, Nina Bilandzic, and Jee-Peng Tan (all Human Development Maria Teresa Chimienti, Massimo Cirasino, Network); Caroline Heider, Andrew Stone, Tito Cordella, Quy-Toan Do, Mary Hall- Beata Lenard, Jack Glen, Leonardo Alfonso ward-Driemeier, Oya Pinar, Mohammad Zia Bravo, Raghavan Narayanan, and Stoyan Qureshi, and Sergio Schmukler. The authors Tenev (all Independent Evaluation Group); also received valuable suggestions and other Jaime Saavedra-Chanduvi, Lucia Hanmer, contributions from Daryl Collins, Augusto and Swati Ghosh (all Poverty Reduction and de la Torre, Shantayanan Devarajan, Xavier Economic Management); Vijay Iyer (Sustain- Faz, Michael Fuchs, Matt Gamser, Xavier able Development Network); and Madelyn Giné, Jasmina Glisovic, Richard Hinz, Antoncic (Treasury). Martin Hommes, Zamir Iqbal, Juan Carlos The background work was presented Izaguirre, Dean Karlan, Alexia Latortue, in 14 Global Financial Development Semi- Timothy Lyman, Samuel Maimbo, Kate nars (http://www.worldbank.org/financial McKee, David McKenzie, Ajai Nair, Evariste development). The speakers and discussants Nduwayo, Douglas Pearce, Tomas Prouza, included, in addition to the core team, the Alban Pruthi, Steve Rasmussen, Rekha following: Abayomi Alawode, Michael Reddy, Mehnaz Safavian, Manohar Sharma, Bennett, Gunhild Berg, Timothy Brennan, Dorothe Singer, Ghada Teima, Hourn Thy, Francisco Campos, Robert Cull, Tatiana Judy Yang, and Bilal Zia. The manuscript Didier, Vincenzo Di Maro, Xavier Giné, also benefitted from informal conversations Mary Hallward-Driemeier, Zamir Iqbal, with colleagues at the International Mon- Leora Klapper, Cheng Hoon Lim, Rafe etary Fund, the Gates Foundation, Banco Mazer, David Medine, Martin Melecký, Bilbao Vizcaya Argentaria (BBVA), the U.K. Bernardo Morais, Florentina Mulaj, Maria Department for International Development, Lourdes Camba Opem, Douglas Pearce, and the World Economic Forum. Valeria Perotti, Mehnaz Safavian, Sergio In the World Bank–wide review, com- S ch mu k ler, Sandeep Sing h, Jonathan ments were received from Makhtar Diop, Spader, P. S. Srinivas, Wendy Teleki, Niraj Yira Mascaro, Xiaofeng Hua, Francesco Verma, and Bilal Zia. Strobbe, Ben Musuku, and Gunhild Berg The report would not be possible with- (all Africa Region); Sudhir Shetty, Nataliya out the production team, including Stephen Mylenko, and Hormoz Aghdaey (all East McGroarty (editor in chief), Janice Tuten Asia and Pacific Region); Laura Tuck, Ulrich (project manager), Nora Ridolfi (print Bartsch, Megumi Kubota, John Pollner, and coordinator), and Debra Naylor (graphic Vahe Vardanyan (all Europe and Central designer). Roula Yazigi assisted the team Asia Region); Hasan Tuluy, Marialisa Motta, with the website. The communications team Holger Kray, Jane Hwang, Marisela Monto- included Merrell Tuck, Nicole Frost, Ryan liu Munoz, P. S. Srinivas, Pierre Olivier Col- Douglas Hahn, Vamsee Kanchi, and Jane leye, Rekha Reddy, and Aarre Laakso (all Zhang. Excellent administrative assistance Latin America and the Caribbean Region); was provided by Hédia Arbi, Zenaida Kran- Shantayanan Devarajan (Middle East and zer, and Tourya Tourougui. Other valuable GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ACKNOWLEDGMENTS xv assistance was provided by Parabal Singh, The authors would also like to thank the Meaghan Conway and Julia Reichelstein. many country officials and other experts who Azita Amjadi, Liu Cui, Shelley Lai Fu, participated in the surveys underlying this Patricia Katayama, Buyant Erdene Khal- report, including the Financial Development tarkhuu, William Prince, Nora Ridolfi, Jomo Barometer. Tariku, and Janice Tuten have been helpful Financial support from Knowledge for in the preparation of the updated Little Data Change Program’s research support budget Book on Financial Development, accompa- and the IFC is gratefully acknowledged. nying this report. Abbreviations and Glossary GDP gross domestic product IFC International Finance Corporation MFI microfinance institution SME small and medium enterprises Note: All dollar amounts are U.S. dollars ($) unless otherwise indicated. GLOSSARY Country A territorial entity for which statistical data are maintained and pro- vided internationally on a separate, independent basis (not necessarily a state as understood by international law and practice). Financial Conceptually, financial development is a process of reducing the costs development of acquiring information, enforcing contracts, and making transac- tions. Empirically, measuring financial development directly is chal- lenging. This report focuses on measuring four characteristics (depth, access, efficiency, and stability) for financial institutions and markets (“4x2 framework”). Financial inclusion The share of individuals and firms that use financial services. Financial services Services provided to individuals and firms by the financial system. Financial system The financial system in a country is defined to include financial insti- tutions (banks, insurance companies, and other nonbank fi nancial institutions) and fi nancial markets (such as those in stocks, bonds, and financial derivatives). It also encompasses the financial infrastruc- ture (for example, credit information sharing systems and payments and settlement systems). GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 xvii xviii A B B R E V I AT I O N S A N D G L O S S A R Y GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Formal financial A commercial bank, insurance company, or any other financial insti- institution tution that is regulated by the state. State The country’s government as well as autonomous or semi-autonomous agencies such as a central bank or a financial supervision agency. Unbanked A person who does not use or does not have access to commercial banking services. Overview F inancial inclusion—typically defined as the proportion of individuals and fi rms that use fi nancial services—has become a cost, travel distance, and amount of paper- work play a key part. It is encouraging that most of these barriers can be reduced by bet- subject of considerable interest among policy ter policies. makers, researchers, and other stakeholders. Indeed, some progress has been achieved. In international forums, such as the Group of For example, in South Africa, 6 million basic Twenty (G-20), financial inclusion has moved bank accounts were opened in four years, up the reform agenda. At the country level, significantly increasing the share of adults about two-thirds of regulatory and supervi- with a bank account. Worldwide, hundreds sory agencies are now charged with enhanc- of millions have gained access to electronic ing financial inclusion. In recent years, some payments through services using mobile 50 countries have set formal targets and goals phone platforms. In the World Bank’s Global for financial inclusion. Financial Barometer (Cihák ˇ 2012; World The heightened interest reflects a better Bank 2012a), 78 percent of the financial sec- understanding of the importance of finan- tor practitioners surveyed indicated that, in cial inclusion for economic and social devel- their assessment, access to finance in their opment. It indicates a growing recognition countries had improved substantially in the that access to financial services has a critical last five years. role in reducing extreme poverty, boosting But boosting financial inclusion is not shared prosperity, and supporting inclusive trivial. Creating new bank accounts does not and sustainable development. The inter- always translate into regular use. For exam- est also derives from a growing recognition ple, of the above-mentioned 6 million new of the large gaps in fi nancial inclusion. For accounts in South Africa, only 3.5 million example, half of the world’s adult popula- became active, while the rest lie dormant. tion—more than 2.5 billion people—do not Moreover, things can go—and do go— have an account at a formal financial institu- badly, especially if credit starts growing rap- tion (figure O.1). Some of this nonuse demon- idly. The promotion of credit without suf- strates lack of demand, but barriers such as ficient regard for fi nancial stability is likely GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 1 2 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE O.1 Use of Bank Accounts and This is where the current report fits in. Self-Reported Barriers to Use It provides a careful review and synthesis of recent and ongoing research on fi nancial Have a bank account Not enough money inclusion, identifying which policies work, and which do not, as well as areas where more evidence is still needed. Box O.1 pro- 11% vides the main messages of this synthesis. Despite the growing interest, the views of policy makers and other fi nancial sector practitioners on the policies that work best are widely split (box O.2), underscoring the 50% major gaps in knowledge about the effects 39% of key policies on financial inclusion. Hence, this Global Financial Development Report, while recognizing the complexity of the ques- tions and the limits of existing knowledge, introduces new data and research and draws on available insights and experience to con- Other reasons for nonuse tribute to the policy discussion. (lack of trust, lack of documentation, distance to bank, religious reasons, FINANCIAL INCLUSION: another family member MEASUREMENT AND IMPACT already has an account) Financial inclusion and access to finance are different issues. Financial inclusion is Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. defi ned here as the proportion of individu- Note: Self-reported barriers to use of formal bank accounts. Respondents als and firms that use financial services. The could choose more than one reason. “Not enough money” refers to the percentage of all adults who reported only this reason. lack of use does not necessarily mean a lack of access. Some people may have access to financial services at affordable prices, but to result in a crisis. A spectacular recent choose not to use certain fi nancial services, example is the subprime mortgage crisis in while many others may lack access in the the United States in the 2000s: the key con- sense that the costs of these services are pro- tributing factors included the overextension hibitively high or that the services are simply of credit to noncreditworthy borrowers and unavailable because of regulatory barriers, relaxation in mortgage-underwriting stan- legal hurdles, or an assortment of market dards. Another example of overextension and cultural phenomena. The key issue is the of credit in the name of access was the cri- degree to which the lack of inclusion derives sis in India’s microfinance sector in 2010. from a lack of demand for fi nancial services Because of a rapid growth in loans, India’s or from barriers that impede individuals and microfinance institutions were able to report firms from accessing the services. high profitability for years, but this resided Globally, about 50 percent of adults have on large indebtedness among clients. While one or more bank accounts, and a nearly these two examples (explored in chapter 1, equal share are unbanked. In 2011, adults box 1.5) are more complex, they illustrate the who were banked included the 9 percent of broader point that deep social issues cannot adults who received loans and the 22 per- be resolved purely with an infusion of credit. cent of adults who saved through fi nancial If not implemented properly, efforts to pro- institutions. mote fi nancial inclusion can lead to defaults Looking beyond global averages, we fi nd and other negative effects. that financial inclusion varies widely around GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 3 BOX O.1 Main Messages of This Report The level of fi nancial inclusion varies widely around environment (for instance, setting standards for the world. Globally, about 50 percent of adults have disclosure and transparency and promoting credit a bank account, while the rest remain unbanked, information–sharing systems and collateral reg- meaning they do not have an account with a for- istries), and educating and protecting consumers. mal financial institution. Not all the 2.5 billion An important part of consumer protection is repre- unbanked need fi nancial services, but barriers such sented by competition policy because healthy com- as cost, travel distance, and documentation require- petition among providers rewards better performers ments are critical. For example, 20 percent of the and increases the power that consumers can exert unbanked report distance as a key reason they do in the marketplace. Policies to expand account pen- not have an account. The poor, women, youth, and etration—such as requiring banks to offer basic or rural residents tend to face greater barriers to access. low-fee accounts, granting exemptions from oner- Among fi rms, the younger and smaller ones are con- ous documentation requirements, allowing corre- fronted by more binding constraints. For instance, spondent banking, and using electronic payments in developing economies, 35 percent of small fi rms into bank accounts for government payments—are report that access to fi nance is a major obstacle to especially effective among those people who are their operations, compared with 25 percent of large often excluded: the poor, women, youth, and rural fi rms in developing economies and 8 percent of large residents. Other direct government interventions— fi rms in developed economies. such as directed credit, debt relief, and lending Financial inclusion is important for development through state-owned banks—tend to be politicized and poverty reduction. Considerable evidence indi- and less successful, particularly in weak institu- cates that the poor benefit enormously from basic tional environments. payments, savings, and insurance services. For firms, New technologies hold promise for expanding particularly the small and young ones that are sub- fi nancial inclusion. Innovations in technology—such ject to greater constraints, access to fi nance is associ- as mobile payments, mobile banking, and borrower ated with innovation, job creation, and growth. But identification using biometric data (fingerprint- dozens of microcredit experiments paint a mixed ing, iris scans, and so on)—make it easier and less picture about the development benefits of micro- expensive for people to use fi nancial services, while fi nance projects targeted at particular groups in the increasing financial security. The impact of new population. technologies can be amplified by the private sector’s Financial inclusion does not mean finance for adoption of business models that complement tech- all at all costs. Some individuals and fi rms have no nology platforms (as is the case with banking corre- material demand or business need for fi nancial ser- spondents). To harness the promise of new technolo- vices. Efforts to subsidize these services are coun- gies, regulators need to allow competing fi nancial terproductive and, in the case of credit, can lead to service providers and consumers to take advantage overindebtedness and fi nancial instability. However, of technological innovations. in many cases, the use of fi nancial services is con- Product designs that address market failures, strained by regulatory impediments or malfunction- meet consumer needs, and overcome behavioral ing markets that prevent people from accessing ben- problems can foster the widespread use of finan- eficial fi nancial services. cial services. Innovative fi nancial products, such as The focus of public policy should be on address- index-based insurance, can mitigate weather-related ing market failures . In many cases, the use of risks in agricultural production and help promote fi nancial services is constrained by market failures investment and productivity in agricultural fi rms. that cause the costs of these services to become Improvements in lending to micro and small fi rms prohibitively high or that cause the services to can be achieved by leveraging existing relationships. become unavailable due to regulatory barriers, legal For example, novel mechanisms have broadened hurdles, or an assortment of market and cultural fi nancial inclusion by delivering credit through retail phenomena. Evidence points to a function for gov- chains or large suppliers, relying on payment histo- ernment in dealing with these failures by creating ries in making loan decisions, and lowering costs by the associated legal and regulatory framework (for using existing distribution networks. example, protecting creditor rights, regulating busi- It is possible to enhance fi nancial capability— ness conduct, and overseeing recourse mechanisms fi nancial knowledge, skills, attitudes, and behav- to protect consumers), supporting the information iors—through well-designed, targeted interven- (box continued next page) 4 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX O.1 Main Messages of This Report (continued) tions. Financial education has a measurable impact channels—such as entertainment education—shows if it reaches people during teachable moments, for promise. In microenterprises, business training pro- instance, when they are starting a job or purchas- grams have been found to lead to improvements in ing a major fi nancial product. Financial education knowledge, but have a relatively small impact on is especially beneficial for individuals with lim- business practices and performance and depend on ited financial skills. Leveraging social networks context and gender, with mixed results. The con- (for example, involving both parents and children) tent of training also matters: simple rule-of-thumb tends to enhance the impact of fi nancial education. training is more effective than standard training in Delivery mode matters, too; thus, engaging delivery business and accounting. BOX O.2 The Views of Practitioners on Financial Inclusion: The Global Financial Barometer To examine views on financial inclusion among barometer, which addresses specifi c questions on some of the World Bank’s clients, the Global Finan- the state’s role in fi nance, were reported in Global cial Development Report team has undertaken the Financial Development Report 2013. Selected second round of the Financial Development Barom- results of the second barometer, with specific ques- eter, following up on the fi rst such survey from the tions on financial inclusion, are reported in this ˇ previous round (Cihák 2012; World Bank 2012a). box. (Additional information is available on the The barometer is a global informal poll of views, report’s website, at http://www.worldbank.org opinions, and sentiments among financial sector /financialdevelopment.) The barometer poll was practitioners (central bankers, fi nance ministry offi- carried out in 2012/13 and covers respondents from cials, market participants, and academics, as well as 21 developed and 54 developing economies. Of the representatives of nongovernmental organizations 265 individuals polled, 161 responded (a response and interdisciplinary research entities focusing on rate of 61 percent). fi nancial sector issues). According to the survey results, a majority of The barometer contains 23 questions arranged respondents see fi nancial inclusion as a big problem in two categories: (1) general questions about fi nan- both for households and for small enterprises. At the cial development and (2) specific questions relating same time, most respondents see an improving trend to the specifi c topic of the relevant Global Finan- in the access to fi nance in the last five years (table cial Development Report. The results of the fi rst BO.2.1, rows 1–3). TABLE BO.2.1 Selected Results of the 2012–13 Financial Development Barometer % of respondents agreeing with the statements 1. “Access to basic financial services is a significant problem for households in my country.” 61 2. “Access to finance is a significant barrier to the growth of small enterprises in my country.” 76 3. “In my country, access to finance has improved significantly over the last 5 years.” 78 4. “State banks and targeted lending programs to poorer segments of the population (social banking) are a 80 useful tool to increase financial access.” 5. “Social banking actually plays an important role in increasing financial access in my country.” 43 6. “The lack of knowledge about basic financial products and services is a major barrier to financial access 78 among the poor in my country.” Source: Financial Development Barometer; for full results, see the Global Financial Development Report website, at http://www.worldbank.org/financialdevelopment. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 5 BOX O.2 The Views of Practitioners on Financial Inclusion: The Global Financial Barometer (continued) On the role of the state, there is an interesting FIGURE BO.2.1 Views on Effective Financial disconnect: 80 percent of the respondents consider Inclusion Policies social banking—that is, state banks and lending Financial programs targeted at poorer segments of the popula- Better legal education tion—as a useful tool to increase fi nancial access, framework but only about 50 percent think that social banking actually plays a major part in expanding fi nancial 18% access (table BO.2.1, rows 4–5). 32% Another interesting result is that 78 percent of the respondents consider the lack of knowledge about basic financial products and services as a 17% major barrier to financial access among the poor (table BO.2.1, row 6). Corresponding to this result, for views about the best policy to improve access to finance among low-income borrowers, the policy Promotion of 33% selected by the greatest number of respondents (32 new lending percent) was fi nancial education (figure BO.2.1). Other technologies Source: Financial Development Barometer; for full results, see the Global Financial Development Report website, at http://www.worldbank.org /financialdevelopment. the world. Newly available user-side data workforce, or less well educated, or who live show striking disparities in the use of finan- in rural areas are much less likely to have an cial services by individuals in developed account (figure O.2). Account ownership also and developing economies. For instance, goes hand in hand with income equality: the the share of adults with a bank account in more even the distribution of income within developed economies is more than twice the a country, the higher the country’s account corresponding share in developing ones. The penetration. What helps is a better enabling disparities are even larger if we examine the environment for accessing financial services, actual use of accounts (map O.1). Worldwide, such as lower banking costs, proximity to 44 percent of adults regularly use a bank financial providers, and fewer documentation account. However, if we focus on the bottom requirements to open an account. 40 percent of income earners in developing While the disparities are less pronounced countries, we find that only 23 percent regu- in the access of fi rms to fi nance, significant larly use an account, which is about half the differences persist across countries and by participation rate among the rest of the popu- characteristics such as firm age and size. lations of these countries (the corresponding Younger and smaller fi rms face greater con- participation rates in developed economies straints, and their growth is affected rela- are 81 percent and 88 percent, respectively). tively more by the constraints. From the viewpoint of shared prosperity, Research highlighted in this report shows it is particularly troubling that the dispari- that financial inclusion matters for economic ties in financial inclusion are large in terms of development and poverty reduction. A range population segments within countries. People of theoretical models demonstrate how the who are poor, young, unemployed, out of the lack of access to fi nance can lead to poverty 6 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP O.1 Adults Using a Bank Account in a Typical Month Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. Note: Percentage of adults (age 15 years or older) depositing to or withdrawing from an account with a formal financial institution at least once in a typical month. FIGURE O.2 Correlates of Financial Inclusion 10 7 Effect on probability of owning an account, % 5 3 2 0 –2 –5 –3 –6 –7 –8 –10 –9 –13 –12 –15 –16 –20 Poorest Second Middle Fourth Age Rural 0–8 yrs of Log of Married Employed Unemployed Out of 20% 20% 20% 20% education household workforce size Source: Based on Allen, Demirgüç-Kunt, and others 2012. Note: Results from a probit regression of a financial inclusion indicator on country fixed effects and a set of individual characteristics for 124,334 adults (15 years of age and older) covered in 2011 in the Global Financial Inclusion (Global Findex) Database (http://www.worldbank.org/globalfindex). The financial inclusion indicator is a 0/1 variable indicating whether a person had an account at a formal financial institution in 2011. See Allen, Demirgüç-Kunt, and others (2012) for definitions, data sources, the standard errors of the parameter estimates, additional estimation methods, and additional regressions for other dependent variables (savings and the frequency of use of accounts). GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 7 traps and inequality. Empirical evidence on FIGURE O.3 Effect of Collateral Registry Reforms on Access to the impact of financial inclusion paints a pic- Finance ture that is far from black and white. The evi- dence varies by type of financial services. For 80 73 basic payments and savings, evidence on the benefits, especially among poor households, 70 is quite supportive. For insurance products, Firms with access to finance, % 60 54 there is also some evidence of a positive 50 impact, although studies on the effects of 50 microinsurance are inconclusive. As regards 41 40 access to credit, evidence on microcredit is mixed, with some cautionary findings on the 30 pitfalls of microcredit. For small and young firms, access to credit is important. 20 The message from the research is thus a 10 nuanced one: for inclusion to have positive effects, it needs to be achieved responsibly. 0 Prereform Postreform Creating many bank accounts that lie dor- mant makes little sense. While inclusion has Registry reformers Nonreformers (matched by region and income) important benefits, the policy objective can- not be inclusion for inclusion’s sake, and the Source: Doing Business (database), International Finance Corporation and World Bank, Washing- goal certainly cannot be to make everybody ton, DC, http://www.doingbusiness.org/data; Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org; calculations by borrow. Love, Martínez Pería, and Singh 2013. Note: Based on firm-level surveys in 73 countries, the study compares the access of firms to credit in seven countries that have introduced collateral registries for movable assets against the access of firms in a sample of countries matched by region and income per capita. PUBLIC POLICY ON FINANCIAL INCLUSION: OVERALL FINDINGS Enhancing financial inclusion requires the such as machines and other equipment, policy and market problems that lead to can greatly spur fi rm access to fi nance (fig- financial exclusion to be addressed. The pub- ure O.3). Importantly, the improvements lic sector can promote this goal by developing in access to fi nance are larger among small the appropriate legal and regulatory frame- firms. work and supporting the information envi- This evidence shows that improvements in ronment, as well as by educating and pro- the legal, regulatory, and institutional envi- tecting the users of fi nancial services. Many ronment, which tend to be helpful for devel- of the public sector interventions are more opment in general, are also quite useful for effective if the private sector is involved. For financial inclusion. example, improvements in the credit environ- How about policies aimed more directly ment, disclosure practices, and the collateral at financial inclusion? New evidence on framework can be more effective with private 142,000 people in 123 countries suggests sector buy-in and support. that policies aimed specifi cally at enhanc- New evidence showcased in this report ing account penetration and payments can suggests that the government has a par- be effective, especially among the poor, ticularly important role in overseeing the women, youth, and rural residents. Spe- information environment. Public policy can cifically, Allen, Demirgüç-Kunt, and others achieve potentially large effects on fi nancial (2012) show that, in countries with higher inclusion through reforms of credit bureaus banking costs, fi nancially excluded individ- and collateral registries. Evidence highlighted uals are more likely to report that they per- in the report indicates that the introduction ceive not having enough money as a barrier or reform of registries for movable collateral, to opening an account. Focusing only on 8 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 fi nancially excluded individuals who report Against this broader policy context, the “not having enough money” as the only bar- report examines three focus areas: (1) the rier, they observe that the presence of basic potential of new technology to increase or low-fee accounts, correspondent bank- financial inclusion; (2) the role of business ing, consumer protection, and accounts to models and product design; and (3) the role receive “government-to-person” (G2P) pay- of financial literacy, financial capability, and ments lower the likelihood that these indi- business training. These three areas can rein- viduals will cite lack of funds as a barrier. force each other. For example, new technol- While these results are not causal, they hint ogy can be used not only to boost fi nancial that government policies to enhance inclu- inclusion, but also to improve product design sion may be related to a higher likelihood and strengthen financial capability (as exam- that individuals consider that fi nancial ser- ined in the studies on the use of text messages vices are within their reach. to promote savings behavior that are refer- In contrast, direct government interven- enced in chapter 2). The report’s emphasis of tions in credit markets tend to be politicized these areas reflects the impact these areas can and less successful, particularly in environ- have on fi nancial inclusion and shared pros- ments with weak institutions. Examples of perity, as well as the fact that there is new such direct interventions include bailouts evidence to highlight. (For help in navigating and debt relief for households, directed the report, see box O.3.) credit, subsidies, and lending via state- owned fi nancial institutions. The challenges FOCUS AREA 1: associated with these direct government THE PROMISE OF TECHNOLOGY interventions are discussed in Global Finan- cial Development Report 2013, where the Technological innovations can lower the cost focus is “Rethinking the Role of the State in and inconvenience of accessing financial ser- Finance.” The current report highlights addi- vices. The last decade has been marked by a tional, novel evidence. For example, recent rapid growth in new technologies, such as in-depth analysis of India’s 2008 debt relief mobile payments, mobile banking, Internet for highly indebted rural households finds banking, and biometric identification tech- that, while the initiative led to the intended nologies. These technological innovations reductions in household debt, it was associ- allow for a significant reduction in trans- ated with declines in investment and agricul- action costs, leading to greater financial tural productivity (Kanz 2012). inclusion. Research also suggests that it mat- While much of the public discussion has ters how the various interventions are put focused on mobile payments and mobile together. Packaging reforms together leads banking, other new technologies are also to scale effects—positive and negative— promising. Recent research suggests that and to sequencing issues. For example, in a biometric identification (such as fi ngerprint- country in which creditor rights are weakly ing, iris scans, and so on) can substantially enforced because of a poorly functioning reduce information problems and moral haz- judiciary, a policy that would center solely ard in credit markets. To illustrate, figure on the computerization and unifi cation of O.4 shows results from a study authored by credit registries for movable collateral would World Bank researchers and based on a field have a limited impact on credit inclusion if experiment involving the introduction of fin- it were not combined with other supportive gerprinting among borrowers. The interven- reforms that may take longer to implement. tion significantly improved the lender’s abil- Considerations of this sort help impart some ity to deny credit in a later period based on welcome realism to aspirational objectives previous repayment performance. This, in of universal access and assist countries in turn, reduced adverse selection and moral operationalizing their national financial hazard, leading to improved loan perfor- inclusion strategies. mance among the weakest borrowers. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 9 BOX O.3 Navigating This Report The rest of the report consists of three chapters, credit, but it also discusses the importance of savings which cover (1) the importance of fi nancial inclusion, and insurance products for fi rms. The chapter high- some key facts, and drivers of fi nancial inclusion; lights three topics that have recently received much (2) fi nancial inclusion for individuals; and (3) fi nan- policy and research attention: (1) whether gender cial inclusion among firms. Within these broader matters in lending and the extent to which differ- topic areas, the report focuses on policy-relevant ences in fi rm growth arise because of differences in issues on which new evidence can be provided. access to fi nance among fi rms owned by women and Chapter 1 introduces the concept of financial fi rms owned by men; (2) the challenges and fi nanc- inclusion and reviews the evidence on its links to ing needs of rural fi rms; and (3) the role of fi nance in fi nancial, economic, and social development. It dis- promoting innovation. cusses the benefits of and the limits to inclusion and The Statistical Appendix consists of three parts. the importance of pursuing this agenda responsi- Appendix A presents basic country-by-country data bly. It highlights evidence based on theoretical and on fi nancial system characteristics around the world. empirical research on the impact of fi nancial inclu- It also shows averages of the same indicators for sion on economic development and identifies trans- peer categories of countries, together with summary mission channels through which fi nancial inclusion maps. It is an update of information in the 2013 contributes to income equality and poverty reduc- Global Financial Development Report. Appendix tion. The chapter introduces cross-cutting issues B provides additional information on key aspects related to fi nancial inclusion, such as the relationship of fi nancial inclusion around the world. Appendix between fi nancial sector structure and inclusion. C contains additional data on Islamic banking and Chapter 2 focuses on fi nancial inclusion among fi nancial inclusion in member countries of the Orga- individuals. It starts with a discussion on the role nization of Islamic Cooperation (OIC). of technology in financial inclusion. This is fol- The accompanying website (http://www.world lowed by an examination of private sector initia- bank.org/fi nancialdevelopment) contains a wealth of tives in financial inclusion, particularly product underlying research, additional evidence, including designs that address market failures, meet consumer country examples, and extensive databases on finan- needs, and overcome behavioral problems to foster cial development. It provides users with interactive the widespread use of fi nancial services. Financial access to information on fi nancial systems. The web- literacy and capability are another area of special site is a place where users can supply feedback on the focus. The chapter ends with an in-depth discus- report, participate in an online version of the Finan- sion of the various public sector policies in fi nancial cial Development Barometer, and submit suggestions inclusion and provides some evidence-based policy for future issues of the report. The website also pres- recommendations. ents an updated and expanded version of the Global Chapter 3 covers fi nancial inclusion among fi rms. Financial Development Database, a data set of 104 It focuses on firms that face market failures that fi nancial system characteristics for 203 economies restrict access to fi nance, such as small fi rms and since 1960, which was launched together with the young fi rms. The discussion covers access not only 2013 Global Financial Development Report. The to formal bank credit, but also to microfi nance, pri- database has now been updated with data on 2011, vate equity, and other forms of fi nance, such as fac- and new series have been added to the data set, espe- toring and leasing. The chapter focuses on access to cially in areas related to the nonbank financial sector. The adoption of new technologies has payment services took off when the coun- taken different paths in different economies. try’s mobile phone penetration rate was only In mobile technology, for instance, neither about 20 percent. This was similar to the ubiquity nor a high penetration of mobile penetration rate in countries such as Afghan- phones is a necessary condition for the devel- istan, Rwanda, and Tanzania, where mobile opment of mobile banking. To illustrate this, payments have not developed to such a high consider the example of Kenya, where mobile degree. 10 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE O.4 Fingerprinting and Repayment Rates, Malawi 100 96 98 91 92 93 88 89 90 79 80 74 70 Repayment rate, % 60 50 40 30 26 20 10 0 Q1 Q2 Q3 Q4 Q5 Fingerprinted Control Source: Calculations based on Giné, Goldberg, and Yang 2012. Note: The repayment rates among fingerprinted and control groups by quartiles of the ex ante probability of default. Individuals in the “worst” quintile (Q1) are those with the highest probability of default and those on whom fingerprinting had the largest effect. The evidence indicates that one of the fac- banks diminish the access of firms to finance. tors that truly make a difference is competi- This effect is stronger if financial develop- tion among providers of financial services. To ment is less advanced, if the share of govern- harness the potential of technologies, regula- ment banks is higher, and if credit informa- tors need to allow competing financial service tion is less available or of lower quality. providers and consumers to take advantage of technological innovations. This may seem FOCUS AREA 2: controversial for two main reasons. First, PRODUCT DESIGN AND regulators have to walk a fi ne line between BUSINESS MODELS providing incentives for the development of new payment technologies (allowing pro- Wider use of fi nancial services can also be viders to capture some monopoly rents to fostered by innovative product designs that recoup investments) and requiring the new address market failures, meet consumer platforms to be open. Second, competition needs, and overcome behavioral problems. without proper regulation and supervision One example of such product design is the could cause credit to become overextended commitment savings account, whereby an among people who are not qualified, which individual deposits a certain amount and could lead to a crisis. But, as noted in the relinquishes access to the cash for a period of fi rst Global Financial Development Report time or until a goal has been reached. These (World Bank 2012a), the evidence on crises accounts have been viewed as a tool to pro- actually underscores that healthy competition mote savings by mitigating self-control issues among providers rewards better performers and family pressures to share windfalls. and increases the power that consumers can One of the studies authored by World Bank exert in the marketplace. The present report researchers and highlighted in this report follows up on this theme, highlighting new (Brune and others 2011) finds that farm- evidence that low rates of competition among ers who were offered commitment accounts GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 11 saved more than farmers who were offered An interesting case that illustrates this checking accounts. This had positive effects point is Mexico’s Banco Azteca (Bruhn on agricultural input use, crop sales, and and Love 2013). In October 2002, the household expenditures. The study finds that bank simultaneously opened more than commitment accounts work primarily by 800 branches in all the stores of its parent shielding the funds of farmers from the social company, Grupo Elektra, a large retailer of networks of the farmers, rather than by help- consumer goods. The bank catered to low- ing the farmers deal with self-control issues. and middle-income groups that were mostly Another example of innovative product excluded from the commercial banking sec- design is index-based insurance. In contrast tor. Capitalizing on the parent company’s to traditional insurance, payouts for index- rich data, established information, collec- based insurance are linked to a measurable tion technology, and experience in making index, such as the amount of rainfall over a small installment loans for merchandise, the given time or commodity prices at a given bank was able to require less documentation date. Index insurance reduces problems of than traditional commercial banks, often moral hazard because payouts occur accord- accepting collateral and cosigners instead ing to a measurable index that is beyond of valid documents. Analysis highlighted in the control of the policyholder. Also, index this report shows that the new bank branch insurance is well suited to protect against the openings led to an increase in the proportion adverse shocks that affect many members of of individuals who ran informal businesses, informal insurance networks simultaneously. but to no change in formal businesses. After It has clear benefits for lenders and a poten- the Banco Azteca branches were opened, tial to increase fi nancial inclusion and agri- the proportion of informal business own- cultural production. Interestingly, however, ers increased significantly in municipalities the take-up of index insurance has often been with Banco Azteca branches (figure O.5). low. For example, in a randomized experi- Additionally, Banco Azteca’s branch open- ment with farmers highlighted in this report, ings generated increases in employment take-up was only 20 percent for loans with and income levels. These fi ndings illustrate rainfall insurance, compared with 33 percent that innovative business models can address for loans without insurance (Giné and Yang some of the failures that lead to financial 2009). New evidence from field experiments exclusion. suggests that lack of trust and liquidity con- straints are significant nonprice frictions that FOCUS AREA 3: constrain demand. Therefore, what helps is FINANCIAL LITERACY AND to design fi nancial products that pay often BUSINESS TRAINING and quickly; endorsements by credible, well- regarded institutions; the simplification of Financial literacy is different from fi nancial products; and consumer education. capability. Research indicates that standard, Beyond product design, innovative busi- classroom-based financial education aimed at ness models can help enhance economic the general population does not have much growth. For example, microenterprises are of an impact on fi nancial inclusion. It takes often financially constrained because of a more than lectures and memorizing defi ni- lack of information. This constraint can be tions to develop the capacity needed to ben- addressed by leveraging existing relation- efit from financial services. This can be illus- ships. Recent years have seen a growth in trated through an analogy with driving cars. innovative channels for the delivery of credit A person can learn the meaning of street through retail chains or large suppliers, reli- signs, but this does not make him capable of ance on payment histories to make loan deci- driving in traffic. Similarly, financial literacy sions, and the reduction of costs through the does not ensure that a person is fi nancially use of existing distribution networks. capable and able to make fi nancially sound 12 OVERVIEW GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE O.5 Individuals Who Work as Informal Business Owners in Municipalities with and without Banco Azteca over Time 15.0 9.0 Banco Azteca opening Individuals who are informal business owners Individuals who are informal business owners (municipalities without Banco Azteca), % (municipalities with Banco Azteca), % 14.5 8.8 8.6 14.0 8.4 13.5 8.2 13.0 8.0 12.5 7.8 12.0 7.6 2000 2001 2002 2003 2004 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th Quarter Quarter Quarter Quarter Quarter Municipalities without Banco Azteca Municipalities with Banco Azteca Source: Bruhn and Love 2013. Note: The study uses the predetermined locations of Banco Azteca branches to identify the causal impact of Banco Azteca branch openings on economic activity through a difference-in-difference strategy. It controls for the possibility that time trends in outcome variables may be different in municipalities that had Grupo Elektra stores and those that did not have such stores (see the text). decisions. Promoting financial capability individuals and fi rms. It is possible to boost through standard financial education has financial capability through well-designed proven to be extremely challenging. and targeted interventions. Interventions that Recent research has identified some inter- use teachable moments, such as starting a job ventions that can raise the fi nancial skills of or purchasing a major financial product, have been shown to have a measurable impact. The evidence also demonstrates that financial FIGURE O.6 Effects of Entertainment Education education is especially beneficial among peo- ple with below-average education and limited 31 financial skills. The impact of fi nancial edu- 30 cation is enhanced by leveraging social net- 26 25 works, which means involving both parents and children in the program or, in the case Respondents, % 19 20 of remittances, both senders and recipients. In 15 microenterprises, business training programs 10 have been found to lead to improvements in knowledge; however, these programs have a 5 relatively small impact on business practices 0 and performance. Recent research suggests Has someone in the household Has someone in the household that education focusing on rules of thumb is used hire purchase in the gambled money in the particularly helpful because it avoids infor- past 6 months? past 6 months? mation overload. Treatment Control New research on financial capability indicates that the mode of delivery can be a Source: Berg and Zia 2013. major factor in the effectiveness of outreach. Note: “Hire purchase” refers to contracts whereby people pay for goods in installments. One example of engaging delivery channels GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 OVERVIEW 13 that show some promise is entertainment statistically and economically significant education, highlighted in this report through effects on the fi nancial behavior of respon- a study that analyzes the impact of South dents who watched the show (figure O.6). At Africa’s soap opera “Scandal.” The soap the same time, the effects of this and other opera’s story line incorporated examples of fi nancial interventions are often short-lived, financially irresponsible behavior and the suggesting the need for these interventions effects of such behavior. Researchers found to be repeated. CHAPTER 1: KEY MESSAGES • Financial inclusion—the proportion of individuals and fi rms that use financial services — varies widely across the world. • More than 2.5 billion adults—about half of the world’s adult population—do not have a bank account. While some of these people exhibit no demand for accounts, most are excluded because of barriers such as cost, travel distance, and amount of paperwork. • Enterprise surveys in 137 countries find that only 34 percent of firms in developing econo- mies have a bank loan, whereas the share is 51 percent in developed economies. In develop- ing economies, 35 percent of small firms identify finance as a major constraint, while only 16 percent in developed economies do so. • Research—both theoretical and empirical—suggests that financial inclusion is important for development and poverty reduction. For the poor, the relevant evidence is especially strong on access to savings and automated payments; it is much weaker on access to credit. For firms, especially for small and medium enterprises and new entrepreneurs, improving access to credit is likely to have significant growth benefits. • If inclusion is to have positive effects, it needs to be promoted responsibly. Financial inclusion does not mean credit for all at all costs. • A diverse and competitive fi nancial sector —one that includes different types of fi nancial providers and financial markets—is helpful in supplying the range of products and services necessary for healthy financial inclusion. FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 1 Financial Inclusion: Importance, Key Facts, and Drivers W ell-functioning financial systems serve a vital purpose by offering savings, payment, credit, and risk manage- (1993), the occupational choices of individu- als (between becoming entrepreneurs or re- maining wage earners) are limited by the ini- ment services to individuals and firms.1 In- tial endowments. These occupational choices clusive financial systems are those with a determine how much the individuals can save high share of individuals and firms that use and what risks they can bear, with long-run financial services. Without inclusiveness in implications for growth and income distribu- financial systems, people must rely on their tion.2 These models show that lack of access own limited savings to invest in education to finance can be critical for generating per- or become entrepreneurs. Newly founded sistent income inequality or poverty traps, as enterprises must likewise depend on their well as lower growth. constrained earnings to take advantage of promising growth opportunities. This can DEFINING FINANCIAL contribute to persistent income inequality and INCLUSION slow economic growth. Development theory provides important Financial inclusion, as defined in this report, clues about the impact of financial inclusion is the proportion of individuals and firms that on economic development. Available mod- use financial services.3 It has a multitude of els illustrate how financial exclusion and, in dimensions, reflecting the variety of possible particular, lack of access to finance can lead financial services, from payments and savings to poverty traps and inequality (Aghion and accounts to credit, insurance, pensions, and Bolton 1997; Banerjee and Newman 1993; securities markets. It can be determined dif- Galor and Zeira 1993). For example, in the ferently for individuals and for firms. model of Galor and Zeira (1993), it is be- Greater financial inclusion is not neces- cause of financial market frictions that poor sarily good. For the most part, more exten- people cannot invest in their education, de- sive availability of financial services allows spite their high marginal productivity of in- individuals and firms to take advantage of vestment. In Banerjee and Newman’s model business opportunities, invest in education, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 15 16 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 1.1 Use of and Access to Financial Services #1: No need for financial services Users of formal Voluntary exclusion financial services (self-exclusion) #2: Cultural, religious reasons not to use, indirect access Population #3: Insufficient income, high risk Nonusers of formal financial services #4: Discrimination, lack of Involuntary exclusion information, weak contract enforcement, product features, price barriers due to market imperfections Source: Adapted from Demirgüç-Kunt, Beck, and Honohan 2008. save for retirement, and insure against risks box in figure 1.1.) Several groups belong to (Demirgüç-Kunt, Beck, and Honohan 2008). this category. One notable group consists of But not all financial services are appropriate individuals and firms that are unbankable for everyone, and—especially for credit— from the perspective of commercial financial there is a risk of overextension. institutions and markets because they have It is important to distinguish between the insufficient income or represent an excessive use of and access to financial services (figure lending risk. In this case, lack of use may not 1.1).4 Actual use is easier to observe empiri- be caused by market or government failure. cally. Some individuals and firms may have Other groups in this category may not have access to, but choose not to use some financial access because of discrimination, lack of in- services. Some may have indirect access, for formation, shortcomings in contract enforce- example, by being able to use someone else’s ment, a poor information environment, short- bank account. Others may not use finan- comings in product features that may make cial services because they do not need them a product inappropriate for some customer or because of cultural or religious reasons. groups, price barriers because of market im- The nonusers include individuals who prefer perfections, ill-informed regulations, or the to deal in cash and firms without promising political capture of regulators. If high prices investment projects. From a policy maker’s exclude large parts of the population, this viewpoint, nonusers do not constitute an is- may be a symptom of underdeveloped physi- sue because their nonuse is driven by lack of cal or institutional infrastructure, regulatory demand. However, financial literacy can still barriers, or lack of competition. Financial ex- improve awareness and generate demand.5 clusion deserves policy action if it is driven by Also, nonuse for religious reasons can be barriers that restrict access by individuals for addressed by allowing the entry of financial whom the marginal benefit of using a finan- institutions that offer, for instance, Shari‘a- cial service would otherwise be greater than compliant financial services (for example, see the marginal cost of providing the service. box 1.4). Box 1.1 explains why one hears about an ac- Some customers may be involuntarily ex- cess problem in markets for financial services cluded from the use of financial services. (This more often than in markets for other prod- is illustrated by the “involuntary exclusion” ucts and services.6 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 17 BOX 1.1 What Makes Finance Different? Moral Hazard and Adverse Selection Financial service markets differ from the markets for barriers and fi nancial exclusion arising because of other services and products. For example, one often high idiosyncratic risk or poor project quality (those hears about access problems in credit markets, but people identified as group #3 in figure 1.1) and those not about an access problem for, say, bubblegum. individuals facing such barriers because of mar- One of the basic rules of economics is that prices ket imperfections such as asymmetric information adjust so that, at market equilibrium, supply equals (group #4 in figure 1.1). demand. Hence, if the demand for bubblegum Rationing may arise even in a competitive credit exceeds the supply, the price of bubblegum will rise market because interest rates and bank charges affect until demand and supply are equated at a new equi- not only demand, but also the risk profi le of a bank’s librium price. If this price is too high for some con- customers: higher interest rates tend to attract riskier sumers, they will not be able to buy bubblegum. But borrowers (adverse selection) and change repayment all those who are willing and able to pay the price incentives (moral hazard). As a result, the expected will be able to buy bubblegum. So, if prices function rate of return of a loan will increase less rapidly than as expected, there should be no access problem. the interest rate and, beyond a point, may actually In a seminal paper, Stiglitz and Weiss (1981) decrease (fi gure B1.1.1, left panel). Because banks provide a compelling explanation of why fi nancial do not have perfect information about the credit- markets, particularly credit and insurance markets, worthiness of prospective borrowers, the supply of are different. a They show that information prob- loans will be backward bending at rates above the lems can lead to credit rationing and exclusion from bank’s optimal rate, r*. This means that fi nancial fi nancial markets even in equilibrium. Credit and exclusion will persist even at market equilibrium (at insurance markets are characterized by serious prin- rate rM). Because it is not profitable to supply more cipal agent problems, which include adverse selec- loans if the bank faces excess demand for credit, the tion (the fact that borrowers less seriously intent on bank will deny loans to borrowers who are obser- repaying loans are more willing to seek out external vationally indistinguishable from those who receive fi nance) and moral hazard (once the loan is received, loans. The rejected applicants would not receive a borrowers may use funds in ways that are inconsis- loan even if they offered to pay a higher rate. Hence, tent with the interest of the lenders). Therefore, in they are denied access. In other words, they may be considering involuntarily excluded users, one must bankable (that is, worthy, in principle, of fi nancial distinguish between those individuals facing price services), but are involuntarily excluded. FIGURE B1.1.1 Financial Exclusion in Market Equilibrium a. Expected return from a loan b. Supply and demand in loan market S Expected return to the bank rM r* D r* S* D* Source: Stiglitz and Weiss 1981. Note: D = demand; S = supply (box continued next page) 18 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 1.1 What Makes Finance Different? Moral Hazard and Adverse Selection (continued) Determining whether individuals or fi rms have who are not easy to oversee. These problems may access to credit, but choose not to use it or are sim- become acute and limit participation in contexts ply excluded is complex, and the effects of adverse where a lack of trust is an important reason for not selection and moral hazard are difficult to separate. opening an account at a formal fi nancial institution Thus, attempts to broaden access beyond level S* in (globally, 13 percent of adults report such a reason). figure B1.1.1, right panel, are associated with chal- The nonprice barriers are often considerable. lenges because they would require a bank to lower For example, potential customers may be discrimi- screening standards and may translate into higher nated against by the design features of a product, or risks for the bank and for borrowers. The global they may face barriers to access because of red tape. fi nancial crisis has highlighted that extending access Some individuals will have no access to financial at the expense of reduced screening and monitor- services because there are no fi nancial institutions ing standards can have severely negative implica- in their area; this is the case in many remote rural tions both for consumers and for fi nancial stability. areas. Yet others may be excluded because of poorly Therefore, in the case of credit, it is preferable to designed regulations, for instance, the documenta- enhance fi nancial inclusion through interventions tion requirements for opening an account, such as that increase supply by removing market imperfec- the existence of a formal address or of formal sector tions. Examples are new lending technologies that employment. For these individuals, the supply curve reduce transaction costs and improved borrower in the right panel of figure B1.1.1 would be vertical identification that can mitigate (even if not fully at the origin, and the supply and demand for services eradicate) the problems of asymmetric information. would not intersect, leading to fi nancial exclusion. Moral hazard and adverse selection issues are Policy makers are also concerned if high prices also well documented in insurance markets. Other and fi xed costs make it impossible for large segments fi nancial services, such as deposits and payments, do of the population to use basic services such as simple not suffer from moral hazard and adverse selection deposit or payment services. This is not an access to the same extent, but they still present policy chal- issue in the strict sense, but it still represents a policy lenges in terms of fi nancial inclusion. For example, challenge to the extent that it reflects a lack of com- agency problems also occur from the perspective petition or underdeveloped physical or institutional of depositors, particularly small depositors, who infrastructures, in resolving which government can entrust their fi nancial resources to intermediaries play a major role. a. For other explanations, see, for example, Keeton (1979) and Williamson (1987). The challenging but important practical is- 1.2). The Global Financial Inclusion (Global sue is that, if nonuse is observed, it is hard to Findex) Database provides a new set of indi- disentangle whether the nonuse is voluntary cators that measure how adults (aged 15 years or involuntary. Therefore, for policy reasons, and above) in 148 economies save, borrow, it is crucial that financial inclusion be prop- make payments, and manage risk by survey- erly measured. This is the subject of the fol- ing over 150,000 individuals with charac- lowing section. teristics representative of 97 percent of the world’s adult population during the 2011 cal- endar year.7 For instance, the share of adults MEASURING FINANCIAL in developed economies with an account at INCLUSION: KEY FACTS a formal financial institution is more than Financial inclusion varies widely across the twice the corresponding share in developing world. Newly available data confirm strik- economies. There is also substantial variation ing disparities in the use of financial services in financial inclusion within countries across in developed and developing economies (box individual characteristics such as income and GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 19 BOX 1.2 Overview of Global Data Sources on Financial Inclusion Until recently, the measurement of financial inclusion other reasons than their lack of knowledge as the around the world has focused on density indicators, main reason for being excluded.) such as the number of bank branches or automatic For now, the Global Findex data are available teller machines (ATMs) per capita. Data of this for one year, 2011. Global Findex will eventually type have been compiled by surveying finan- produce time series on usage (complete updates are cial service providers (for example, see Beck, planned for 2014 and 2017). The questionnaire, Demirgüç-Kunt, and Martínez Pería 2007; Kendall, translated into 142 languages to ensure national rep- Mylenko, and Ponce 2010). Much of this provider- resentation in 148 economies, can be used by local side in formation on fi nancial inclusion is now col- policy makers to collect additional data. lected as part of the Financial Access Survey, which The compilation of user-side data includes a grow- has annual data for 187 jurisdictions from 2001 to ing body of survey data on fi nancial capability, living 2011.a While these indicators have made it possible standards and measurement surveys, and enterprise to obtain basic provider-side information on the use surveys.c There are also user-side data that, while of fi nancial services, relatively little has been known limited in country coverage, have had a great impact until recently about the global reach of the fi nancial on financial inclusion (for example, the FinScope sur- sector, that is, the extent of fi nancial inclusion and veys conducted by FinMark Trust in South Africa the degree to which the poor, women, and other pop- and now expanding beyond the region).d Data col- ulation segments are excluded from formal fi nancial lected through the financial diaries methodology systems. used by Collins and others (2009) have also provided This gap in data has now been addressed with the important insights such as the sheer number of fi nan- release of the Global Financial Inclusion (Global Fin- cial transactions undertaken by the poor. dex) Database, built by the World Bank, in coopera- For the access of firms to finance, the World tion with the Bill and Melinda Gates Foundation and Bank’s Enterprise Surveys are the leading data set for Gallup, Inc. (see Demirgüç-Kunt and Klapper 2012).b the measurement of fi nancial inclusion by firms of all These user-side indicators, compiled using the Gal- sizes across countries.e The World Bank also compiles lup World Poll Survey, measure how adults in 148 informal surveys, similar in format to the Enterprise economies around the world manage their day-to- Surveys, but focused on fi rms in the informal sector.f day fi nances and plan for the future. The indicators The Global Financial Development Database con- are constructed with survey data from interviews tains additional variables, such as cross-country indi- with more than 150,000 nationally representative cators on the access of fi rms to securities markets.g It and randomly selected adults over the 2011 calendar has been updated and expanded as part of the work year. The database includes over 40 indicators related on this report (see the Statistical Appendix). to account ownership, payments, savings, borrowing, At a summit in June 2012, the leaders of the G-20 and risk management. As a survey-based data set, the endorsed the “G20 Basic Set of Financial Inclusion user-side data face certain challenges that have been Indicators,” which integrates existing global data addressed in the survey design and execution. (For efforts to compile indicators on the access to and instance, self-reported barriers could be misleading usage of fi nancial services. They include the Finan- because respondents with limited awareness of the cial Access Survey, the Global Findex Database, and benefits of formal fi nancial products may mention the Enterprise Surveys.h a. See Financial Access Survey (database), International Monetary Fund, Washington, DC, http://fas.imf.org/. b. See Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org /globalfi ndex. c. See World Bank (2013a) and Responsible Finance (database), World Bank, Washington, DC, http://responsiblefi nance .worldbank.org/. d. See FinScope (database), FinMark Trust, Randjespark, South Africa, http://www.fi nscope.co.za/. e. Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. f. For additional information on the informal surveys, see “Enterprise Surveys Data,” World Bank, Washington, DC, http://www.enterprisesurveys.org/Data. g. Global Financial Development Report (database), World Bank, Washington, DC, http://www.worldbank.org /fi nancialdevelopment. h. See “G20 Basic Set of Financial Inclusion Indicators,” Global Partnership for Financial Inclusion, Washington, DC, http://datatopics.worldbank.org/g20fidata/. 20 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 gender. While the disparities are less sizable in The data clearly indicate a wide disper- the access of firms to finance, considerable dif- sion in financial inclusion around the world. ferences exist across countries and by certain For example, user-side indicators such as characteristics, such as firm age and firm size the share of adults who own an account (see Beck, Demirgüç-Kunt, and Martínez Pería (shown on the vertical axes in figure 1.3) var- 2007; Cull, Demirgüç-Kunt, and Morduch ies from less than 1 percent (Turkmenistan) 2012; Demirgüç-Kunt and Klapper 2012). to more than 99 percent (Denmark). Simi- larly, provider-side indicators, such as bank branch density and ATM density (shown on Basic trends in financial inclusion the horizontal axes in figure 1.3), vary widely The available proxies suggest that the use of from country to country. The user-side and financial services has been slowly, but steadily, the provider-side measures are correlated, expanding over time. For example, the num- but their correspondence is far from perfect, ber of deposits and loan accounts with com- as illustrated in figure 1.3.9 For example, mercial banks has been increasing for the Bulgaria had 84 commercial bank branches whole period on which consistent data are per 100,000 adults in 2004–11, substantially available (figure 1.2).8 above the global average of 19 branches (with The growth in the number of deposits and a standard deviation of 19). At the same time, loan accounts dipped with the onset of the according to Global Findex, only 53 percent global financial crisis in 2008, but, despite of adult Bulgarians had an account with the this dip, the number of accounts continued formal financial system.10 In comparison, the to expand. Low-income countries did record Czech Republic, a country of approximately slightly positive growth in the number of ac- similar size and population, had account pen- counts, but the growth rate was generally etration of 81 percent, with 22 branches per lower than the corresponding rate in high- 100,000 adults. These differences underscore income countries, thereby increasing the that variables such as bank branch density, wedge between the two country groups in while useful, provide only a rough proxy for terms of financial inclusion. financial inclusion. FIGURE 1.2 Trends in Number of Accounts, Commercial Banks, 2004–11 1,300 a. Deposit accounts 1,300 b. Loan accounts 1,200 1,200 1236 1,100 High income 1,100 1,000 1,000 Deposit accounts, 1,000 adults Loan accounts per 1,000 adults 900 846 900 804 802 800 800 High income 700 Middle income 738 700 600 600 500 423 World average 500 381 400 400 459 World average 386 300 300 204 320 200 200 Middle income 100 201 166 116 Low income 100 Low income 24 21 0 0 2004 2005 2006 2007 2008 2009 2010 2011 2004 2005 2006 2007 2008 2009 2010 2011 Source: Calculations based on data from the Financial Access Survey (database), International Monetary Fund, Washington, DC, http://fas.imf.org. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 21 FIGURE 1.3 Provider-Side and User-Side Data on Financial Inclusion a. Branches per capita b. Branch density 100 100 Account penetration, % Account penetration, % 80 80 60 60 40 40 y = 8.70x0.57 y = 21.45x0.28 20 20 R 2 = 0.48 R 2 = 0.34 0 0 0 50 100 0 50 100 150 Commercial bank branches Commercial bank branches per 100,000 adults per 1,000 km2 c. ATMs per capita d. ATM density 100 100 80 80 Account penetration, % Account penetration, % 60 60 40 40 y = 17.23x0.30 y = 7.67x0.48 R 2 = 0.46 R 2 = 0.58 20 20 0 0 0 50 100 150 200 0 100 200 300 400 ATMs per 100,000 adults ATMs per 1,000 km2 Sources: Calculations based on data from the Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank .org/globalfindex for account penetration rates and the Financial Access Survey (database), International Monetary Fund, Washington, DC, http://fas.imf .org/ for the other four indicators. Note: ATM = automatic teller machine. Ownership of accounts example, in developing economies, on aver- age, the top 20 percent of the wealthiest adults Accounts are a key measure of financial in- are more than twice as likely as the poorest clusion because essentially all formal finan- adults to have a bank account (Demirgüç- cial activity is tied to accounts. In developed Kunt and Klapper 2012). Also, in developing economies, 89 percent of adults report that economies, women are 20 percent less likely they have an account at a formal financial in- stitution, while the share is only 24 percent in to have an account than men (box 1.3). low-income economies. Globally, 50 percent of the adult population—more than 2.5 bil- Payments lion people—do not have a formal account. In many countries in Africa, the Middle East, Noncash methods of payment are becoming and Southeast Asia, fewer than 1 in 5 adults more important, but they still lag behind cash has a bank account (map 1.1). methods in terms of penetration. Debit and Account penetration varies considerably credit cards account for a large part of non- not only among countries, but also across cash retail transactions. Only a small fraction individuals within the same country. For of adults are using mobile payments, although 22 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP 1.1 Adults with an Account at a Formal Financial Institution Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. this area has shown much promise (figure 1.4; commonly found in low-income econo- also see chapter 2). mies. (For example, in Sub-Saharan Africa, 19 percent of adults reported they had used savings clubs and similar methods in 2011.) SAVINGS As with account penetration, formal sav- Globally, 36 percent of adults report that they ings behavior also varies by country category saved or set aside money in the previous year. or by individual characteristics within coun- In high-income economies, this ratio is 58 tries (Demirgüç-Kunt and Klapper 2012). percent, while in low-income economies, it For instance, 43 percent of account holders is only 30 percent. Similar to account owner- worldwide saved at a formal financial insti- ship, the propensity to save differs across and tution in 2011. This ratio ranged from less within countries. than 1 percent in Armenia, Cambodia, the Only a portion of these savings is held Arab Republic of Egypt, the Kyrgyz Republic, in formal financial institutions. Worldwide, Tajikistan, Turkmenistan, and Uzbekistan to 22 percent of adults report they saved at a more than 60 percent in Australia, New Zea- bank, credit union, or microfinance institu- land, and Sweden. tion (MFI) in 2011. The share ranges from Worldwide, 12 percent of bank account 45 percent in high-income economies to holders save solely using methods other than 11 percent in low-income economies. Many bank accounts. The reasons for this include other people, including some who own a the high costs of actively using the account, formal account, rely on different meth- such as balance and withdrawal fees, as well ods of saving. Community-based savings as costs associated with physical distance. methods, such as savings clubs, are widely Policy makers or commercial bankers could used around the world as an alternative introduce new products to encourage existing or complement to saving at a formal finan- account holders to save in formal financial cial institution. These methods are most institutions. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 23 BOX 1.3 The Gender Gap in Use of Financial Services Gender is a powerful determinant of economic and countries, even after one controls for differences in financial opportunities. Women use the formal gross domestic product (GDP) per capita. fi nancial sector less than men, especially in develop- In a reflection of this discrimination, more ing economies. Globally, 47 percent of women own women than men turn to alternative means to man- or co-own an account, compared with 55 percent of age their day-to-day finances and plan for their men.a There are other major differences. For exam- future fi nancial needs. In Sub-Saharan Africa, for ple, the gap is largest among people living on less example, 53 percent of women who save do so using than $2 per day and among people living in South an informal community-based savings method, such Asia and the Middle East. But, according to analysis as a rotating savings and credit association, in com- by Demirgüç-Kunt, Klapper, and Singer (2013), the parison with 43 percent of the men who save. broad pattern holds in all regions of the world and In Findex surveys, women also cite the abil- across income groups within countries. ity to use another person’s account as a reason for The gender gap is large even within the same not opening one themselves.a Indeed, 26 percent of income groups, by 6 to 9 percentage points in devel- unbanked women in developing economies—com- oping economies. This signals that the gap is not pared with 20 percent of men—say they do not have merely a function of income. Indeed, the economet- an account because a family member already has ric analysis of Demirgüç-Kunt, Klapper, and Singer one. (2013) highlights that significant gender differ- The lack of asset ownership, moreover, may ences remain even after one controls for income and have an adverse effect on empowerment and self- education. employment opportunities. Several interesting stud- The gap extends beyond the opening of a bank ies use randomized controlled trials to show that account. Women lag signifi cantly behind men also providing access to personal savings instruments in the rate of saving and borrowing through formal boosts consumption and productive investment, institutions, even after we account for individual especially among women, thereby contributing to characteristics such as age, education, income, and the empowerment of women (Ashraf, Karlan, and urban or rural residence. In Latin America, for Yin 2010; Dupas and Robinson 2013 ). It is thus instance, 8 percent of women report that they had probably not enough for women to have access to saved in a formal institution in the previous year, an account; they also need to be the owners of their compared with 12 percent of men. accounts and savings. This can put women, often the main caregivers in In some countries, the gender gap in terms of use their families, at a disadvantage. If they do not own may be even greater than the gap in terms of account an account, women face more difficulties saving for- ownership. For example, a World Bank study in mally or receiving government payments or remit- Pakistan estimates that 50–70 percent of the loans tances from family members living abroad. The lack made to women clients may actually be used by their of a formal account can also create barriers to the male relatives (Safavian and Haq 2013). This high- access of formal credit channels, which may hinder lights the challenges of measuring inclusion, as well entrepreneurial or educational ambitions. as of designing policies to enhance fi nancial access: Several factors are to blame for this phenom- if, for instance, credit for women were subsidized, enon. Legal discrimination against women, such as an expansion in loans to women clients might not in employment laws and inheritance rights, explains necessarily translate into women gaining more a large portion of the gender gap across developing access to fi nance. a. Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org /globalfi ndex. A large share of adults around the world Adults saving through alternative methods, who report they saved or set aside money in such as accumulating gold or livestock or put- 2011 used neither formal financial institu- ting money “under the mattress,” account for tions nor community-based savings methods. 29 percent of savers globally. 24 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 1.4 Selected Methods of Payment, 2011 a formal financial institution in the previous 12 months. But, in developing economies, adults are three times more likely to borrow from Pays with debit card 59 family and friends than from formal financial 29 institutions. In high-income economies, the most commonly cited purpose of an outstand- 36 Pays with credit card 15 ing loan is to purchase a home, while emer- gency and health reasons are most frequently 24 cited by adults in the developing world. Pays electronically 16 Developed economies The introduction of credit cards has had Developing economies a large effect on the demand for and use of 10 short-term formal credit. In high-income Pays with mobile device 3 economies, half of the adult population re- 0 20 40 60 80 ports they have a credit card. Despite a surge Adults, % in recent years, credit card ownership in devel- oping economies still lags far behind the rates Source: Calculations based on data from the Global Financial Inclusion (Global Findex) Database, in high-income economies; in the former, only World Bank, Washington, DC, http://www.worldbank.org/globalfindex. Note: The response on mobile payments may subsume some of the other categories. (For example, 7 percent of adults report they have a credit using a credit card to make a payment by phone may be seen as a mobile payment by the card (notable exceptions are Brazil, Turkey, respondent.) and Uruguay, where the proportion of adults with a credit card exceeds 35 percent). Insurance As a result of the extensive ownership of Insurance is crucial for managing risks, both credit cards, adults in high-income economies the risks associated with personal health and may have less need for short-term loans from the risks associated with livelihoods. In devel- financial institutions. This may help explain oping economies, 17 percent of adults report why the share of adults in these economies they paid for health insurance (in addition to who report they received a loan in the pre- national health insurance, where applicable). vious year from a formal financial institution The share ranges from 3 to 5 percent in Sub- (such as a bank, cooperative, credit union, or Saharan Africa, Europe, Central Asia, and MFI) is not particularly high. Indeed, if the South Asia. The corresponding ratio for East adults in high-income economies who report Asia and the Pacific is 38 percent, which is they own a credit card are included in the driven by China, where 47 percent of adults share of those adults who report they bor- report they paid for health insurance; without rowed from a formal financial institution in China, the regional average would drop to the previous year (a measure that may not 9 percent. include credit card balances), the share rises People who work in farming, forestry, from 14 percent to 54 percent. or fishing are critically vulnerable to severe Individuals in higher-income economies are weather events. Nonetheless, only 6 percent more likely to borrow from formal sources, of these people report they purchased crop, while those in lower-income economies tend rainfall, or livestock insurance in 2011. In to rely more heavily on informal sources. To Europe and Central Asia, only 4 percent re- illustrate, in Finland, 24 percent of adults port they purchased such insurance. report they borrowed money from a formal financial institution, such as a bank, credit union, or MFI, in the previous year (map Credit 1.2). In Ukraine, only 8 percent of adults Most borrowing by adults in developing econ- report they did so, and, in Burundi, only omies occurs through informal sources, such 2 percent of adults report they used formal as family and friends. Globally, 9 percent of credit. The pattern is reversed with respect adults report they originated a new loan from to the proportion of adults who received GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 25 MAP 1.2 Origination of New Formal Loans Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. credit from informal sources (the shares of FIGURE 1.5 Reasons for Loans Reported by adults who have done so in Finland, Ukraine, Borrowers, Developing Economies and Burundi are 15 percent, 37 percent, and Adults with an outstanding loan, by purpose of loan 44 percent, respectively). This propensity toward informal borrowing persists across 12 low- and middle-income countries. Friends 10 and family are the most commonly reported 8 Adults, % sources of new loans in upper-middle-, lower- 6 middle-, and low-income economies, but not 4 in high-income economies. In low-income 2 economies, 20 percent of adults report that 0 friends or family were their only sources of Home School Emergency/ Funerals/ new loans in the previous year. In contrast, construction fees health weddings only 6 percent of adults report that a formal fi- Source: Global Financial Inclusion (Global Findex) Database, World Bank, nancial institution is their only source. Adults Washington, DC, http://www.worldbank.org/globalfindex. in poorer countries are also more likely to re- port they borrowed money from a private in- formal lender in the previous year. An impor- of an outstanding loan (figure 1.5). More in- tant caveat to this finding is that social norms depth analysis points to a fine line between may have a significant effect on the degree to lending for business purposes and lending for which this type of borrowing is reported. household purposes. For example, evidence The purpose of borrowing varies across from Mongolia indicates that about half of economies and by individual characteristics, all microcredit business loans are used for pur- similar to the differences in sources of credit. poses associated with the household, such as Data gathered in developing economies high- the purchase of domestic applicances (Attana- light that emergencies or health issues are par- sio and others 2011). Similar results have been ticularly common reasons for the existence obtained for Bangladesh, Indonesia, Peru, and 26 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 the Philippines (Collins and others 2009; John- example, firms in developed economies are, ston and Morduch 2008; Karlan and Zinman on average, more likely to use bank credit 2011). The key point is fungibility: a business (figure 1.6). Indeed, the dividing line between loan diverted to consumption can free up oth- firm finance and individual finance becomes er resources for business investment later; so, it cloudy in some cases, especially for small is still possible to see a positive net impact on firms and the poorer segments of the popula- business and income. Households need such tion, segments that policies aimed at enhanc- general use finance in part for self-employment ing financial inclusion are often designed to and in part for other priorities. reach. Frequently, the family is essentially a Data on the use of mortgages show large productive unit, that is, a firm. Thus, quasi- differences across economies at different experimental studies and randomized con- income levels. In high-income economies, trolled trials regularly tend not to make fine 24 percent of adults report they have an distinctions between the two.12 outstanding loan to purchase a home or Comparing the use of checking and sav- apartment. The corresponding share is only ings accounts by (formal sector) firms and 3 percent in developing economies. Even the use of loans by the same type of firms within the European Union, there is large reveals that the differences between devel- variation in the use of mortgages. For exam- oped and developing economies are relatively ple, while 21 percent of adults in Germany small (figure 1.6). In both sets of countries, a have an outstanding mortgage, only 3 per- vast majority of firms use bank accounts. In cent in Poland do so. Such differences may, comparison with their counterparts in high- in part, reflect differences in housing finance income economies, firms in low- and middle- systems across economies, such as in product income economies are more financially con- diversity, types of lenders, mortgage funding, strained. Interestingly, a significant share of and the degree of government participation. firms in both the high-income and the low- and middle-income groups (48 percent and 39 percent, respectively) reported they did FINANCIAL INCLUSION AMONG not need a loan. The geographic variation in FIRMS the use of bank accounts and loans by firms Many of the findings on financial inclu- is relatively large. While more than 90 per- sion among individuals are also applicable cent of firms in Eastern Europe have bank to financial inclusion among firms.11 For accounts, and 44 percent have loans, 76 per- cent of firms in the Middle East report they have accounts, and only 27 percent report FIGURE 1.6 The Use of Accounts and Loans by Firms they have loans. Within-region variations are large, too. In Azerbaijan, for example, 100 93 98 77 percent of firms have accounts and 80 20 percent have loans, while in Romania, 50 percent of firms have accounts and Firms, % 60 51 48 about 42 percent have loans. 39 40 34 Firms do differ in developed and develop- 20 ing economies in terms of their identification 0 of access to finance as a major hindrance to Firms with a checking Firms with a bank loan Firms not needing a loan their operations and growth. More so than or savings account or line of credit firms in high-income economies, firms in low- Developing economies Developed economies and middle-income economies, on average, Source: Enterprise Surveys (database), International Finance Corporation and World Bank, Wash- identify access to finance as a major problem ington, DC, http://www.enterprisesurveys.org. (figure 1.7). The lack of access to finance is Note: The sample includes 137 countries from 2005 to 2011. The income groups are based on negatively correlated with firm size in both World Bank definitions; high income refers to countries with gross national income per capita of at least $12,476. country-income types; a relatively higher share GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 27 of smaller firms identify access to finance as a FIGURE 1.7 Finance Is a Major Constraint among Firms, Especially major constraint. Small Firms About 66 percent of investments are inter- nally financed by firms. Among the external 40 35 35 sources of finance available, banks are the 30 29 25 mostly widely used option and represent, on 25 Firms, % average, 18 percent of investment financing. 20 16 15 15 Credit from suppliers and the issuance of 10 8 new equity, are also commonly used options 5 (figure 1.8). The use by firms of internal fi- 0 < 20 employees 20–99 employees > 100 employees nancing varies considerably across regions. Developing economies Developed economies For example, on average, firms in South Asia finance 75 percent of their new investments Source: Enterprise Surveys (database), International Finance Corporation and World Bank, internally, while, in Latin America, firms use Washington, DC, http://www.enterprisesurveys.org. Note: The sample includes 137 countries from 2005 to 2011. The income groups are based on internal resources to finance 58 percent of World Bank definitions; high income refers to countries with gross national income per capita of their new investment. Big differences also ex- at least $12,476. ist within regions: while an average firm in Pakistan finances 80 percent of its new invest- ments internally, the share is around 52 per- nomic instability (figure 1.9). This reflects the cent for an average Sri Lankan firm. A look at negative effects of the recent financial crisis bank financing—the most widely used exter- on aggregate demand across the globe, the nal source of finance—reveals that Pakistani reduced exports of developing countries, and firms, on average, finance less than 15 per- macroeconomic volatility. Tax rates are also cent of their new investments through banks, considered an important obstacle by firms, while, in Sri Lanka, bank financing accounts along with electricity shortages, corruption, for around 30 percent. and political instability. The lack of access The biggest constraints affecting the opera- to finance, which includes problems in both tions and growth of firms include macroeco- the availability and the cost of financing, is FIGURE 1.8 Sources of External Financing for Fixed Assets 25 24.2 20 18.3 15 Firms, % 10.8 10 6.7 6.4 5.2 5 4.0 3.9 3.1 1.8 1.9 1.8 0.9 0.6 0.4 0.3 0.5 0.8 0 Banks Supplier credit Equity shares NBFIs Informal sources New debt Small firms (< 20) Medium firms (20–99) Large firms (> 100) Source: Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. Note: The sample includes 120 countries from 2006 to 2012. The numbers in parentheses represent the number of employees. NBFI = nonbank financial institution. New debt = issuance of new debt instruments such as commercial paper and debentures. 28 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 1.9 Business Constraints 2.0 2.0 1.9 1.9 1.8 1.7 1.7 1.6 1.5 1.5 1.5 1.5 1.3 1.2 1.1 Responses, % 1.1 1.1 1.1 1.0 1.0 0.9 0.5 0.0 s s ity ty ion ity ce s ce ns er n ts ns ts d ns s or ion on tio lan te ici mi ur rd an or tio tio tio bil bil pt tit cti ra rta Co iso lat ctr rkf er fin to ta sta tra ica ula rru pe tri x dp po gu Ele ins wo dd Ta ss om nis Co un to eg l in es ns re an ce an mm rr ic gr ss of mi rc ica Tra de Ac om ing bo ce t, nin ad to ills lit co ra ef La Ac on ec ns Po Zo Sk x dt le th Ta ls ec ice Te an e, ma ro sl im rs ac or es Cr me M Inf sin sto Bu Cu Source: Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. Note: The sample includes 120 countries from 2006 to 2012. The figure shows the percentage of total responses to a question asking firms about the biggest obstacle they face. ranked as the sixth greatest obstacle faced by old get financing from informal sources (Cha- firms. Ayyagari, Demirgüç-Kunt, and Maksi- vis, Klapper, and Love 2011). The downside movic (2012) show that there is a distinction of financing through friends and family in- between what firms report and what actually cludes that this method might be unreliable constrains their growth, and that, regardless or untimely or that it might bear large non- of perception, finance represents the most financial costs. Indeed, studies find that bank constraining factor. credit is associated with higher growth rates Young firms and start-ups are particu- if one controls for other factors (for example, larly credit constrained because of principal see Ayyagari, Demirgüç-Kunt, and Maksi- agent problems. Because they have not been movic 2012). in the market for long, there is no or little Firms in the informal sector face a specific information on their performance or credit- set of issues in accessing finance. According worthiness. Data from the World Bank En- to data from World Bank surveys of firms in terprise Surveys show that, across a wide the informal sector, 47 percent of informal range of countries, younger firms rely less on sector firms do not have an account to run bank financing and more on informal financ- their business, and 88 percent do not have a ing from family and friends.13 Globally, only loan (figure 1.10). On the other hand, formal 18 percent of firms that are 1–2 years old sector firms have relatively greater access to use bank financing, compared with 39 per- bank accounts and loans. About 86 percent cent of firms that are 13 or more years old; of formal firms report having an account and in contrast, 31 percent of firms that are 1–2 47 percent claim to have loans. Differences in years old rely on informal sources of financ- the use of finance between formal and infor- ing, such as family and friends, whereas only mal sector firms can also be seen at the coun- 10 percent of firms that are 13 or more years try level. For example, in the formal sector of GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 29 Rwanda, 86 percent of firms have an account, FIGURE 1.10 Formal and Informal Firms with Accounts and Loans and 18 percent have loans, while, in the infor- mal sector, only 36 percent of firms have an 100 86 account, and 9 percent have loans. A sizable share of firms in both the in- 80 formal and formal sectors (37 percent and 42 60 53 Firms, % percent, respectively) reported they did not 47 42 need a loan. The firms that did need a loan 37 40 but did not apply for one reported the fol- lowing reasons for not applying: complex ap- 20 12 plication procedures, interest rates that were 0 too high, a lack of the required guarantees, or Firms with a bank account Firms with a loan Firms not needing a loan a higher collateral requirement (figure 1.11). Informal Formal The reasons for not applying for a loan were similar for firms in the formal and informal Sources: World Bank informal surveys; Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. sectors. Note: The sample includes 13 countries from 2008 to 2011. A study of the access of firms to finance would be incomplete without an examina- tion of access to securities markets, where some firms are able to use nonbank sources Access to securities markets tends to be rela- of funding, such as the issuance of stocks or tively easier for smaller firms in high-income bonds. Map 1.3 illustrates the differences in countries. It is also relatively easier in large the access of firms to the stock market by us- economies, such as those of China and India. ing the share of market capitalization outside But, even in such economies, there are serious the top 10 largest companies: the higher this disparities between larger and smaller firms; share, the easier it generally is for smaller the larger ones are much more likely to issue firms to issue securities in this market. In most new equity (Didier and Schmukler 2013). In countries, the top 10 issuing firms capture a contrast, smaller, lower-income economies ei- large fraction of the total amount raised. ther have no securities markets to speak of, or FIGURE 1.11 Reasons for Not Applying for a Loan 20 Firms that have not applied for loans, % 18 17 17 16 15 12 12 10 8 8 5 4 4 3 0 Complex High Lacking Not Other High High Complex Wouldn't Informal Loan size applications interest guarantees registered interest collateral applications be payments and rates rates approved to get maturity bank loans Informal Formal Sources: World Bank informal surveys; Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. Note: The sample includes 13 countries from 2008 to 2011. 30 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP 1.3 Access by Firms to Securities Markets Source: Global Financial Development Report (database), World Bank, Washington, DC, http://www.worldbank.org/financialdevelopment. Note: The map is based on the share in market capitalization of firms that are outside the top 10 largest companies. The World Federation of Exchanges provided data on individual exchanges. The variable is aggregated up to the country level by taking a simple average over exchanges. The four shades of blue in the map are based on the average value of the variable in 2008–11: the darker the blue, the higher the quartile of the statistical distribution of the variable. are characterized by markets in which most Demirgüç-Kunt and Klapper (2012) show of the capitalization is concentrated among a that within-country inequality in the use of small number of the largest issuers, while ac- formal accounts is correlated with the coun- cess by smaller firms is limited. try’s income inequality. They find a relatively high correlation between account penetration CORRELATES OF FINANCIAL and the Gini coefficient as a proxy for income INCLUSION AND EXCLUSION inequality. This association seems to hold even if one controls for national income and Income seems to matter, as does other variables. inequality Nonetheless, these correlations need to Country-level income, approximated by GDP be taken with a grain of salt. In-depth analy- per capita, seems to account for much of ses suggest that other factors, especially the the massive variation in account penetration quality of institutions in an economy, drive worldwide. In most countries with GDP per account penetration as well as income level capita above $15,000, account penetration and income inequality. One therefore needs is essentially universal (notable exceptions to look beyond basic factors in explaining the are Italy and the United States, with account variance in account penetration. penetration of 71 percent and 88 percent, re- spectively). Indeed, national income explains Geography matters, too 73 percent of the variation around the world in the country-level percentage of adults with Within individual countries, the use of finan- a formal account.14 cial services is often rather uneven: densely Account ownership also appears to populated urban areas show a much higher go hand in hand with income equality. density in retail access points (such as bank GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 31 MAP 1.4 Geography Matters: Example of Subnational Data on Financial Inclusion U N I T E D S TAT E S O F A M E R I C A Baja California Sonora Chihuahua Coahuila de Zaragoza Baja Nuevo California Gulf of Mexico León Sur Durango Sinaloa Tamaulipas Zacatecas Aguascalientes San Luis Querétaro Arteaga Nayarit Potosí Veracruz de Ignacio de la Llave Yucatán Guanajuato Distrito Federal Jalisco 280 – 750 Formal accounts Hidalgo MEXICO CITY Tlaxcala Quintana Roo Michoacán Campeche 240 – 280 per 100 adults Colima de Ocampo México Puebla Tabasco 280–750 BELIZE 200 – 240 240–279 Guerrero Oaxaca Chiapas 200–239 160 – 200 160–199 PA C I F I C Morelos GUATEMALA 130–159 OCEAN HONDURAS IBRD 40409 130 – 160 EL SEPTEMBER 2013 SALVADOR Source: Calculations based on data from Comisión Nacional Bancaria y de Valores, Mexico. Note: The indicator is the sum of all formal accounts, including deposit accounts, savings accounts, open market transaction accounts, debit card accounts, credit card accounts, and payroll transaction accounts. branches, ATMs, and agents) and a greater Map 1.4 underscores the significance of ge- use of financial services than rural areas. De- ography in financial inclusion by illustrating spite the growth in mobile money and other it through granular data on Mexico. Formal recent technologies, being near a retail access financial services are in much greater use in point is still important for the use of finan- the capital city district (the small central area cial services by individuals. It is especially in black) than elsewhere in the country. The important for the poor, who are less mobile number of accounts per capita is relatively and have less access to modern technologies. higher in the north of the country and in areas Building retail access points is costly for fi- with an active tourism industry. More gener- nancial service providers; so, they are keen ally, urban areas show much greater numbers to place these points optimally in relation of accounts, reflecting factors such as higher to their existing and potential customers. Fi- population density and greater proximity to nancial inclusion advocates also want broad retail access points. coverage of the population by financial access points, but would like to see that certain types of customers, such as the poor, also have ac- Financial inclusion vs. depth, efficiency, cess, a concern not always shared by provid- and stability ers. This leads to major policy questions— Large amounts of credit in a financial sys- such as the identification of policies that tem—both commercial and consumer cred- work and how much policy interventions can it—do not always correspond to the broad use and should push providers to offer access in of financial services because the credit may be places in which they do not want to provide concentrated among the largest firms and the access—that are discussed in greater depth in wealthiest individuals. Indeed, the use of for- chapter 2.15 mal accounts is imperfectly correlated with a 32 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 1.12 Financial Inclusion vs. Depth, Efficiency, and Stability (Financial Institutions) Financial Institutions a. Depth b. (In)efficiency c. Stability 300 50 60 Lending minus deposit rate, percentage points 250 50 Private sector credit to GDP, % 40 200 40 30 Z-scorea 150 30 R 2 = 0.0012 R 2 = 0.5805 2 100 R 2 = 0.2204 20 10 50 10 0 0 0 0 50 100 0 50 100 0 50 100 Population with account, % Population with account, % Population with account, % Financial Markets d. Depth e. Efficiency f. Stability 300 300 80 70 250 250 Market capitalization to GDP, % Market turnover/capitalization Standard deviation/average price 60 R 2 = 0.0065 200 200 50 R 2 = 0.2733 150 150 40 2 R = 0.11178 30 100 100 20 50 50 10 0 0 0 0 50 100 0 20 40 60 80 0 20 40 60 80 Market cap outside top 10, % Market cap outside top 10, % Market cap outside top 10, % Source: Global Financial Development Report (database), World Bank, Washington, DC, http://www.worldbank.org/financialdevelopment. Note: Each point corresponds to a country; averages are for 2008–11. For definitions and a discussion of the 4x2 matrix for benchmarking financial systems, see Cˇ ihák and others (2013). “Market cap” is short for market capitalization. a. The z-score is a ratio, defined as (ROA + equity)/assets)/sd(ROA ), where ROA is average annual return on end-year assets and sd(ROA ) is the standard deviation of ROA. common measure of financial depth: domestic GDP), has relatively high account penetration credit to the private sector as a percentage of (81 percent). This suggests that financial depth GDP (figure 1.12, panel a). Country examples and financial inclusion are distinct dimensions bear this out. Vietnam has domestic credit to of financial development and that financial the private sector amounting to 125 percent systems can become deep while showing low of GDP, but only 21 percent of the adults degrees of inclusion.16 in the country report they have a formal ac- The greater use of formal accounts is asso- count. Conversely, the Czech Republic, with ciated with higher efficiency in financial insti- relatively modest financial depth (with domes- tutions. This is illustrated in panel b of figure tic credit to the private sector at 56 percent of 1.12 by the negative correlation between the GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 33 FIGURE 1.13 Correlates of Financial Inclusion 10 7 Effect on probability of owning an account, % 5 3 2 0 –2 –5 –3 –6 –7 –8 –10 –9 –13 –12 –15 –16 –20 Poorest Second Middle Fourth Age Rural 0–8 yrs of Log of Married Employed Unemployed Out of 20% 20% 20% 20% education household workforce size Source: Based on Allen, Demirgüç-Kunt, and others 2012. Note: The figure shows probit regression results of a financial inclusion indicator on country fixed effects and a set of individual characteristics, for 124,334 adults (15 years and older) covered by the Global Financial Inclusion (Global Findex) Database (http://www.worldbank.org/globalfindex) in 2011. The financial inclusion indicator is a 0/1 variable indi- cating whether a person had an account at a formal financial institution in 2011. See Allen, Demirgüç-Kunt, and others (2012) for definitions, data sources, standard errors of the parameter estimates, additional estimation methods, and additional regressions for other dependent variables (savings and the frequency of use of accounts). account penetration rate and the lending- massive data set underlying the Global Findex deposit spread. This relationship is robust with database provides a unique source of informa- respect to different measures of efficiency and tion on what drives financial inclusion around different proxies for account penetration. the world.17 Allen, Demirgüç-Kunt, and oth- Finally, there is no significant correlation ers (2012) subject the data set to a battery of between account penetration and financial statistical tests to address the question of what stability. This is illustrated in panel c of figure drives financial inclusion. They consider a host 1.12. The result holds up also for alternative of other country-level characteristics and poli- measures of financial stability. For example, cies as potential determinants of account use. if one compares the crisis and noncrisis coun- Their analysis focuses on three indicators of tries as identified by Laeven and Valencia account use: (1) ownership of an account, (2) (2012), the degree of account use is, on aver- use of the account to save, and (3) frequent age, not significantly different. use of the account, defined as three or more Cross-country data on financial markets withdrawals per month. They find that these provide a similar picture: financial inclusion is indicators are associated with a better enabling associated positively with depth and efficiency environment for accessing financial services, (figure 1.12, panels d and e) while having no such as lower banking costs, greater proximity significant association with stability (figure to financial providers, and fewer documenta- 1.12, panel f). tion requirements to open an account. People who are poor, young, unemployed, out of the workforce, or less well educated, or who live in An econometric examination of financial rural areas are relatively less likely to have an inclusion account (figure 1.13). Policies targeted at en- What can data on some 124,000 people tell us hancing financial inclusion—such as offering about the drivers of financial inclusion? The basic or low-fee accounts, granting exemptions 34 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 from onerous documentation requirements, al- member already has an account. The other lowing correspondent banking, and using bank reasons reported (in order of importance) are accounts to make government payments—are bank accounts being too expensive, excessive especially effective among those people who distance of banks, lack of the necessary docu- are most likely to be excluded: the poor and mentation, lack of trust in banks, and religious rural residents. reasons. Adults in developing economies are significantly more likely to cite distance, cost, documentation, and lack of money compared BARRIERS TO FINANCIAL with adults in developed economies. INCLUSION The second most frequently cited reason was that another member of the family al- Income levels and individual characteristics ready has an account, a response that identi- clearly help explain some of the differences in fies indirect users. Women and adults living the use of accounts around the world. What in high-income and upper-middle-income do people say when asked why they do not economies (where relatives are most likely have an account? The Global Findex survey to have an account) were substantially more provides novel data on the barriers to finan- likely to choose this reason. A recent study cial inclusion, based on the responses of more shows that lack of account ownership (and than 70,000 adults who do not have a formal personal asset accumulation) limits women’s account.18 ability to pursue self-employment opportuni- Globally, the reason most frequently cited ties (Hallward-Driemeier and Hasan 2012). for not having a formal account is the lack Hence, while such voluntary exclusion may of enough money to have and use one (figure be linked to individual preferences or cul- 1.14). Thirty percent of adults who are with- tural norms, it may also indicate a lack of out a formal account cite this as the only rea- awareness of financial products or a lack of son. The next most commonly cited reason for financial literacy more generally.19 not having an account is that another family Affordability is a key barrier to account ownership. High costs are cited by a quarter FIGURE 1.14 Reported Reasons for Not Having a Bank Account of unbanked respondents, on average, and by 32 percent of unbanked respondents in low- income economies. Fixed transaction costs Not enough money 30 and annual fees tend to make small transac- tions unaffordable for large parts of the popu- Family member already has account 25 lation. For example, annual fees on a check- ing account in Sierra Leone are equivalent to Too expensive 23 27 percent of GDP per capita. Not surpris- Too far away 20 ingly, 44 percent of adults without accounts in the country cite high cost as a reason for Lack of documentation 18 not having a formal account. Analysis finds a significant relationship between cost as a Lack of trust 13 self-reported barrier and objective measures of costs (Demirgüç-Kunt and Klapper 2012). Religious reasons 5 Even more importantly, the high costs of opening and maintaining accounts are asso- 0 5 10 15 20 25 30 35 ciated with a lack of competition and un- Adults without an account, % derdeveloped physical or institutional infra- structure in a country. This is a key point: it Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. suggests that the high costs associated with Note: Respondents could choose more than one reason. a small account do not simply represent the GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 35 fixed costs of provision, but there is now evi- discrimination against certain segments of dence that some of the costs are associated the population, past episodes of government with underlying distortions (Allen, Demirgüç- expropriation of banks, or economic crises Kunt, and others 2012). and uncertainty. In Europe and Central Asia, Twenty percent of unbanked respondents 31 percent of non–account-holders cite lack cited distance as a key reason they did not of trust in banks as a reason for not having an have a formal account. The frequency with account, a share almost three times the share which this barrier was cited increases sharply in other regions, on average. as one moves down the income level of coun- About 5 percent of unbanked respondents tries, from 10 percent in high-income econo- cite religious reasons for not having a formal mies to 28 percent in low-income economies. account. The proportion is higher in some Among developing economies, there is a sig- Middle Eastern and South Asian economies, nificant relationship between distance as a where the development of financial products self-reported barrier and objective measures compatible with religious beliefs (in particu- of providers, such as bank branch penetra- lar, Islamic finance) could potentially increase tion. To illustrate this point: Tanzania has a account penetration and the use of various fi- large share of non–account-holders who cite nancial services (box 1.4). distance as a reason they do not have an ac- count—47 percent—and also ranks near the bottom in bank branch penetration, averag- FINANCIAL SECTOR ing less than 0.5 bank branches per 1,000 STRUCTURE, COMPETITION, square kilometers. AND INCLUSION The documentation requirements for open- ing an account may exclude workers in the Which type of financial sector structure rural or informal sector who are less likely to works best in reaching out to a broad set of have wage slips or formal proof of residence. individuals and firms? In terms of financial There is a significant relationship between inclusion and economic development, how subjective and objective measures of docu- do financial systems based on banks compare mentation requirements as a barrier to ac- with those based on financial markets? Stud- count use that holds even after one takes into ies consistently find that what matters for eco- account GDP per capita (Demirgüç-Kunt and nomic growth is the overall development of Klapper 2012). Indeed, the Financial Action the financial system, rather than the relative Task Force, an intergovernmental body aimed shares of banks and financial markets. In oth- at combating money laundering, terrorist fi- er words, at similar levels of financial devel- nancing, and related threats to the integrity of opment, more highly bank- or market-orient- the international financial system, recognized ed economies are not associated with greater that overly cautious safeguards against money economic growth rates, industry growth, or laundering and terrorist financing can have the the access of firms to external finance (Beck unintended consequence of excluding legiti- and Levine 2004; Demirgüç-Kunt and Maksi- mate businesses and consumers from financial movic 2002; Levine 2002). systems. It has therefore called for such safe- While the structure of the financial system guards that also support financial inclusion.20 does not seem to be associated with growth Distrust in formal financial institutions is outcomes, Demirgüç-Kunt and Maksimovic a nontrivial barrier to wider financial inclu- (2002) find that this does affect the types of sion and one that is difficult to address in firms and projects that are financed, which the short term. Thirteen percent of adults has strong implications for financial inclu- without a formal account cite lack of trust in sion. By looking at firm-level data for 40 banks as a reason for not having an account. countries, they conclude that, in economies This distrust can stem from cultural norms, with more well developed capital markets, the 36 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 1.4 Islamic Finance and Inclusion Shari‘a-compliant financial products and instru- religiously minded Muslim individuals and fi rms in ments can play a significant role in enhancing fi nan- need of fi nancing. cial inclusion among Muslim populations. About Based on a 2010 Gallup poll, about 90 percent of 700 million of the world’s poor live in predomi- the adults residing in Organization of Islamic Coop- nantly Muslim-populated countries. In recent years, eration (OIC) member countries consider religion an there has been growing interest in Islamic fi nance as important part of their daily lives (Crabtree 2010). a tool to increase fi nancial inclusion among Muslim This may help explain why only about 25 percent populations (Mohieldin and others 2011). of adults in OIC member countries have an account The main issue relates to the fact that many in formal fi nancial institutions, which is below the Muslim-headed households and micro, small, and global average of about 50 percent. Also, while 18 medium enterprises may voluntarily exclude them- percent of non-Muslim adults in the world have selves from formal financial markets because of formal saving accounts, only 9 percent of Muslim Shari‘a requirements. Islamic legal systems, among adults have these accounts (Demirgüç-Kunt, Klap- other characteristics, prohibit predefi ned interest- per, and Randall, forthcoming). Moreover, 4 percent bearing loans. They also require fi nancial provid- of respondents without a formal account in non-OIC ers to share in the risks of the business activities countries cite religious reasons for not having an for which they provide financial services (profit account, compared with 7 percent in OIC countries and loss sharing). Given these requirements, most (table B1.4.1) and 12 percent in the Middle East and conventional fi nancial services are not relevant for North Africa. TABLE B1.4.1 OIC Member Countries and the Rest of the World % of respondents, unless otherwise indicated All OIC countries Non-OIC countries Have an account at a formal financial institution* 50 25 57 Reason for not having an account religious reasons* 5 7 4 distance* 20 23 19 account too expensive* 25 29 23 lack of documentation* 18 22 16 lack of trust 13 13 13 lack of money* 65 75 61 family member already has an account* 23 11 28 Source: Calculations based on the Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. * The means t -test between the Organization of Islamic Cooperation (OIC) and non-OIC countries is significant at the 1 percent level. Muslim countries are far from uniform in terms having an account (table B1.4.2). This correlation of fi nancial inclusion. For example, 34 percent of is particularly strong if one focuses on the group the unbanked Afghan population cite religious rea- of OIC countries and, even more, on those OIC sons for not having an account in a formal fi nancial countries that show a religiosity index exceeding institution, while only 0.1 percent of Malaysians 85 percent. do so, although both countries have similarly high Based on the Global Findex, for religious rea- Gallup religiosity indexes (97 percent and 96 per- sons, some 51 million adults in the OIC countries do cent, respectively; see the Statistical Appendix). a not have accounts in a formal fi nancial institution.b This can be traced to the extent to which Islamic Given that a majority of the OIC population lives in fi nancial institutions are present in a given country. poverty, Islamic microfi nance could be particularly An analysis suggests that the size of Islamic assets attractive. For example, 49 percent and 54 percent per adult population is negatively correlated with of adults in Algeria and Morocco, respectively, pre- the share of adults citing religious reasons for not fer to use Islamic loans even if these loans are more (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 37 BOX 1.4 Islamic Finance and Inclusion (continued) TABLE B1.4.2 Islamic Banking, Religiosity, and Household Access to Financial Services OIC countries Indicator All countries OIC countries with religiosity > 85% Non-OIC countries OIC dummy 5.79*** .. .. .. GDP per capita, US$, 1,000s 0.02 0.38** 0.43** −0.005 Islamic assets per adult, US$, 1,000s −0.18* −0.61*** −0.65** −3.85 Observations 137 41 32 96 R -squared 0.21 0.06 0.08 0.00 Sources: Based on the Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex; World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators; Bankscope (database), Bureau van Dijk, Brussels, http://www.bvdinfo.com/en-gb/products/company-information/international/bankscope. Note: Dependent variable: percentage of adults citing religious reasons for not having an account. Regressions include a constant term. Robust standard errors are reported. .. indicates that the variable could not be included in the regression. OIC = Organization of Islamic Cooperation. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. expensive than conventional loans (Demirgüç-Kunt, well as the astonishing growth of the overall Islamic Klapper, and Randall, forthcoming). fi nance industry, all point to the growing attractive- Global surveys on Islamic microfinance com- ness of Shari‘a-compliant financial products and the pleted by the Consultative Group to Assist the Poor supply shortage of such products.c (CGAP) in 2007 and 2012 provide some initial Religiosity also has an impact on the access of insights into the rapidly growing Islamic microfi- fi rms to fi nance in OIC countries. The number of nance industry. The 2007 CGAP survey found fewer Islamic banks per 100,000 adults is negatively corre- than 130 and 500,000 Islamic MFIs and custom- lated with the proportion of fi rms identifying access ers, respectively (Karim, Tarazi, and Reille 2008). to fi nance as a major constraint. The negative corre- Within five years, these figures more than doubled, lation is greater if one focuses on OIC countries and reaching 256 MFIs and 1.3 million active clients (El- greater still if one focuses on a subset of OIC coun- Zoghbi and Tarazi 2013). These figures are on the tries with a religiosity index above 85 percent (table conservative side because they are based on data for B1.4.3). These fi ndings, which are mainly driven by 16 of the 57 OIC member countries (excluding econ- small fi rms (fi gure B1.4.1), suggest that increasing omies such as the Islamic Republic of Iran, Malay- the number of Shari‘a-compliant fi nancial institu- sia, and Turkey, which have active Islamic fi nance tions can make a positive difference in the opera- industries). In short, the estimated unmet demand tions of small fi rms (0–20 employees) in Muslim- for Shari‘a-compliant financial products, in con- populated countries by reducing the access barriers junction with the rapid growth of Islamic MFIs, as to formal fi nancial services. TABLE B1.4.3 Islamic Banking, Religiosity, and Firm Access to Financial Services Indicator All countries OIC countries OIC countries with religiosity > 85% OIC dummy 8.59** .. .. GDP per capita, US$, 1,000s −1.23*** −6.12*** −5.79*** Islamic banks per 100,000 adults −52.70* −61.97* −108.76** Observations 107 32 24 R -Squared 0.25 0.35 0.38 Sources: Calculations based on Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys .org; World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators; Bankscope (database), Bureau van Dijk, Brussels, http://www.bvdinfo.com/en-gb/products/company-information/international/bankscope. Note: Dependent variable: percentage of firms identifying access to finance as a major constraint. All regressions include a constant term. Robust standard errors are reported. .. indicates that the variable was not included in the regression. OIC = Organization of Islamic Cooperation. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. (box continued next page) 38 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 1.4 Islamic Finance and Inclusion (continued) FIGURE B1.4.1 Islamic Banking, Religiosity, and Access of Firms to Financial Services 0 number of Islamic banks per 100,000 adults –25 Size of the regression coefficients for –75 –125 –175 All countries OIC countries OIC countries with religiosity index > 85% All firms Small firms Medium firms Sources: Calculations based on Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org; World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators; Bankscope (data- base), Bureau van Dijk, Brussels, http://www.bvdinfo.com/en-gb/products/company-information/international/bankscope. Note: The dependent variable is the percentage of firms identifying access to finance as a major constraint. All regressions include a constant term and GDP per capita. Robust standard errors are used. Only coefficients significant at 10 percent or less are included in the figure. OIC = Organization of Islamic Cooperation. Efforts to increase fi nancial inclusion in jurisdic- that are operating based on Shari‘a specifications tions with Muslim populations thus require sustain- and institutions that are not. Another difficulty has able mechanisms to provide Shari‘a-compliant fi nan- been the lack of information and training on Islamic cial services to all residents, especially the Muslim finance. For example, only about 48 percent of poor, estimated at around 700 million people who adults in Algeria, Egypt, Morocco, Tunisia, and the are living on less than $2 per day. One obstacle is the Republic of Yemen have heard about Islamic banks lack of transparency and the absence of a broadly (Demirgüç-Kunt, Klapper, and Randall, forthcom- accepted standardized process for assessing the ing). Finally, in their infancy and smaller in scale, compliance of financial institutions with Shari‘a Islamic fi nancial products tend to be more expensive guidelines, which makes it difficult for many indi- than their conventional counterparts, reducing their viduals to distinguish between fi nancial institutions attractiveness. a. The Gallup religiosity index captures the percentage of adults who responded affi rmatively to the question “Is religion an important part of your daily life?” in a 2009 Gallup poll (Crabtree 2010). The question does not distinguish which type of religion (and the analysis presented here may apply to other religions as well). The reliability of this index depends on how truthfully people respond to the survey question. Nonetheless, the variable has been used in previous research. b. Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org /globalfi ndex. c. Globally, Islamic fi nancial assets have more than doubled since 2006 (Mohieldin and others 2011). See also Cihák ˇ and Hesse (2010) on the growth and stability of the Islamic fi nancial sector. use by firms of long-term financing is greater. Consistent with these results, Demirgüç- In contrast, firms rely on short-term financ- Kunt, Feyen, and Levine (2011) find that fi- ing to a much larger extent in countries with nancial structures evolve with economic de- more financial structures that are bank-based. velopment because capital markets provide GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 39 FIGURE 1.15 Ratio of Cooperatives, State Specialized Financial Institutions, and Microfinance Institution Branches to Commercial Bank Branches 700 per 1,000 commercial bank branches 600 Financial institution branches 500 400 480 300 49 146 44 22 200 74 31 269 53 209 149 100 37 138 58 4 104 71 44 35 61 0 Upper-income East Asia Europe and Latin America and Middle East and South Sub-Saharan countries and Pacific Central Asia the Caribbean North Africa Asia Africa Cooperatives State specialized financial institutions Microfinance institutions Source: Financial Access (database) 2010, Consultative Group to Assist the Poor and World Bank, Washington, DC, http://www.cgap.org/data/financial-access-2010-database-cgap. financial services that are different than the Countries also differ in the mix of financial services provided by banks. Particularly, the institutions constituting the market because significance of market-based financing in- certain types of institutions are more preva- creases relative to bank-based financing. They lent in particular regions (figure 1.15). In show that, as economies develop, the services the Middle East and North Africa, for every provided by securities markets become more 1,000 commercial bank branches, there are important for economic activity, while those 35 cooperatives, 209 state specialized finan- provided by banks become less important. cial institutions, and 44 MFI branches. How- Another part of this literature has moved ever, in Sub-Saharan African countries, there from the market- vs. bank-financing analysis are 480 MFIs for every 1,000 commercial to explore a different angle of financial diver- banks, and only 61 cooperative and 37 public sity, more focused on the types of financial bank branches. Countries in Latin America institutions (such as niche banks, coopera- and East Asia and the Pacific are more reliant tives, and MFIs) and their link with access to on cooperatives than are developing countries finance and with economic growth. in other regions. The ratio of cooperatives to In developing economies, the financing bank branches is the highest in upper-income needs of a large fraction of households and countries, where there are 269 cooperatives enterprises are supplied by alternative finan- per 1,000 commercial bank branches. This cial institutions such as cooperatives, credit pattern is particularly strong in Western Euro- unions, MFIs, and factoring or leasing com- pean economies, where financial systems rely panies. While banks are the most prominent more on these institutions.22 institution across regions, their relative im- Some argue that a comparative advantage portance varies substantially. For instance, for of institutions such as cooperative banks each MFI, there are 46 bank branches in Eu- may be that they rely on more flexible lend- rope and Central Asia, 32 in Latin America, ing technologies, making it possible to ex- and only 2 in Sub-Saharan Africa.21 tract information about more opaque clients, 40 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 such as micro, small, and medium enter- find such enterprises profitable for several prises or households that do not have avail- reasons. One such reason is the increased able the type of information or documenta- use of different transactional technologies tion that banks traditionally request (Berger, that benefit from the economies of scale of Klapper, and Udell 2001; Stein 2002). Oth- larger institutions (for example, the use of ers argue that banks can extend financing credit scoring models requires a large pool of to more opaque clients by applying differ- clients, thereby benefiting from larger bank ent transactional technologies that facilitate size). arm’s-length lending and that, through more Relationship lending is one of several competition, improved financial infrastruc- other ways in which banks extend financ- ture, and appropriate incentives, banks can ing to more opaque clients (Berger and be encouraged to reach out for new clients Udell 2006). Two examples of large banks (Berger and Udell 2006; de la Torre, Gozzi, that have adopted innovative lending tech- and Schmukler 2007). nologies and business models are Banco A study by Beck, Demirgüç-Kunt, and Azteca in Mexico and BancoSol in Bolivia. Singer (forthcoming) explores the role of dif- BancoSol is arguably the first commercial ferent kinds of financial institutions, as well bank to specialize in microfinance. Its lend- as their average size, in easing the access of ing technology relies on a solidarity group firms to financial services and finds heteroge- lending strategy, whereby members organize neous impacts across countries and firm sizes. small joint liability credit groups, and the More specifically, in low-income countries, bank lends simultaneously to all group mem- a higher share of low-end financial institu- bers (Gonzalez-Vega and others 1997). Ban- tions (such as cooperatives, credit unions, co Azteca, on the other hand, targets clients and MFIs) and specialized lenders (such as employed in the informal sector by using factoring and leasing companies) is associated their durable goods as collateral for loans. with better access to finance. The evidence Ruiz (2013) shows that, in municipalities in the authors present indicates that the average which this bank has opened a branch, house- size of financial institutions matters for inclu- holds with members working in the informal sion. In contrast to earlier studies suggesting sector are more likely to borrow from banks, that smaller financial institutions are better are less likely to rely on more expensive able to serve the credit needs of small, opaque credit suppliers, and are thus better able to borrowers (for example, Berger and Udell smooth their consumption and accumulate 1995; Keeton 1995), the authors reject that more valuable durable goods. smaller institutions are better at easing the ac- Beyond financial structure, evidence cess of firms to finance. On the contrary, in points to competition in the financial sector countries with low levels of GDP per capita, as a key factor in enhancing financial inclu- larger banks and low-end institutions seem sion. Examining firm-level data for 53 coun- to improve access to financial services, and tries from 2002 to 2010, Love and Martínez larger banks and specialized lenders seem to Pería (2012) find that bank competition sub- facilitate access to loans and overdraft use by stantially increases the access of firms to fi- smaller enterprises. nance. An advantage of their study relative These results are in line with the findings to others is that their data allow them to of more recent studies. For instance, Berger, isolate within-country variations in competi- Klapper, and Udell (2001) find that larger tion and access to finance more effectively. banks may be as well equipped as smaller In a more microlevel analysis, Lewis, Mo- ones to serve small clients because they use rais, and Ruiz (2013) examine competition a different lending technology. De la Torre, among Mexican banks. Their results show Gozzi, and Schmukler (2007) likewise find that large banks are more likely to engage in that, contrary to the belief that large banks less competitive practices and confirm that, are less capable of reaching out to opaque as expected, competition leads to less collu- small and medium enterprises, most banks sive practices. Importantly, less competition GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 41 disproportionally affects access to finance McKee, Lahaye, and Koning 2011). Financial among smaller firms. inclusion does not mean increasing access for To summarize, recent studies indicate more the sake of access, and it certainly does not financially diverse markets are associated mean making everybody borrow. with improved access to finance. Policy rec- ommendations to support a more financially The effects of savings and payments diverse landscape encompass (1) improv- ing competition within the financial system, Newly available global data point to a strong but also allowing a variety of financial in- correlation between income inequality and stitutions to operate (from specialized lend- inequality in the use of bank accounts. For ers and low-end institutions to banks with example, in Sweden—a country with one lending technologies or business models that of the most even income distributions in the reach out to new clients in responsible ways); world—the share of people having bank ac- (2) strengthening financial and lending infra- counts is the same for the rich and the poor. structure, including commercial laws, bank- On the other end of the spectrum are coun- ruptcy laws, and contract enforcement; and tries such as Haiti, where income inequality (3) creating the conditions for capital market is very high and where the richest 20 percent development to improve access to longer-term are about 14 times more likely to have a bank finance. account than the poorest 20 percent. Figure 1.16 illustrates that across a broad spectrum of countries, this measure of inequality in THE REAL EFFECTS OF account penetration (financial inequality) is FINANCIAL INCLUSION ON THE closely correlated with income inequality (the POOR: EVIDENCE correlation coefficient being 0.33). Our discussion has so far focused on why fi- It thus appears that financial inequality and nancial inclusion is important for development income inequality go hand in hand. The strong in economic theory and on defining financial relationship holds even when controlling for inclusion, measuring it, explaining what drives national income. However, the correlation is it, and examining the barriers that limit it. In not perfect. For example, the measure of finan- the following, we turn to the empirical evi- cial inequality in the Philippines is very close dence on the relationship between financial in- to that of Haiti, but incomes are much more clusion and economic development. evenly distributed in the Philippines. Moreover, Recent empirical evidence on the impact of the correlations only suggest that financial and financial inclusion on economic development economic inequality are closely associated, not and poverty varies by the type of financial ser- necessarily that one causes the other. vice in question.23 In the access to basic pay- Field experiments provide more direct evi- ments and savings, the evidence on benefits, dence about the causal linkages between ac- especially among poor households, is quite cess to savings and payments services and real- supportive. In insurance products, there is economy variables. For example, a range of also some evidence of a positive impact. For randomized controlled experiments finds that firms, particularly small and young firms that providing individuals with access to savings ac- face greater constraints, access to finance is counts or simple informal savings technologies associated with innovation, job creation, and increases savings (Aportela 1999; Ashraf, Kar- growth. However, in access to microcredit, lan, and Yin 2006), women’s empowerment the data on dozens of microcredit experi- (Ashraf, Karlan, and Yin 2010), productive in- ments and from other cross-country research vestment (Dupas and Robinson 2011, 2013), paint a rather mixed picture (for example, see consumption, investment in preventive health, Bauchet and others 2011; Roodman 2011).24 productivity, and income (Ashraf, Karlan, and A common message of the underlying re- Yin 2010; Dupas and Robinson 2013).25 search is that effective financial inclusion The findings from the randomized con- means responsible inclusion (for example, trolled experiments are in line with those of 42 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 1.16 Correlation between Income Inequality and Inequality in the Use of Financial Services 70 Haiti 60 Income inequality (Gini coefficient) 50 Philippines 40 30 Sweden 20 0 5 10 15 Account penetration in the richest 20% as a multiple of that in the poorest 20% Source: Calculations based on Demirgüç-Kunt and Klapper 2012; World Development Indicators (database), World Bank, Washington, DC, http://data .worldbank.org/data-catalog/world-development-indicators. Note: Higher values of the Gini coefficient mean more inequality. Data on Gini coefficients are for 2009 or the latest available year. Account penetration is the share of adults who had an account at a formal financial institution in 2011. several other studies. For example, an in- more nuanced, but, on balance, still positive. depth examination of the effect of bank de- Evidence based on total volumes of written in- regulation in the United States shows that surance premiums casts doubts on the aggre- greater financial inclusion accelerates eco- gate impact. Ward and Zurbruegg (2000), us- nomic growth, intensifies competition, and ing cointegration analysis for nine countries of boosts the demand for labor. It is also usually the Organisation for Economic Co-operation associated with relatively bigger benefits to and Development (OECD) from the 1960s those people at the lower end of the income to the 1990s, find that, in some countries, the distribution, thus contributing to inclusive insurance industry Granger causes economic growth (Beck, Levine, and Levkov 2010). growth, while, in other countries, the reverse Together, these studies provide robust is true. They conclude that the relationship justification for policies that encourage the between insurance and growth is nation spe- provision of basic accounts for savings and cific. Subsequent authors (for example, Kugler payments.26 Increasing financial inclusion in and Ofoghi 2005) have pointed out that it is terms of savings and payments, if done well, possible to have cointegration at the aggregate can both help reduce extreme poverty and level but not at the disaggregate level, and vice boost shared prosperity. versa; so, looking more closely at the disaggre- gated data is important. Recent evidence from disaggregated data The effects of insurance is encouraging. In particular, Cai and others For insurance products, the evidence on the (2010) have conducted a large randomized impact on economic development is slightly experiment in southwestern China to assess GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 43 the impact of insurance on sows. They find associated with growth (Ayyagari, Demirgüç- that providing access to formal insurance sig- Kunt, and Maksimovic 2008). There is also nificantly increases the propensity of farmers evidence that returns on capital are high (de to raise sows. In another study, Cole, Giné, Mel, McKenzie, and Woodruff 2008, 2012a). and Vickery (2012) examine how the avail- Improving access to finance for potential ability of rainfall insurance affects the invest- entrepreneurs therefore promises significant ment and production decisions of small- and welfare gains not only for the entrepreneurs, medium-scale Indian farmers. They observe but also for society as a whole. little effect on total expenditures. However, Some evidence indicates that access to cred- they find that increased insurance induces it is associated with a decline in observable farmers to substitute production activities measures of poverty. For example, Burgess toward high-return high-risk cash crops. and Pande (2005) suggest that bank branch- Finally, Shapiro (2012) evaluates the effects ing regulation in India has had a substantial of a Mexican government disaster relief pro- impact on poverty reduction. Between the gram with insurance-like features. Specifi- 1970s and the 1990s, India’s bank branch- cally, the program provides fixed indemnity ing regulations required banks to open four payments to rural households the crops or branches in unbanked locations for every new assets of which have been damaged by a nat- branch opened in an urban area. This led to ural disaster. The evaluation finds that the the establishment of up to 30,000 rural bank availability of insurance against losses from branches, and the study provides direct evi- natural disasters changes how rural house- dence that this expansion of the bank branch holds invest in their farms. In particular, in- network had a positive impact on financial sured farmers utilize more expensive capital inclusion and contributed to a considerable inputs and purchase better seeds. decline in rural poverty. At the same time, bearing in mind that India’s branch expan- sion was a government initiative in a banking The effects of credit system dominated by state-owned banks, one Economic theory suggests that improved ac- ought to ask whether the cost-benefit calcu- cess to credit can have positive implications lation for this program exceeds that of other for poverty alleviation and entrepreneurial ac- forms of government assistance and to what tivity. Better access to credit makes it easier for extent the benefits of this policy need to be households to smooth out consumption over adjusted for the cost of the longer-run credit time and provides a de facto insurance against market distortions arising from India’s social many of the common risks facing households banking experiment. After all, the program and small enterprises in the developing world. was discontinued in 1991 in part because of By the same token, improved access to credit the high default rates among rural branches. can also encourage entrepreneurial activity by New research provides insights into the attenuating investment constraints and mak- impact of access to credit on poverty allevia- ing it easier for small businesses to grow be- tion and into the channel through which this yond subsistence. is likely to occur, that is, the greater propen- There is ample evidence that limited sity of individuals to transition into entrepre- access to credit poses a substantial obstacle neurship. In a recent study, Bruhn and Love to entrepreneurship and firm growth, espe- (forthcoming) evaluate the opening of a new cially among small and young firms (Banerjee commercial bank, Banco Azteca, focused on and Duflo 2007; Beck, Demirgüç-Kunt, and low- and middle-income borrowers in Mex- Maksimovic 2005; Beck and others 2006; ico. They find that the opening of Banco Az- Evans and Jovanovic 1989). Cross-country teca, which represented a significant improve- research analyzing 10,000 firms in 80 coun- ment in access to credit among households tries shows that financing constraints are at the “bottom of the pyramid,” led to a associated with slower output growth, while 7.6 percent increase in the number of informal other reported constraints are not as robustly businesses in areas with a new bank branch, 44 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 a 1.4 percent decrease in unemployment, and (Kaboski and Townsend 2011, 2012; Karlan an increase in income levels of up to 7 per- and Zinman 2010; Khandker 2005; Pitt and cent in the two years following the branch Khandker 1998). By contrast, studies that opening. The opening of Banco Azteca led to explore the impact of microfinance on en- no change among formal businesses, which trepreneurship find relatively modest effects is consistent with the new bank’s focus on (Giné, Jakiela, and others 2010; Morduch lower-income individuals and with the bank’s and Karlan 2009). In a study that experimen- low documentation requirements. In contrast, tally varies access to microcredit, Banerjee formal business owners have easier access to and others (2013) find a large effect of access commercial bank credit and likely prefer it to microfinance on investment in fixed assets, because of the higher interest rates charged but also note that this effect is concentrated by Banco Azteca. The measured impacts were among wealthier households that already larger among individuals with below median own a business, while households with a low income levels and in municipalities that were initial probability to transition into entrepre- relatively underserved by the formal banking neurship use credit to consume rather than sector before Banco Azteca opened. (For more invest. Many of the limitations of microcredit on Banco Azteca, see chapter 3, box 3.3.) as a tool to finance entrepreneurship are likely Overall, there is plenty of evidence that to be the result of the rigidity of microcredit, access to finance is important for firms, espe- including the lack of grace periods, frequent cially for the smaller and younger ones. Eco- payments, and joint liability that may prevent nomic growth would come to a halt if firms risk taking (Field and others, forthcoming; could not get credit. But there are major Giné, Jakiela, and others 2010). While joint ongoing debates on the pros and cons of liability contracts have made the extension of microcredit, which are discussed in the next credit to marginal clients possible, such con- subsection. tracts may be poorly suited for loans to busi- nesses for which the cash flows and risks are difficult to observe. (See also the discussion The effects of microcredit on product design in chapter 2.) In many parts of the world, substantial im- At the macroeconomic level, however, the provements in access to credit among house- impact of broader access to microcredit may holds and small businesses in recent decades be mostly redistributive and not without have occurred through the rapid expansion risks for financial stability. Recent research of microcredit. The original narrative empha- in this area has supplied intriguing new in- sized that microcredit could serve not only as sights using applied general equilibrium a tool to alleviate extreme poverty, but also modeling. For example, Buera, Kaboski, and as a means to unleash the entrepreneurship Shin (2012) offer a quantitative evaluation potential of the poor. Recent evidence has, of the aggregate and distributional impact of however, highlighted some of the limitations microfinance and find that, if general equi- of microcredit and suggests a more nuanced librium effects are accounted for, scaling up narrative. Although microcredit can have sig- microfinance programs has only a small im- nificantly positive welfare effects if used as a pact on per capita income because increases means for consumption smoothing and risk in total factor productivity are counterbal- management, most studies have found that anced by the lower capital accumulation re- the effects of microfinance on investment and sulting from the distribution of income from entrepreneurship are relatively small. high savers to low savers. The benefits occur Many studies have documented the ef- largely through wage increases and greater fects of microfinance on household welfare access to finance by poorer entrepreneurs. At and income. This includes positive effects on the same time, there are also costs: the re- consumption, economic self-sufficiency, and distribution of credit toward new borrower some aspects of mental health and well-being segments may lead to losses in the efficiency GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 45 of financial intermediation (for example, be- the effects on the risk profile of bank and cause of higher screening and information nonbank lending. The expansion of credit, if costs) and change the risk profile of bank not well managed and if combined with low lending as banks make loans to new borrow- capitalization, can lead to financial crises, as ers who are, on average, riskier clients. This illustrated quite powerfully by the examples means that the economic benefits of financial of Bosnia and Herzegovina, India, and the access need to be carefully weighed against United States illustrated in box 1.5. BOX 1.5 Three Tales of Overborrowing: Bosnia and Herzegovina, India, and the United States The debate over the sources of the U.S. mortgage cri- lion in 2003. In a forensic analysis of the sources sis highlights the issues and trade-offs in attempts to of increased mortgage risk during the 2000s, Rajan, increase credit. One of the key explanations of the Seru, and Vig (forthcoming) show that more than crisis (for example, Rajan 2010) is that it was precip- half of the mortgage losses that occurred in excess of itated by an overextension of credit and a relaxation the rosy forecasts of the expected loss at the time of in mortgage-underwriting standards (combined with mortgage origination reflected low-doc and no-doc aggregating the resulting junk bonds and reselling lending. them as highly rated bonds). There is evidence that The second example of the overextension of credit the losses in the two massive government-sponsored in the name of access comes from the other side of enterprises, Fannie Mae and Freddie Mac, reflected the globe. In October 2010, the microfi nance sec- these relaxed standards. The decline in underwriting tor in India’s Andhra Pradesh was in the middle of standards was at least partly a response to mandates a major crisis. An analysis shows that the roots of that required Fannie Mae and Freddie Mac steadily the crisis were in the rapid rise of loans disbursed to increase the mortgages or mortgage-backed secu- by specialized MFIs since the late 1990s (Mader rities they issued to target low-income or minority 2013). The liberalization of India’s economy and its borrowers and underserved locations. The turning fi nancial sector after 1991 changed the composition point, as documented by Calomiris (2011), was the of lending: credit from the private sector (especially year 2004. Fannie and Freddie had kept their expo- MFIs and nonbank financial institutions) rapidly sures low to loans made with little or no documenta- rose, even as the state remained a driving force in the tion (low-doc and no-doc loans), owing to their risk background. Evidence suggests that the expansion management guidelines that limited such lending. In of Indian microfinance did have some—relatively early 2004, however, senior management realized limited—impacts (Banerjee and others 2013). At the that the only way to meet the political mandates was same time, the spectacular growth and profitabil- to cut underwriting standards. Risk managers, espe- ity of Indian MFIs in many cases also led to mul- cially at Freddie Mac, complained, as documented tiple borrowing and excessive indebtedness among by publicly released e-mails to senior management. low-income clients. While India’s MFI crisis had its They refused to endorse the move to no-docs and roots in the rapid and, at times, insufficiently regu- battled unsuccessfully against reduced underwriting lated growth of MFIs, there were also other factors standards. Politics—not shortsightedness or incom- that contributed to the crisis. First, the development petent risk managers—drove Freddie Mac to elimi- of an appropriate institutional infrastructure lagged nate its previous limits on no-doc lending. After a behind the rapid growth of the MFI sector. This is decision by Fannie and Freddie to embrace no-doc particularly true for the establishment of reliable lending, the volume of new, low-quality mort- credit reporting systems for MFI borrowers that gages increased to $715 billion in 2004 and more could have limited problems of overindebtedness and than $1 trillion in 2006, compared with $395 bil- of borrowing from multiple lenders. These problems (box continued next page) 46 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 1.5 Three Tales of Overborrowing: Bosnia and Herzegovina, India, and the United States (continued) were aggravated by the absence of well-functioning the microfi nance industry contributed to the indus- personal bankruptcy laws that could have allowed try problems: (1) lending concentration and multi- for the orderly discharge of excessive debts. Second, ple borrowing, (2) overstretched MFI capacity, and the MFI sector faced competition from grossly sub- (3) a loss in MFI credit discipline (Chen, Rasmussen, sidized state government programs that extended and Reille 2010). Fierce competition among fi nan- credit to borrowers at the bottom of the pyramid cial institutions in concentrated markets enabled cli- under soft conditions, and this arguably contributed ents to borrow from multiple lenders and increase to problems of overindebtedness and moral hazard their total loan amount, without creditors knowing. in loan repayment. Finally, the sector was affected The country began credit information bureau proj- by overt political interventions in the credit market: ects in 2005, but the bureaus became operational state governments encouraged MFI clients to stop only after the repayment crises had already started. repaying their loans ahead of elections. Hence, the This resulted in repayment problems for clients, and Indian case illustrates how the rapid growth of low- around 16 percent of MFI clients were identified as documentation lending is particularly problematic in close to exceeding their repayment capacity (Maurer environments with an insufficiently developed legal and Pytkowska 2010). To survive in an extremely and institutional framework and environments in competitive environment, MFIs started using more which political interventions in the credit market are persuasive sales techniques, hired new, less experi- common. An important lesson of the crisis—one that enced staff, and disbursed loans quickly based on has begun to be translated into policy—is that appro- shallow assessments of the repayment capacity of priate consumer protection regulation, credit market borrowers. Around 60 percent of the clients going infrastructure, and legal provisions for the orderly through repayment difficulties reported that intensi- discharge of excessive debt burdens (personal bank- fied sales behavior by fi nancial institutions contrib- ruptcy) can reduce the need for political interventions uted to their problems, and 30 percent of the prob- in the credit market, which are prone to introduce lem clients stated that they had not been visited by additional distortions into the credit market. a loan officer at the time of loan appraisal (Maurer The third study on overborrowing involves the and Pytkowska 2010). To keep up with the grow- microfi nance industry in Bosnia and Herzegovina. It ing business, MFIs also increased their staff size provides a vivid view on how excessive competition at a pace that could not keep up with the training among credit providers, riskier lending, rapid insti- staff needed to be able to monitor existing loans and tutional growth, lapses in credit discipline, and lack identify new low-risk clients. The imprudent lend- of transparency in client indebtedness can result in a ing and the decline in monitoring quality led to dete- loan delinquency crisis. The country’s microfi nance riorations in MFI loan portfolios. The portfolios- industry grew quickly from 2004 to 2008. Delin- at-risk over 30 days rose from 1 percent in 2003 to quency problems started arising in late 2008, coin- 11 percent in 2009 (Lützenkirchen and Weistroffer ciding with the recession in Europe. Even though the 2012). Strained by target-driven growth in an envi- economic downturn aggravated the pace and scope ronment mired in low levels of oversight compliance of the repayment crisis, it was not the main cause. undermined internal controls and eroded the credit The recession merely brought to the surface a cri- discipline of MFIs. Staff working under incentives sis that had been looming for some time and that that emphasized growth and market share often was primarily caused by structural deficiencies and overlooked their lending discipline. This increased excessive market growth. About 58 percent of micro- credit risk contributed to the subsequent repayment credit borrowers in the country had accumulated crises. The lack of industry standards for a code of loans from several microlenders, while 28 percent conduct also added to the problem because irrespon- were seriously indebted or overindebted (Maurer sible lending could not be curbed due to the lack of and Pytkowska 2010). Three main vulnerabilities in standards of responsible fi nance. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 47 The evidence on the overall impact of in- indefinite storage. Still, 20 percent of adults in creased microcredit access on economic de- Brazil report receiving government transfers velopment and poverty alleviation is weak via a bank account, among the highest rates in at best. Kaboski and Townsend (2011) use the developing world.29 In India, the govern- data from the Townsend Thai Survey to ment recently began depositing government evaluate the Thai Million Baht Village fund pension and scholarship payments directly into program, which involved the transfer of the bank accounts of almost 250,000 people in 1.5 percent of the Thai GDP to the nearly 20 districts; officials plan to expand the pro- 80,000 villages in Thailand to start village gram with the aim of preventing corruption banks and was one of the largest government as well as increasing financial access. These microfinance initiatives of its kind.27 The types of reforms have the potential to extend evaluation finds that some households val- the reach of the formal financial sector to the ued the program at much more than the per poorest individuals. household cost, but, overall, the program cost The challenge for public policy is to en- 30 percent more than the sum of the benefits. sure that enhancements in access are achieved One conclusion is relatively clear: micro- through the removal of market and regula- credit does have a substantial redistributive tory distortions rather than through price potential. While microcredit is costly and its regulation or other anticompetitive policies overall impact on economic growth is subject that may exacerbate distortions and threaten to debate, the available studies indicate that financial stability. Policies can enhance finan- a majority of the population is positively af- cial inclusion by addressing imperfections in fected through increases in wages. the supply of financial services (for example, through modern payment and credit informa- tion systems, the use of new lending technolo- PUBLIC AND PRIVATE SECTOR gies, support for competition in the provision BREAKTHROUGHS IN FINANCIAL of financial services) and in the demand for INCLUSION financial services (for instance, through fi- nancial literacy initiatives that raise aware- There is clear potential for private sector– and ness and lead to the more responsible use of public sector–led breakthroughs in expanding finance). Policies to support financial inclu- financial inclusion. For example, the Global sion also include initiatives to remove non- Findex data show that, in Kenya, 68 percent market barriers that prevent equitable access of adults report they had used a mobile phone to financial services (for example, through in the past 12 months to pay bills or send or consumer protection and antidiscrimination receive money.28 The spread of mobile money laws). More generally, the challenge for policy products, the proliferation of bank agents, and is to guarantee that financial service provid- the growing movement toward dispensing gov- ers are delivering their services as widely and ernment payments via formal accounts all of- inclusively as possible and that the use of such fer potential to alter substantially the ways in services is not hampered by inappropriate which adults manage their day-to-day finances. regulatory policies or nonmarket barriers that The public sector can bring about change in limit the use of financial services.30 how adults around the globe interact with the There are many important links between formal financial sector. Increasingly, govern- the public sector and financial inclusion. ments are using formal accounts to disburse Weak public sector institutions are detrimen- transfer payments. In Brazil, the government tal to financial inclusion; so, improvements allows recipients of conditional cash transfers in public sector governance can have a posi- (as part of its Bolsa Família Program) to receive tive impact on the use of and the access to payments via no-frills bank accounts, though financial services in an equitable way. There many more choose to receive payments via a are also key effects in the other direction: if virtual account that does not allow deposits or electronic payments are widely available, this 48 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 can boost the efficiency of public sector pro- 8. The caveat is that the tracking of trends in grams. The remainder of this report discusses, inclusion has been possible so far based in greater depth, policies to influence financial exclusively on data from financial service providers. User-side data have only started to inclusion among individuals (chapter 2) and be collected recently, and time series have yet among firms (chapter 3). to be built up. Even in the producer-side data, consistent worldwide data sets are available NOTES only since 2004. 9. The correlations are lower for geographic 1. For further discussion of the key functions of density, reflecting mobile technologies and the financial system, see the inaugural Global other forms of remote access. Financial Development Report (World Bank 10. See Global Financial Inclusion (Global Fin- 2012a) and the related study by Čihák and dex) Database, World Bank, Washington, others (2013). DC, http://www.worldbank.org/globalfindex. 2. See also Demirgüç-Kunt and Levine (2008) 11. From the viewpoint of development, the rela- and the references therein. tionship between household inclusion and 3. Financial inclusion can be defined in different firm inclusion is far from trivial. For example, ways. The advantage of this chapter’s defini- in Latin America and Central and Eastern tion is that it can be expressed in operational Europe in the early 2000s, household inclu- terms (by specifying “use”) and measured (for sion was expanding rapidly, whereas firm example, in map 1.1, “use” is approximated finance was slow to develop. Recent studies by “having at least one account”; in map O.1, point out that there may be costs associated it is measured more narrowly as “depositing with expanding household access before firm to or withdrawing from an account at least access has been opened up and that the impact once per month”). of financial deepening on growth and income 4. In addition to use and access, the quality of inequality derives from enterprise credit, financial services is a relevant issue. Quality rather than from household credit (Beck and is more difficult to measure empirically than others 2008). The caveat is that, in many coun- usage. Nonetheless, the issue is important, for tries, the dividing line between households and example, for consumer protection (see chap- firms is blurred in the case of small firms. ter 2). 12. An area of active recent research has focused 5. Lack of financial literacy may also result in on whether and how financial services could enable microentrepreneurs to expand their excessive use of financial services. For exam- businesses and employ additional workers. ple, people may buy insurance policies or take Microcredit impact studies show that greater on credit they do not really need, put money availability of credit had no impact on the in savings accounts with high fees that are not profits of microbusinesses owned by women welfare enhancing, and so on. in India, Morocco, or the Philippines (Baner- 6. Unbanked does not necessarily mean exclu- jee and others 2013; Crépon and others 2011; sion from the use of financial services. Some Karlan and Zinman 2011). of the people who are without bank accounts 13. See Enterprise Surveys (database), Interna- have access to transactional services or tional Finance Corporation and World Bank, accounts (for example, accounts provided by Washington, DC, http://www.enterprise mobile phone operators). Even more broadly, surveys.org. financial services are also provided by infor- 14. This is based on a country-level ordinary least mal service providers, such as moneylenders. squares regression of account penetration on This fills some of the gaps in formal finan- the log of GDP per capita (Demirgüç-Kunt cial inclusion, but is also associated with and Klapper 2012). important drawbacks because the use of such 15. New technologies—inexpensive global posi- informal financial services may involve less tioning system receivers, mapping software, protection and higher costs. and widely available spatial data—have made 7. See Global Financial Inclusion (Global Fin- it feasible to map out the geographic reach dex) Database, World Bank, Washington, of financial systems in many countries. This DC, http://www.worldbank.org/globalfindex. facilitates a range of analyses that can be use- See also Demirgüç-Kunt and Klapper (2012). ful to the commercial sector, the public sector, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION: IMPORTANCE, KEY FACTS, AND DRIVERS 49 regulators, and donor agencies. For example, Levine and Zervos 1998; Rajan and Zingales the Bill and Melinda Gates Foundation has 1998), and on poverty reduction and income sponsored a project to use mapping software inequality (Beck, Demirgüç-Kunt, and Levine and spatial data to assess the geographic dis- 2007; Clarke, Xu, and Zou 2006; Li, Squire, tribution of financial access points relative to and Zou 1998; Li, Xu, and Zou 2000). But, population in Kenya, Peru, and other coun- as illustrated in figure 1.12, deep financial tries. On Thailand, data on 960 households in sectors are not necessarily inclusive ones, 64 rural Thai villages have been made avail- which is why recent and ongoing research able via the Townsend Thai Survey and are and empirical work have focused on examin- studied in Kaboski and Townsend (2011). For ing financial inclusion and the access to and the survey, see “The Townsend Thai Project: use of different types of financial services Baseline Survey (‘The Big Survey’),” National separately from financial depth. Bureau of Economic Research, Cambridge, 24. For example, Karlan and Zinman (2011), MA, http://cier.uchicago.edu/data/baseline- based on an innovative experiment in the survey.shtml. Philippines, find that access to credit led to 16. This discussion highlights the usefulness of a decline in the number of business activi- measuring financial systems using the 4x2 ties and employees in the treatment group matrix introduced in the first Global Finan- relative to controls, and subjective well-being cial Development Report, that is, measuring declined slightly. However, they did find that depth, access, stability, and efficiency in both microloans increase the ability to cope with financial institutions and financial markets risk, strengthen community ties, and boost (World Bank 2012a; Čihák and others 2013). the access to informal credit. These find- 17. See Global Financial Inclusion (Global Fin- ings suggest that microcredit has a positive dex) Database, World Bank, Washington, impact, but that the channels may not neces- DC, http://www.worldbank.org/global sarily be those hypothesized by proponents. findex. 25. Ashraf, Karlan, and Yin (2010) also find a 18. See Global Financial Inclusion (Global Fin- reduced vulnerability to illness and other dex) Database, World Bank, Washington, unexpected events. DC, http://www.worldbank.org/globalfindex. 26. For brevity, the focus in this section is on 19. The institutional barriers to financial inclu- savings and transactions related to savings sion are analyzed in Allen, Demirgüç-Kunt, accounts. Nonetheless, there is also evidence and others (2012). of the strong impact of access to payment 20. For more, see Financial Action Task Force services. One aspect of this is the benefits of (2013). international remittances on the incomes and 21. Calculations based on the Financial Access living standards of the families of migrants, (database) 2010, Consultative Group to on which there is some evidence. (See chapter Assist the Poor and World Bank, Washing- 3, box 3.1 for a discussion on remittances and ton, DC, http://www.cgap.org/data/financial financial inclusion.) -access-2010-database-cgap. 27. For the survey, see “The Townsend Thai 22. Within Western Europe, in countries such as Project: Baseline Survey (‘The Big Survey’),” Austria and Germany, the number of coop- National Bureau of Economic Research, eratives is even higher than the number of Cambridge, MA, http://cier.uchicago.edu banks. In 2010, there were 15.9 commercial /data/baseline-survey.shtml. bank branches and 17.4 cooperatives per 28. Global Financial Inclusion (Global Findex) 100,000 adults in Germany. In Austria, the Database, World Bank, Washington, DC, corresponding number of branches was 27.5 http://www.worldbank.org/globalfindex. and 30.6, respectively. 29. Global Financial Inclusion (Global Findex) 23. Earlier research on the impact of the financial Database, World Bank, Washington, DC, sector on economic development highlighted http://www.worldbank.org/globalfindex. the contributions of aggregate financial depth 30. Beck and de la Torre (2007) rely on the notion on economic growth (Demirgüç-Kunt and of the access possibilities frontier to explain Maksimovic 1998; King and Levine 1993; this. CHAPTER 2: KEY MESSAGES • Promoting the use of financial services by individuals requires that market distortions—such as information asymmetries or the abuse of market power—preventing the widespread use of financial products be addressed, ways be found to deliver services at lower costs for sup- pliers and consumers, and consumers be educated and protected so they use products that meet their needs and avoid costly mistakes. Governments can confront market failures and enhance inclusion by developing an appropriate legal and regulatory framework, supporting the information environment, promoting competition, and facilitating the adoption of busi- ness models by providers to enhance financial inclusion. • Technological advances hold promise in the expansion of financial inclusion. Transaction costs become an obstacle for fi nancial inclusion if providers cannot profitably serve low- income consumers. Innovations in technology, such as mobile banking, mobile payments, and the biometric identification of individuals, help reduce transaction costs. Which tech- nology is appropriate for financial inclusion depends on the development of the traditional banking sector; market size, structure, and density; and the level of development of support- ing infrastructure. • Product designs that deal with market failures, meet consumer needs, and overcome behav- ioral problems can foster the widespread use of financial services. Certain business models and delivery channels can also enhance inclusion by reducing the cost of using financial ser- vices. The regulatory stance of governments can influence the product designs and business models of financial institutions. Hence, governments should strike a delicate balance between financial stability concerns and supporting innovations in product design and business mod- els that allow for greater financial inclusion. • How best to strengthen financial capability, that is, fi nancial knowledge, skills, attitudes, and behaviors, remains a focus of research and discussion, but some lessons are emerging: the importance of using teachable moments to deliver financial knowledge, the value of social networks (such as between parents and children or between remittance senders and receiv- ers), the relevance of psychological traits such as impulse control, and the possible benefits of new delivery channels such as entertainment education and text messaging. • Evidence points to the role of government in setting standards for disclosure and transpar- ency, regulating aspects of business conduct, and overseeing effective recourse mechanisms to protect consumers. To avoid confl icts of interest, prudential regulation may be sepa- rated from the regulation of financial consumer protection. Competition is also a key part of consumer protection because it creates a mechanism that rewards better performers and increases the power that consumers can exert in the marketplace. • Governments can also subsidize access to fi nance and undertake other direct policies to enhance financial inclusion, but more evidence on the effectiveness of these approaches is needed. Financial exclusion is often a result of high debt levels, especially in rural economies. Debt restructuring may be preferable to unconditional debt relief to minimize the incentives for moral hazard and restore financial inclusion. FINANCIAL INCLUSION FOR INDIVIDUALS 2 Financial Inclusion for Individuals P romoting the use of financial services by individuals requires dealing with market failures, such as asymmetric information and barriers are major impediments to the provi- sion of financial services. Innovative technol- ogies—including mobile banking, electronic moral hazard, that prevent the widespread credit information systems, and biometric use of financial products. It also involves individual identification—can reduce such designing products that fit consumer needs transaction costs and can thus help overcome and delivering services at prices that indi- some of the traditional barriers to financial viduals can afford. It entails educating and access. protecting consumers so they avoid making Major innovations in retail payment sys- costly mistakes upon entering into financial tems date back to the rise of card-based pay- contracts. Both private sector and govern- ment services. Credit cards were introduced ment engagement is necessary to expand the in the 1950s, and their use grew rapidly over financial inclusion of individuals. Techno- the next three decades based on infrastruc- logical progress, likely driven by the private ture developed and managed mainly by the sector and facilitated by the public sector, is card associations Visa and MasterCard. Dur- expected to help increase the financial inclu- ing the late 1980s and the 1990s, because of sion of individuals. growing sophistication in information pro- This chapter reviews the roles of tech- cessing and telecommunications technologies, nology, product design, financial capability, which, among other features, allowed online financial education programs, consumer pro- transaction authorization by issuers, credit tection and market conduct, and government cards became a widely accepted form of pay- policies in fostering financial inclusion. ment in many countries. In the 1980s, debit cards started to evolve as a key electronic payment instrument. Today, in some coun- THE ROLE OF TECHNOLOGY tries where credit card adoption has been The last two decades have seen a proliferation slow because of the limited infrastructure for of new technologies with significant potential credit information and other issues, such as to improve financial access. As illustrated in cultural preferences, debit cards have become chapter 1, transaction costs and geographical the most popular electronic instrument for GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 51 52 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 retail payments.1 The growth in debit cards negative implications for competition and has been dramatic over the last 25–30 years. product innovation. At first, debit cards were enablers for mov- More recently, much attention has been ing customers from bank teller counters to focused on the role of mobile banking applica- the newly deployed automatic teller machine tions in financial inclusion. The reason is that (ATM) systems. Over time, instead of using mobile phones have been adopted by con- the card to withdraw cash from an ATM to sumers at a rapid rate, becoming almost ubiq- pay merchants, bank customers could simply uitous. They are now well within the reach of present the card to the merchants and have many poor individuals around the world. In their bank account debited directly. Given many low- and middle-income countries, the this tremendous potential, debit card prod- share of the population that has access to a ucts evolved globally and became based on mobile phone is considerably larger than the the infrastructure that was already in place share of the population that has a formal for processing credit card transactions at the bank account. For example, in 2011, there point of sale. were 127 mobile phone subscriptions for The market for prepaid cards, or stored- every 100 inhabitants in South Africa, while value cards, has also become one of the most only 54 percent of the population had a bank rapidly growing segments in the retail pay- account. There were 123 mobile phone sub- ment industry.2 In the 1990s, when prepaid scriptions per 100 inhabitants in Brazil, while cards first became available, they were mostly only 56 percent had a bank account. And, in issued by nonfinancial businesses and used India, 72 of every 100 inhabitants had a in limited deployment environments, such as mobile phone, while only 35 percent had a mass transportation systems. In recent years, bank account.3 These numbers illustrate the prepaid cards have grown substantially as vast potential of mobile telephony to enhance financial institutions and nonbank organi- financial inclusion. zations target the unbanked and migrant Mobile banking has been perhaps the most remittance segments of populations. Some visible example of the use of new technolo- prepaid cards already rely on the existing gies to advance financial inclusion, but new infrastructure for traditional credit and debit technologies have also had an impact in other cards. Technological innovations in the way areas. For instance, modern information tech- information is stored (such as magnetic stripe nologies have allowed banks to serve previ- or computer chip), the physical form of the ously unbanked locations through banking payment mechanism, and biometric account correspondents. Improved technologies for access and authentication are converging to credit reporting and borrower identification create efficiencies, reduce transaction times at have dramatically reduced the cost of finan- the point of sale, and lower transaction costs. cial intermediation and allowed banks to pro- Although innovations in card-based retail vide financial services to clients who would payments have expanded access to finance have been excluded from the use of formal and the method remains a dominant mode financial services in the absence of these tech- of transactions in many countries, there are nologies. As computing power has grown, limitations. From a consumer protection per- banks are also able to leverage data on histor- spective, card-based payment systems can be ical client behavior to better assess credit risk problematic because of hidden, nontranspar- and deliver credit to previously underserved ent fees. From a technology perspective, a individuals. potential concern is that the huge investments While new technologies are, in principle, that banks and payment system providers available globally, their adoption and use have made into card-based payment infra- for financial inclusion have been uneven structure may inhibit interoperability and across countries, and there has also been create market segmentation, which may have great variation in whether such technologies GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 53 have been pioneered by the traditional finan- make the success of some technology-based cial sector or other players. India’s financial solutions difficult to replicate elsewhere. inclusion strategy, for example, relies on providing basic financial services through Mobile banking and payments traditional bank branches and technologi- cally based correspondent banking, both Mobile banking and payment technologies led by the country’s large public sector are among the most significant financial sec- banks. At the other end of the spectrum, tor innovations of the last decades. The wide Kenya’s popular mobile payment service, geographical reach and the rapid growth of M-PESA, is operated by a private telecommu- mobile phone technology has dramatically nications provider and has reached nation- reduced communication costs from prohibitive wide appeal independently of the traditional levels to prices that are well within the reach banking sector. of many low- and middle-income individuals This section discusses the role of tech- across the developing world (map 2.1; figure nology in financial inclusion. It reviews the 2.1). In the early stages of this technological growth of mobile banking and payment sys- revolution, users started transferring air time tems and discusses technology-based business credits as a mode of payment within the net- models and the role of improved borrower work. This soon gave rise to the first mobile identification and credit reporting technolo- payment systems that were formalized by gies in financial inclusion. The section high- telecommunications providers (such as Safari- lights that technology-based strategies for com in Kenya) or by banks that began allow- financial inclusion have varied substantially ing customers to receive, transfer, and deposit across countries and examines the features of money over the mobile phone network. national market environments that determine There is still considerable scope for expan- which technologies are best suited to enhance sion in mobile technology to translate into financial inclusion, as well as issues related to greater access to financial services. For exam- market structure and regulation that might ple, Alonso and others (2013), using data on MAP 2.1 Mobile Phones per 100 People, 2011 Source: Calculations based on World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators. 54 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 2.1 Mobile Phones per 100 People, by Country Income Group, 1990–2011 125 100 Mobile phones per 100 people 75 50 25 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 World High income countries Middle income countries Low income countries Source: World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators. Mexico, estimate the potential demand gap Mobile banking and payment technologies for mobile banking—the difference between could serve as a key stepping-stone into the the possession of mobile phones and the use of formal financial services, but a major access to current bank accounts—at about challenge is to design secure mobile applica- 40 percent. tions in a way that makes them easy to access To put matters in perspective, although and use in everyday transactions. Because mobile money has changed the economics mobile banking and payment applications of banking across the globe, the aggregate have network externalities, they become volumes of mobile banking transactions are more cost effective to operate and use as still small compared with the value transacted more participants join the system. In practice, through traditional payment instruments. this means that the success of mobile banking Returning to the example of Kenya, one of and payment systems depends crucially on the countries where the adoption of mobile the number of potential users a participant payments has been most successful, the value can transact with. Because mobile money of transactions among banks is nearly 700 applications in developing countries are set times larger than the value of all transactions up in the context of what are still largely cash among M-PESA mobile accounts (Jack and economies, other crucial factors determin- Suri 2011). ing their adoption are the number of cash-in One area where mobile technologies can be points and the ease with which cash can be an important driver of change is remittance transferred into and out of the mobile system. flows. While mobile payments are becoming What are some of the requirements that a popular channel for sending domestic remit- would need to be in place for a country to tances, technological and regulatory barriers replicate Kenya’s extraordinarily successful still constrain their widespread use for inter- experience in the adoption of mobile pay- national remittances. Box 2.1 provides an ments? The economics of mobile banking in-depth discussion of the links among remit- rely on network externalities and economies tances, technology, and financial inclusion. of scale. In the case of Kenya, the rapid adop- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 55 BOX 2.1 Remittances, Technology, and Financial Inclusion Remittances are among the most important fi nancial tances as a share of gross domestic product (GDP) transactions for populations with limited access to and less account penetration (figure B2.1.1, left formal banking services. The total value of remit- panel). Drawing conclusions from this observation tances has been rising steadily over the past decade. is difficult because, within countries, lower-income The World Bank (2013b) estimates that officially population segments are more likely to have a family recorded international migrant remittances (defi ned member sending remittances and not have a bank as the sum of worker remittances, the compensation account. While data on fi nancial inclusion are now of employees, and transfers initiated by migrants) to available through the Global Findex database, data developing countries totaled $401 billion in 2012. on remittances by income decile still need to be col- Remittances within countries—from one province lected more systematically through surveys. a (Sub- or urban area to another—are several times this stantial progress in this area is expected through the amount. Many countries receive remittances to the planned addition of a module in the Gallup World extent that the remittances have significant mac- Poll in 2014.) roeconomic effects, including real exchange rate Indeed, to send and receive remittances, house- appreciation. holds increasingly rely on mobile banking and There should be a strong relationship between other modern retail payment applications. For remittance flows and financial inclusion. A key many households, this can serve as a primary point reason is that remittances are usually regular and of entry to the financial system and to the use of predictable flows, which should, in principle, make fi nancial services that extend beyond payment sys- remittance recipients relatively more inclined to tems. The potential of mobile banking is evident join the formal fi nancial sector. This is not obvious, from data showing that poorer countries lag much however, from cross-country data. Lower-income more in terms of account penetration than in mobile countries tend to have both higher levels of remit- telephony (figure B2.1.1, right panel). When people FIGURE B2.1.1 Remittances and Financial Inclusion a. Remittances to GDP b. Remittances per capita 100 100 90 90 Account penetration among adults, % 80 80 70 70 60 60 50 50 40 40 30 30 20 20 R 2 = 0.253 10 10 0 0 0 10 20 30 40 50 0.0 1.0 2.0 3.0 4.0 Gross remittance flows, % of GDP Log, gross remittance flows per capita > 1 mobile per person < 1 mobile per person Source: Calculations based on Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. Note: Gross remittance flows is the sum of remittance payments and receipts. (box continued next page) 56 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.1 Remittances, Technology, and Financial Inclusion (continued) come to a bank or credit union to engage in remit- 7.6 percent. However, mobile services for cross-bor- tance transactions, they often end up opening a der remittances are more difficult to access in most bank account after several visits. Many providers countries, and there is substantial scope to make the of fi nancial services have recognized this enormous costs of mobile money transfers more transparent. potential of introducing new client groups to fi nan- This will be helped by regulatory initiatives such cial services through remittances and are actively as the U.S. remittance transfer rule and efforts to offering additional services along with remittance improve consumer awareness.b accounts. Several countries have integrated remittance The positive impact on fi nancial inclusion of bun- products into national policies on fi nancial inclu- dling remittance accounts with other fi nancial prod- sion. Under India’s National Financial Inclusion ucts is also apparent in other ways. The fi xed costs Strategy, for instance, many public sector banks of sending remittances tend to make remittance offer accounts that charge no fees for remittances. flows lumpy and seasonal. This boosts the demand The Philippine Development Plan (2011–16) explic- for savings instruments that offer households a safe itly notes the need to promote financial inclusion place to store temporary savings and to use income and facilitate remittances both internally and from for consumption smoothing (Anzoategui, Demirgüç- abroad (NEDA 2011). To these ends, the central Kunt, and Martínez Pería 2011). Remittances often bank has approved alternative ways of making also improve access to credit because the processing remittances, such as the Smart Padala, G-Cash, and of remittance fl ows provides fi nancial institutions stored-value cards, and competition is helping both with information about the creditworthiness of poor to lower transaction costs and to reduce the time recipient households, which helps make financial needed for delivery. institutions more willing to supply credit and micro- At the global level, the value of facilitating remit- loans. Finally, remittances can facilitate microinsur- tance flows and reducing costs has been repeatedly ance, especially the purchase by migrant relatives, emphasized by G-8 and G-20 leaders (see, among for example, of health insurance for their families in others, the G-8 L’Aquila Declaration and the G-20 their countries of origin. Cannes declarations). The efforts of the Global Given the potential role of remittances in raising Remittances Working Group (chaired by the World fi nancial inclusion, it is important to make transfer Bank’s Financial and Private Sector Development systems less costly, more efficient, and more trans- Vice President) were successful in securing G-8 and parent. According to recent survey data (World G-20 commitments to reducing the cost of remit- Bank 2013c), account-to-account products between tances by 5 percentage points in five years (the 5x5 nonpartner banks are the most expensive, with an objective). The implementation of remittance price average cost of about 13.6 percent, while transfers comparison databases is proving effective in reach- within the same bank or to a partner bank cost ing or monitoring the progress toward achieving this about 8.4 percent, on average. Mobile services are objective.c among the cheapest product types, at a cost of about a. See Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org /globalfi ndex. b. See “Remittance Transfer Rule (Amendment to Regulation E)” Consumer Financial Protection Bureau, Iowa City, IA, http://www.consumerfi nance.gov/remittances-transfer-rule-amendment-to-regulation-e/. c. See the Remittance Prices Worldwide Database, at http://remittanceprices.worldbank.org. tion of mobile banking was driven by Safari- settings where the market for mobile bank- com, the largest mobile network operator in ing is more fragmented, the adoption of these the country, which provided an extensive net- technologies has often been much slower work that ensured banking was viable for the because providers have insufficient incen- provider as well as for potential clients. In tives to invest in the extensive infrastructure GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 57 required to make mobile banking economi- 31,000 agents (covering 90 percent of Paki- cally viable. stan’s districts) who process more than 10 This illustrates that regulators walk a fine million payments per month. Several mobile line between providing incentives for the banking providers currently operate in Paki- development of mobile payment platforms stan, and nearly all of them are involved in the and requiring these platforms to be open. digitized disbursement of G2P payments (Rot- From a consumer perspective, it is clearly man, Kumar, and Parada 2013). The case of desirable to require different electronic bank- Pakistan is also a good example of how gov- ing and payment platforms to be interoper- ernments can work with private sector part- able so that users can interact across mobile ners to leverage the benefits of mobile banking banking applications without technological technologies: Pakistan’s government agencies barriers or extra charges. However, if opera- are keen to channel G2P payments through tors are required to interconnect at an early mobile banking to reduce administrative costs stage of development, this may weaken the and increase transparency, while mobile bank- incentives to invest in infrastructure and ser- ing providers view G2P payments as a useful vices that can be leveraged throughout the way to grow their client base among low- and system. Striking the right balance between middle-income individuals. As a result, the these competing policy goals remains a key share of G2P payments made through Paki- challenge in the regulation of new payment stan’s mobile network has steadily risen, and technologies. it is expected that, within five years, up to 75 Aside from issues of market structure and percent of all G2P payments in Pakistan could regulation, there are a variety of govern- be digitized. ment policies that can support the adoption The effect of mobile banking on savings of mobile banking. One policy that has been has been a matter of some dispute. Because showing some promise in encouraging the many mobile banking platforms were origi- adoption of mobile banking technologies is nally designed as pure payment systems, one the use of government-to-person (G2P) pay- concern is that mobile banking facilitates pay- ments. G2P payments provide an attractive ments and consumption but does not gener- opportunity to draw previously unbanked ate incentives to save and to engage in other beneficiaries into formal financial services by welfare-enhancing financial behaviors. There channeling a consistent flow of money into is some evidence that mobile banking services financial accounts. Several governments have are used much more as a mode of transaction begun to experiment with the use of mobile than as a store of value. For example, in a payment technologies on a small scale, for study of mobile banking in Kenya, Mbiti and example, as a way to channel welfare pay- Weil (2011) suggest that access to M-PESA ments to low-income individuals. These are has only a limited effect on the ability of indi- only relatively limited programs in terms of viduals to save. In a related study, Demom- size and aggregate effect so far; most govern- bynes and Thegeya (2012) show that enroll- ments use card-based or direct deposit tech- ment in Kenya’s M-PESA system increases the nologies for G2P payments. However, two likelihood of saving by up to 20 percentage instances of large government transfer pro- points. People who have only M-PESA save, grams that have gone fully mobile are Colom- on average, K Sh 1,305 per month (about bia’s Familias en Acción Program and Paki- $13), compared with K Sh 2,282 per month stan’s Benazir Income Support Program. among people who save only with other The digitalization of government payments accounts and K Sh 2,959 among people who in Pakistan is an example of how G2P pay- save with M-PESA and other accounts. ments over mobile networks can be used as Taking into consideration the fact that an effective tool for supporting financial inclu- people who save using non–M-PESA sion. There are currently 1.8 million mobile accounts tend to be wealthier individuals who banking accounts in Pakistan and more than save more, one may conclude that mobile 58 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 banking—in addition to reaching unprec- and point-of-sale technologies to reduce edented numbers of low-income individuals transaction costs and expand access to finan- with basic payment services—may encour- cial services beyond locations covered by a age savings. Reliable evidence on this point bank’s existing network of branches. is, however, scarce. One reason for this is The innovative aspect of mobile technology– that many electronic retail payment systems based correspondent banking is the combina- still focus heavily on facilitating transactions, tion of mobile technologies and new delivery but do not offer sufficiently attractive sav- channels, for example, by using retail outlets ings products. In some cases, this is because as banking correspondents. This enables regulators do not permit mobile banking banks to offer more convenient points of providers to act as deposit-taking institu- access to existing customers, decongest their tions. There is, nonetheless, wide variation in branches, collect payments, and gain a larger regulatory regimes, and some providers, such geographical presence without having to as M-Shwari in Kenya, have partnered with invest in brick and mortar branches. Most banks and started offering credit products. of the underlying technologies are similar to those that enable basic mobile banking. However, correspondent banking can also Innovative delivery channels reach people without mobile phones. More- In addition to enhancing financial inclusion over, it typically provides a broader suite of directly, new mobile banking and payment financial services—including insurance and technologies have also given rise to technol- savings products—than mobile banking. ogy-based business models that can broaden More recently, some of the more mature access to basic financial services. Banking mobile payment platforms have linked up correspondents that use a combination of with banking partners to offer a broader card- and mobile-based technologies are an range of products to customers, often example of how banks use new technologies through the retail channel of an established to provide financial services to previously banking correspondent network.4 unbanked customers and locations. Evidence shows that correspondent bank- A banking correspondent is a represen- ing has had a substantial impact on financial tative of a bank who operates transactions inclusion. For example, Allen, Demirgüç- on behalf of one or more banks outside the Kunt, and others (2012) find that, among bank’s branch network. The term “banking adults in the bottom income quintile, the correspondent” is often used broadly and likelihood of using a formal financial account may include post offices, supermarkets, gro- increases by up to 5 percentage points through cery stores, gasoline stations, and lottery out- the introduction of correspondent banking. lets that offer basic financial services. Banking In many countries, including Brazil, India, correspondents can be fixed retail locations Kenya, and Mexico, correspondent bank- that offer banking services to clients on a ing has been instrumental in enhancing the commission basis in previously unbanked access to basic financial services. For instance, locations or mobile banking agents that visit technology-based business models played a remote locations regularly to offer basic key part in India’s policies to enhance finan- financial services. It is worth distinguishing cial inclusion. In 2006, India adopted a bank- between banking correspondents, who offer a led, technology-driven banking correspondent broader range of financial services on behalf model. The Reserve Bank of India called on of a bank, and mobile agents, who act on banks to provide basic financial services in all behalf of a mobile phone or payment opera- unbanked villages in two phases: first, to all tor and typically provide only elementary villages with a population of at least 2,000 transaction services. Both types of correspon- and, second, to all villages with a popula- dent models have grown rapidly through the tion of less than 2,000. Banks have used a advent of new mobile banking and payment combination of new branches, fixed location technologies. They rely on mobile banking business correspondent outlets, and mobile GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 59 technology–based banking correspondents to also created an immediate use for the corre- meet this target. As of March 2012, 96,828 spondent networks. Today, these are largely new customer service points had been set up grouped under the Bolsa Família conditional under the program.5 According to Findex cash transfer program, which serves nearly data, only about 35 percent of India’s adult 14 million families (the largest program of its population had a bank account in 2011.6 kind in the developing world). Central bank Brazil leads the way globally in terms of regulations likewise encourage financial insti- the coverage and number of correspondents. tutions to reach out to more distant consum- Brazil’s strong payment system infrastruc- ers and to communities where they had not ture has set the technological foundation for previously been active. The Brazilian example the successful and rapid deployment of cor- illustrates not only the enormous successes, respondent banking. Brazil has 1,471 point- but also some of the challenges that this inno- of-sale terminals per 100,000 adults, more vative channel faces. For example, despite than three times the number in Chile (450 per progress, correspondents in poorer and more 100,000), and also leads the region in ATMs, remote areas tend to be limited to providing at 121 per 100,000, a coverage rate similar basic access to payments, and services such to the rates in Germany and other European as savings, credit, and insurance are not read- countries. The large government transfer pro- ily available for many low-income consumers grams handled through public banks have (box 2.2). BOX 2.2 Correspondent Banking and Financial Inclusion in Brazil The development of a widespread correspondent a private bank. This has raised the incentive for other banking network is one of the main factors behind banks to look for alternative correspondent networks. improvements in financial inclusion in Brazil. The Between 2005 and 2010, the number of cor- central bank encouraged financial institutions to respondents approximately doubled, exceeding reach out to more distant consumers and to com- 150,000 in December 2010 (table B2.2.1). The num- munities where they had not previously been active, ber of bank branches grew by about 12 percent in including lower-income areas, through partnerships the same period, from 17,627 to 19,813. Banks and with a variety of retail establishments. In response fi nancial institutions have partnered with a variety to early successes with the program, regulators have of retail establishments, including some with public gradually reduced the restrictions on correspondent ties such as the post office network and lottery agen- banking, for instance, on individual approval pro- cies. Several fi nancial institutions, Banco Bradesco cesses. The legal framework has facilitated healthy and Caixa, for instance, have also developed river- expansion by putting the onus on regulated institu- boat banks to reach distant communities along the tions to train and monitor their correspondents. An Amazon (figure B2.2.1). important step was the auctioning off of the right to The example of Brazil highlights how innovative use post offices as correspondents, which was won by business models can harness the possibilities of new TABLE B2.2.1 Correspondent Banking, Brazil, December 2010 Links, % Activities authorized, operational correspondents, % Financial Open Funds Payments Loans and Credit Region Number Bank company account transfers (send and receive) fi nance cards North 6,850 91 14 34 48 79 53 42 Northeast 31,752 93 13 29 47 81 51 41 Central-West 11,948 86 19 25 40 72 58 40 Southeast 67,878 82 31 23 31 61 66 38 South 31,195 80 27 28 37 68 66 38 All 151,623 84 24 26 37 68 62 39 Source: Calculations based on data of the Central Bank of Brazil. (box continued next page) 60 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.2 Correspondent Banking and Financial Inclusion in Brazil (continued) FIGURE B2.2.1 Voyager III: Bradesco’s Correspondent tional bank branches. (The total number of bank Bank in the Amazon branches was 14,631 in 2011.) Nearly half of all banking correspondents in Mexico (9,964 as of 2011) are located in convenience stores, which are part of the OXXO retail chain. Walmart, 7-Eleven, and several pharmacies also serve as banking cor- respondents. The main publicly owned banking correspondent partner is Telecomunicaciones de México (with 1,597 banking locations in 2011), the third largest of the correspondent network partners in Mexico, but significantly smaller than the public sector network partners in Brazil, such as the lottery or the post offices. Mexican regulations are flexible and permit even individuals with a laptop to act as agents, as well as large retailers that offer complete banking services. Banco Azteca, which operates out Source: Banco Bradesco. Used with permission; further permission required of the Elektra retail chain, offers an example of the for reuse. high end of this market; it provides comprehensive banking services through Elektra stores to a largely technologies to extend financial services to previously low-income clientele. unbanked households and locations. The data in table In Colombia in 2006, the government created an B2.2.1 illustrate some of the ongoing challenges this innovative program, Banca de las Oportunidades, channel faces. For example, wealthier regions (such as to support fi nancial inclusion through a combina- the South and Southeast) have a higher share of corre- tion of policy actions, including regulatory reforms, spondents, and the correspondents in these areas are financial capability initiatives, and incentives for more likely to provide a full range of services (such as providers to meet the demand of low-income con- account opening and access to loans), whereas many sumers for banking services. Correspondent bank- correspondents in the North and Northeast only ing is one of the approaches to fi nancial inclusion handle payments (such as those related to utility bills that Banca de las Oportunidades has supported and the provision of monthly government benefits). both through regulatory reforms such as the cre- Basic fi nancial services beyond payments (including ation of nonbank correspondent agents and simple savings, credit, and insurance) are still not readily know-your-customer rules. As of the fi rst quarter of available for many low-income consumers. Govern- 2013, there were 35,765 correspondents in Colom- ment authorities are therefore now aiming to extend bia, roughly equivalent to slightly more than 10 per the use of banking correspondents to government 10,000 adults, compared with only 7,183 traditional payments and to provide greater incentives to use cor- branches of fi nancial institutions.a respondent banking to enable savings. These three examples from Latin America all Other countries in Latin America have likewise demonstrate the importance of the regulatory envi- expanded access through the use of correspondent ronment for the expansion of correspondent bank- banking networks. In Mexico, changes in the regu- ing. In each case, Brazil, Colombia, and Mexico, latory framework in 2009 and 2010 resulted in the the steps taken by the regulator to facilitate agent number of correspondent banks more than dou- or correspondent banking, such as greater flexibil- bling, from 9,303 in the fourth quarter of 2010 to ity in documentation and the know your customer 21,071 one year later. This represents an estimated requirements for small balance accounts, were criti- 2.64 correspondent banking sites per 10,000 adults, cal in enabling banks and other formal financial compared with only 1.83 per 10,000 among tradi- institutions to expand the networks. a. See http://www.bancadelasoportunidades.gov.co/portal/default.aspx, the Banca de las Oportunidades website. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 61 Technologies for improved borrower largest effort at biometric borrower identifi- identification and credit reporting cation is currently under way in India, where the Unique Identification Authority of India’s Technologies that reduce asymmetric infor- Aadhaar Program is assigning identification mation in credit markets are another type of numbers to all of the country’s citizens. When technological innovation that can enhance the process is finished in 2014, each person financial inclusion. Many financial markets will have an identification number linked to are plagued by severe information problems, biometric data, including a photograph, iris which are a major cause of financial exclu- scans, and fingerprints. It is envisioned that sion. Lenders will try to compensate for the the identification numbers could be linked to lack of reliable information on the identities Aadhaar Enabled Payment Systems, as well as and credit histories of borrowers by raising a borrower’s credit history (such as bank and collateral requirements, engaging in the costly microfinance loans) to improve transparency screening of borrowers prior to approval, or and reduce information problems in the credit refusing to lend to certain segments of the market. borrower population altogether. This leads to Recent research indicates that technolo- credit rationing and financial exclusion even gies for improved borrower identification can among otherwise creditworthy borrowers. substantially reduce information problems Most developed economies have national and moral hazard in credit markets. Giné, identification systems that make it easy to Goldberg, and Yang (2012) report on a field identify borrowers uniquely and track indi- experiment in Malawi that introduced fin- vidual credit histories. For a credit reporting gerprinting as a form of biometric borrower system to function effectively, it must be pos- identification. In line with theoretical predic- sible to identify individuals uniquely. This is tions, the intervention improves the lender’s a challenge in many low- and middle-income ability to implement dynamic incentives (that countries where no universal identification is, the ability to deny credit in a later period system exists. Even where there is some form based on earlier repayment performance) and of formal identification, it is often hard to reduces adverse selection and moral hazard verify the authenticity of the documentation. (box 2.3). The paper is part of the broader This means that borrowers can easily renege literature on the topic. In particular, a related on financial commitments and makes lend- paper by Karlan and Zinman (2009) finds ers reluctant to provide financial services and experimental evidence of moral hazard and credit to new clients. Microfinance institutions weaker evidence of adverse selection in urban (MFIs) have traditionally circumvented this South Africa. A number of recent papers sup- problem by relying on group lending and fre- ply empirical evidence on the existence and quent personal interaction between borrow- impacts of asymmetric information in credit ers and loan officers. However, in the wake of markets in both developed and developing recent default crises in microfinance (see chap- economies (for instance, Edelberg 2004; Giné ter 1, box 1.5), even MFIs are increasingly and Klonner 2005; Visaria 2009). relying on traditional credit reporting, which Although efforts to create better borrower underscores the need for reliable identification identification schemes are currently under at the bottom of the financial pyramid. way in many countries, governments often To address this challenge, many countries use multiple purpose-specific identification have resorted to innovative technological schemes. This can run counter to the objective solutions for improved borrower identifica- of creating a base identification infrastructure tion. Local and national governments have, on which the private sector can build financial for example, introduced biometric forms of inclusion products. Hence, one policy impli- identification, which utilize biometric data cation for the creation of an effective unique on individuals, such as fingerprints. These identification system with wide applicability can be linked to credit histories. The world’s is the need to separate the provision of unique 62 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.3 The Credit Market Consequences of Improved Personal Identification Until recently, there has been virtually no empirical experiment was carried out in rural Malawi, which evidence on the impact of personal identification in is characterized by an imperfect identification system credit markets. Three questions are of broad interest. and limited access to credit (Figure B2.3.1). Accord- First, how do improvements in personal identification ing to the Global Financial Development Report affect borrower and lender behavior and, ultimately, database, Malawi ranked 124 out of 153 jurisdic- loan repayment rates? Second, how prevalent are tions for which 2011 data were available on private adverse selection and moral hazard in credit markets? credit to GDP, a frequently used measure of fi nancial And, fi nally, how does improved personal identifica- sector depth.a Malawi also received low marks in the tion affect the operation of credit reporting systems? depth of credit information index, which proxies for Giné, Goldberg, and Yang (2012) present results the amount and quality of information about bor- from a randomized field experiment that sheds light rowers available to lenders.b Few rural Malawian on the above questions. The experiment random- households have access to loans for production activ- izes the fi ngerprinting of loan applicants to test the ities: only 12 percent report any production loans in impact of improved personal identification. The the past 12 months, and, among these loans, only FIGURE B2.3.1 Fingerprinting in Malawi a. Repayment: balance paid on time b. Repayment: balance, eventual 100 98 8,000 91 92 93 96 7,609 88 89 90 7,000 80 79 74 6,000 70 Paid on time, % 60 5,000 50 4,000 3,888 MK 40 2,975 3,000 30 26 2,000 1,506 1,486 1,737 20 1,133 1,024 10 1,000 572 197 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Quintiles Quintiles c. Land allocated to paprika d. Market inputs used on paprika 25 14,000 23 23 12,378 21 22 21 11,803 12,000 11,262 20 19 19 9,600 9,858 10,000 Allocated land, % 15 16 8,874 8,381 8,088 15 8,000 MK 11 10 6,000 4,911 4,000 5 2,503 2,000 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Quintiles Quintiles Fingerprinted Control Source: Calculations based on Giné, Goldberg, and Yang 2012. Note: The graphs show the repayment rates among fingerprinted (gold) and control (blue) groups by quintiles of the ex ante probability of default. Individuals in the worst quintile (Q1) are those with the highest probability of default and those on whom fingerprinting had the largest effect. MK = Malawi kwacha. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 63 BOX 2.3 The Credit Market Consequences of Improved Personal Identification (continued) 40 percent are from formal lenders. In the experi- ers to choose smaller loan sizes, which is consistent ment, farmers who applied for agricultural input with a reduction in adverse selection. In addition, loans to grow paprika were randomly assigned to farmers with a high risk of default who are fi nger- either a control group or a treatment group where printed also divert fewer inputs away from the con- each member had a fi ngerprint collected as part of tracted crop (paprika), which represents a reduction the loan application. in moral hazard. If these benefits are compared to The authors fi nd that fi ngerprinting led to sub- the estimated costs of implementation, the adoption stantially higher repayment rates for the subgroup of of fingerprinting is revealed to be cost-effective, with borrowers with the highest ex ante default risk. This a benefit-to-cost ratio of about 2.3:1. The paper also suggests that fi ngerprinting, by improving personal has implications for the perceived benefits of a credit identification, enhanced the credibility of the lend- reporting system. Despite the absence of a credit er’s dynamic incentive, that is, the threat of denying bureau in Malawi, study participants were told that credit based on earlier repayment performance. The their fingerprints and associated credit histories could impact of fi ngerprinting on repayment in the high- be shared with other lenders. Since fi ngerprinting led est default risk subgroup (20 percent of borrowers) to positive changes in borrower behavior, the paper is large: the average share of the loan repaid (two underscores the belief of borrowers that improved months after the due date) was 67 percent in the identification would allow the lender to condition control group, compared with 92 percent among credit decisions on past credit performance. This is fingerprinted borrowers. In other words, among important because it suggests how borrowers may these farmers, fi ngerprinting accounts for roughly respond to the introduction of a credit bureau. These three-quarters of the gap between repayment in fi ndings also complement recent work by de Janvry, the control group and full repayment. By contrast, McIntosh, and Sadoulet (2010), who study the intro- fi ngerprinting had no impact on repayment among duction of a credit bureau for microfi nance borrow- farmers with low ex ante default risk. ers in Guatemala and find that the resulting improve- The authors also collected unique additional evi- ment in borrower information leads to large efficiency dence that points to the presence of both moral haz- gains through the reduction of moral hazard and ard and adverse selection. Fingerprinting leads farm- adverse selection. a. Global Financial Development Report (database), World Bank, Washington, DC, http://www.worldbank.org /fi nancialdevelopment. b. See Doing Business (database), International Finance Corporation and World Bank, Washington, DC, http://www.doingbusiness.org/data. (biometric) identification from customization generated by new borrower identification and to a specific purpose that could narrow the credit reporting technologies (see also IFC scope of the applications. 2012a; World Bank 2011a). While many developing economies are First, where credit reporting institutions still facing the basic technological challenge exist, information sharing between differ- of establishing an infrastructure for unique ent financial institutions is often difficult to borrower identification, there is a range of achieve in practice. Because lenders benefit policy interventions available to countries from exclusive access to information about with a more advanced credit reporting infra- the creditworthiness of prospective borrow- structure. For the most part, these policies are ers, established lenders are frequently reluc- aimed at creating the appropriate legal and tant to share credit information with com- competitive framework for the possibilities petitors, especially with market participants 64 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 who challenge traditional lending models, Indeed, neither ubiquity, nor a high penetra- such as the providers of consumer credit or tion of mobile phones is a necessary condi- MFIs. Ensuring equitable and transparent tion for the development of mobile banking. access to credit information allows custom- For example, M-PESA began offering mobile ers to use their credit histories as reputa- payment services when Kenya’s mobile phone tional collateral, strengthens credit market penetration rate was only 20 percent, not competition, and enhances access to finance. so different from the levels in Afghanistan, In recent years, these basic functions of credit Rwanda, and Tanzania when they introduced reporting institutions have been supported similar services. In cross-country compari- by new technologies, including the biomet- sons, the correlation between mobile phone ric borrower identification tools described subscriptions and the use of mobile phones above, and improved banking technologies for payments and sending money is insignifi- that make it easy to share and collect bor- cant (figure 2.2). rower information. To understand more clearly the reasons Second, existing credit reporting institu- for this surprisingly low correlation, one may tions often cover only borrowing and transac- usefully focus on two country cases from tions within the traditional banking sector. To opposite ends of the spectrum: the Russian strengthen the role of credit reporting insti- Federation and Somalia. Russia has one of the tutions as a tool for financial inclusion, one highest rates of mobile phone subscriptions in should expand the coverage of credit report- the world (179 per 100 people), while ranking ing systems to nontraditional lenders, such as among the lowest in terms of mobile phone nonbank financial institutions and microfi- use for financial transactions (less than 2 per nance borrowers. This would not only help 100 adults). Part of the explanation for this graduate microfinance clients into formal phenomenon is a long-standing preference for banking. It could also serve to prevent over- using cash in transactions rather than other indebtedness among low-income borrowers, methods of payment; many private sector which has become a cause for concern in the employers pay employee wages in cash, and— wake of recurring default crises in microfi- despite growth in e-commerce—the preferred nance and the rapid expansion of consumer method of payment for online orders is cash finance in emerging markets. on delivery. About 50 percent of Russians Finally, where comprehensive credit report- are skeptical of debit and credit cards, even if ing institutions are in place, they have more they own one. Thus, 83 percent of the more effect on financial inclusion if they are gov- than 140 million active bank cards are payroll erned by an adequate legal framework that cards, and 90 percent of the activities regis- safeguards the rights of consumers, mitigates tered on these cards are ATM withdrawals on some of the risks associated with the availabil- paydays (Adelaja 2012; Evdokimov 2013). A ity of large amounts of personal credit infor- contributing factor is that Russian banks have mation on individuals, and ensures equity and been slow to develop electronic payment and transparency in information sharing among mobile banking services. In mid-2012, only various market participants. 4 of the 30 largest Russian banks provided a complete set of mobile banking services, and only 7 enabled money transfers between Adopting new technologies: The role of individuals.7 Only in 2011 was legislation on the market environment and competition electronic payments adopted that legalized New technologies, particularly mobile bank- electronic payments and set up the legal and ing and retail payment systems, have become security framework to make such payments an integral part of financial inclusion around possible (Adelaja 2012). Because of these fac- the world, but the use of new technologies tors, cash is the main method of payment in has taken different paths across economies. Russia, while payments and transfers through GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 65 FIGURE 2.2 Mobile Phone Penetration and Mobile Payments a. Mobile phones and bill payments b. Mobile phones and sending money 30 70 Mobile phone used to send money, % of adults Mobile phone used to pay bills, % of adults 25 60 50 20 40 15 30 10 20 R 2 = 0.00823 5 10 R 2 = 0.01836 0 0 0 50 100 150 200 0 50 100 150 200 Mobile phone subscriptions per 100 people Mobile phone subscriptions per 100 people Sources: Calculations based on data from the World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development -indicators; Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. mobile phones are less frequent compared There is now wide agreement that no sin- with countries with similar levels of educa- gle strategy will work to enhance financial tion, wealth, technological advancement, and inclusion in all markets and economic envi- mobile phone penetration.8 ronments. Indeed, when researchers have Somalia is a quite different example. It has investigated conditions such as the amount the fourth-lowest mobile penetration rate in of regulation, the extent of legal rights, the the world, but is among the three countries cost of alternatives, market size, and the size with the highest usage of mobile phones for of the financially excluded population, the payments. This somewhat contradictory pic- results have been inconclusive (for example, ture largely reflects Somalia’s challenging see Flores-Roux and Mariscal 2010). This security conditions. When the country’s larg- raises the question: what enabling conditions est telecommunications company launched must be in place to support the development, mobile banking services, the services quickly for example, of mobile banking and payments proved to be of great benefit. Individuals could services sufficiently comprehensive to increase now easily transfer money to other subscrib- financial inclusion? ers, facilitating shopping and the payment While no single factor can explain the stark of bills without the need to carry cash. The cross-country differences in the adoption of new service has made it possible for Somalis mobile banking and payment technologies, to receive remittances from their family mem- some patterns do stand out. In many ways, the bers and friends abroad, an important change provision of financial services to lower-income given that remittances, equivalent to about customers has traditionally been similar to the 70 percent of the Somali GDP, have been the physical retail business. The profitability of backbone of Somalia’s war-torn economy providing financial services to the mass market (Maimbo 2006). The example of Somalia is determined by the number of potential cus- underscores that the degree of use of mobile tomers within the range of a physical branch phones for payments, while a good indicator, and the revenue per potential customer. should not be considered an end in itself. The first of these components is affected by 66 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 2.3 Share of Adults with an Account in a Formal Financial Institution 40 Adults, % 40 Population density, population per square kilometer 20 0 Access to formal account 1,000 Adults, % 40 30 20 0 Access to formal account 125 60 48 Adults, % Adults, % Adults, % 40 40 34 40 18 20 20 20 0 0 0 Access to formal account Access to formal account Access to formal account $4,000 $8,000 $15,000 GDP per capita, current US$ Sources: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex; World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators; Faz and Moser 2013. population density; the second is a function of This classification of financial inclusion the per capita income of potential customers. environments can highlight the settings in Population density and per capita income which mobile banking and mobile payments are two key factors that are systematically can have the greatest impact on financial inclu- correlated with financial inclusion. Figure 2.3 sion. It is possible to identify three broad finan- shows the percentage of adults who are 15 cial inclusion environments, corresponding to years or older and who have a formal bank three of the four corners in figure 2.3, in which account according to per capita income (hori- one would expect the adaptation of mobile zontal axis) and population density (vertical banking services to follow different paths: axis). If measured in terms of access to bank (1) low-income and low–population density accounts, financial inclusion is an increasing environments, characterized by the absence function of a country’s per capita income. of a widespread banking infrastructure and, There is also a strong positive relationship often, of a dominant telecommunications pro- between population density and financial vider; (2) low-income and high–population inclusion. This is largely explained by the density environments with a widespread net- economies of scale in countries such as Ban- work of banks and MFIs and well-developed gladesh, India, and Indonesia, where high telecommunications providers; and (3) high- population densities have been an enabling income and low–population density environ- condition for financial inclusion through tra- ments with a developed banking sector, a ditional bank branches, despite low levels of widespread organized retail sector, and strong per capita income. telecommunications providers. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 67 Mobile banking and payment technolo- to customers, and expanding access to more gies change the economics of banking because remote areas where the high population crite- they dramatically reduce the cost of providing rion may not hold. financial services. This reduction in transac- Harnessing the potential of new technolo- tion costs is especially large in environments gies for financial inclusion is perhaps most with low population densities and low per challenging in the third financial inclusion capita income, precisely the settings that have environment, characterized by high income been underserved by traditional providers of and low or medium population density (bot- financial services. These are the settings in tom right corner of figure 2.3). Examples which mobile banking technologies have the include much of Eastern Europe and Latin greatest potential welfare benefits because the America. In this environment, banks, retail- technologies offer a commercially viable way ers, and telecommunications providers tend of reaching locations and customers that were to be well established so that the introduction previously excluded from formal financial of new banking and payment technologies is services due to the prohibitively high costs often more likely to redistribute the market of providing such services. Kenya, the Philip- shares of the services provided to an already pines, and Tanzania are examples of markets banked population rather than to incorporate in which new technologies have had an instru- new client groups into the formal financial mental role in expanding financial inclusion. sector. Because of the value of existing bank- In the other parts of the matrix in figure ing relationships in such a setting, this is also 2.3, one may see that mobile banking and an environment in which entrenched provid- technology-based business models can still be ers of financial services tend to be reluctant important, but they are more likely to func- to adopt new technologies that could threaten tion as a complement to rather than a substi- their market position or displace existing tute for traditional bank-based financial inclu- technologies with clearly defined market sion policies. This point is most obvious in the shares and revenue structures, such as credit case of economies classified as high popula- cards. Hence, this financial inclusion environ- tion density and low income (upper left corner ment represents a much greater burden for of figure 2.3). Examples include Bangladesh, regulators, who need to ensure that, first, new India, Indonesia, and parts of China. In these technologies are adopted and, second, that economies, the large number of customers they are priced and made available in a way that can be served by a single bank branch that makes them accessible to the unbanked. compensates somewhat for comparatively Increasing the challenges is the fact that, in low average account balances and transaction this environment, the risk of regulatory cap- volumes. As a result, these economies have a ture is heightened. well-developed network of traditional bank Finally, an important prerequisite for the branches and microfinance providers that adoption of technologies that can enhance cater to low-income customers. Mobile bank- financial inclusion across all market environ- ing and technology-based business models ments is an adequate legal and regulatory can nonetheless help in expanding financial framework. The regulatory framework needs inclusion within these environments. India’s to create enabling conditions for the provid- financial inclusion policies, for example, ers of technology-based financial services, focus on a bank-led model that combines the while protecting the rights of consumers. The advantages of a widespread network of physi- case of mobile payment systems is a good cal bank branches with technology-based example: regulating new payment systems as correspondent banking solutions to reach if they were conventional banks is likely to previously unbanked locations and unbanked reduce competition and may create risks for segments of the population. Mobile technolo- consumers. The same applies to new technol- gies can act as a complement by decongest- ogies that make it easier to collect and share ing existing branches, bringing services closer borrower information. 68 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 THE IMPORTANCE OF but not under complete information. At the PRODUCT DESIGN same time, peer monitoring can help mitigate ex ante moral hazard by discouraging risky The design of financial products can have a investments. Giné, Jakiela, and others (2010) major impact on the use by individuals of have obtained similar results from games financial services. Recent studies show that conducted with borrowers in Peru. Giné and product design features can affect both the Karlan (2009), based on experiments in the extent and the impact of the use by individu- Philippines, show that joint liability is not als of financial services.9 necessary to achieve high repayment rates. Another important issue in credit prod- Credit uct design is related to the timing of repay- ments. Field and others (forthcoming) have Certain design features of credit products conducted a randomized evaluation of the aid in mitigating market imperfections and effect of offering a grace period prior to the increasing inclusion. For example, products start of loan repayment. In particular, they that help reveal hidden information assist in offered one set of borrowers from the Vil- channeling credit by mitigating asymmetric lage Welfare Society in West Bengal, India, a information problems. Drugov and Macchia- traditional group microcredit loan with semi- vello (2008) and Ghosh and Ray (2001) show weekly payments with no grace period, while how small, initial tester loans can provide a second set of borrowers received loans with information that is useful for assessing the a two-month grace period. The authors found risk in subsequent larger loans. that the borrowers offered the grace period Group lending is an often-discussed and invested 6 percent more of their loans relative arguably the most controversial solution to to the borrowers who were not given a grace information asymmetries in developing econ- period; the former set saw an average of 30 omies. Ghatak (1999) shows how group lend- percent higher profits. Household income was ing can mitigate adverse selection. In select- also higher, on average, among the borrowers ing fellow borrowers with whom they will be given the grace period. However, these aver- jointly liable for loans, potential clients will age results mask significant differences within exploit information known to borrowers, the grace-period borrowers: while some bor- but not to banks, so as to screen out bad bor- rowers did well, others did not, and 9 percent rowers. Group lending also addresses moral of the individuals among the grace-period bor- hazard by providing incentives for clients to rowers defaulted on their loans relative to the employ peer pressure to ensure that funds are 2 percent default rate among the borrowers invested properly and effort exerted until the not given a grace period. The experiment sug- loans are repaid. Karlan (2007) uses quasi- gests that, while a grace period can allow bor- random variation in the group-formation pro- rowers to invest more and, hence, sometimes cess at a Peruvian MFI to show that groups obtain higher profits, institutions will have to with greater levels of social connection (ethnic balance this out against the potential losses ties and geographical proximity) have lower associated with the higher default rates among default and higher savings rates. There are such borrowers, perhaps by raising loan rates. also concerns, however, that group lending While grace periods seem to have nega- may create problematic incentives at the com- tive consequences on default rates, dynamic munity level, such as the use of coercion to incentives—whereby borrowers are subject oblige family or neighbors to participate.10 to future penalties (no access to loans) or Based on a series of investment games in rewards (discounts on loan prices) depending India, Fischer (2008) finds that joint liabil- on their present repayment behavior—seem ity (a characteristic of group lending) pro- to improve loan repayment. In one study in duced free-riding and higher levels of risk South Africa, Karlan and Zinman (2010) taking under limited information conditions, worked with a lender who randomly offered GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 69 some clients a dynamic incentive, a discount treatment led to increases in deposits at the on future loans if the clients repaid the current bank and, over the next agricultural year, was loans. This offer led to a 10 percent reduction associated with rises in agricultural input use, in the default rate, and the responsiveness was crop sales, and household expenditures. The proportional to the size of the incentive. commitment savings accounts seemed primar- ily to have helped farmers less by mitigating self-control issues than by shielding funds Savings from the social networks of the farmers, A number of psychological factors help because participants who were identified with explain people’s limited use of savings prod- self-control issues experienced no different ucts. One is lack of self-control: people want effect from the commitment savings accounts to save, but self-control issues make it dif- relative to their peers. On the other hand, the ficult to resist the temptation to spend the commitment savings accounts had a higher cash immediately. This type of behavior is impact on wealthier individuals, who may interpreted as a sign of hyperbolic discount- have faced greater pressure to share funds. ing, whereby people disproportionally value Other innovations in savings product today’s money over tomorrow’s (Laibson design try to pin the attention of savers on 1997). Another explanation is that individu- long-term savings goals to reduce the ten- als face pressures from family members and dency to become distracted and spend funds others to share their excess funds, which eats in the short term. Two approaches have been away at their savings. Finally, lack of fore- tested in recent studies. One involves the use thought may make people lose sight of the of reminders, and the other involves offering fact that they might need savings in the future. labeled accounts, whereby individuals create Commitment savings accounts, which accounts with explicit savings goals, such as allow individuals to deposit a certain amount to finance housing or education. and relinquish access to the cash for a period Accounts with general automatic savings of time or until a goal has been reached, have reminders lead to a boost in savings, but been examined as a possible tool to boost accounts with specific goal reminders are even savings by mitigating the self-control issues more effective in raising savings. Karlan and and family pressures to share windfalls. The others (2010) find that reminders can influ- evidence comes primarily from two studies: ence the use of deposit accounts for savings. Ashraf, Karlan, and Yin (2006) and Brune and They designed field experiments with three others (2011). The first study involved a ran- banks, one each in Bolivia, Peru, and the domized evaluation of commitment accounts Philippines. In each experiment, individuals in a financial institution in the Philippines. opened a bank savings account that included The take-up of the accounts was 28 percent. varying degrees of incentives or commitment There was one treatment group, and the study features designed to encourage individuals did not show differential take-up. to reach a savings goal. Some individuals The second study focused on an institution were randomly assigned to receive a monthly in Malawi, with take-up of the commitment reminder via text message or letter, while a accounts at 21 percent. The study had two control group received no reminder. Remind- treatment groups; one was offered a check- ers increased the likelihood of reaching a sav- ing account only, and the other a checking ings goal by 3 percent, and the total amount account, plus a commitment savings account. saved in the reminding bank by 6 percent. There was no differential take-up between the Reminders that highlighted the client’s par- two groups, but the group with the check- ticular goal, that is, reminders that made a ing and commitment accounts saved more. particular future expenditure opportunity, In other words, the observed results did not such as school fees, more salient, were two arise because of differences in take-up. Brune times more effective than reminders that did and others (2011) found that the commitment not mention the goal. 70 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Labeling, a practice of designating a spe- the first place (adverse selection), and access cific savings goal for an account, has been to insurance coverage may lead households to shown to be highly effective in increasing behave in ways that make a negative outcome savings. Karlan, Kutsoiati, and others (2012) more likely (moral hazard). Because insurance offered a random group of clients at a bank providers can often form only a poor estimate in eastern Ghana the opportunity to open of an individual client’s true risk profile, the separate savings accounts labeled according market price of insurance is often substan- to specific goals. The study found that the tially higher than the actuarially fair value of treatment group that had access to labeled the contract, which is the price that would accounts saved 31 percent more, on average, be justified given a customer’s risk profile. In than the control group. many developing economies, this means that Basic accounts (accounts with low fees basic insurance products either are not avail- and minimum requirements), if well designed, able, or are outside the reach of the most vul- can be valuable in encouraging the use of nerable households. accounts. Germany, the United Kingdom, An important innovation in agriculture and numerous other European countries use insurance has been the introduction of index- basic savings accounts as entry products. In based insurance products. Traditional agricul- the United States, the Bank On public-private ture insurance products have been indemnity partnership program of generic accounts based, meaning the company insures against through commercial banks has considerably crop loss or damage. The farmer buys insur- expanded the use of basic accounts by pre- ance up to a given amount of loss, and, if an viously unserved consumers.11 Some of the event materializes that leads to that level of efforts aimed at generic products were not loss, the farmer receives the insurance once the successful; the experiment in Brazil is one validity of the claim is established. The prob- such example. Several countries currently lem is that the moral hazard and adverse selec- have savings promotions associated with lot- tion problems involved in this kind of insur- teries that serve as a simple commitment sav- ance often lead to rationing or high premiums. ings product. By determining payouts based on an objective rainfall-index, for example, index-based insur- ance can circumvent some of the problems of Insurance indemnity-based insurance (box 2.4). Design features also matter in insurance prod- In many cases, index insurance products ucts. Households in developing economies are differ from other types of insurance policies exposed to risks that can generate extreme that consumers may be used to. Understand- income volatility. This is particularly true in ing the potential benefits that index insurance the case of covariate risks, such as droughts or contracts offer over conventional insurance natural disasters, which affect large geograph- requires a relatively high level of financial lit- ical areas or large segments of the population eracy. In practice, this has often meant that and are not adequately covered by informal the take-up of index insurance has been slow insurance mechanisms. Modern insurance and has not yet had the widespread impact products can substantially reduce the result- envisioned by policy makers when these prod- ing welfare losses. However, much of this ucts were first introduced. potential remains unfulfilled because extend- ing access to basic insurance products among THE IMPORTANCE OF vulnerable populations is extremely challeng- BUSINESS MODELS ing (Cole and others 2013). An important barrier to access derives from the fact that The business models used by banks and other insurance markets are plagued by moral haz- financial institutions can have a considerable ard and adverse selection: riskier clients are impact on financial inclusion. Throughout more likely to demand and buy insurance in this report, there are numerous examples GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 71 BOX 2.4 Insurance: Designing Appropriate Products for Risk Management Two main approaches have been proposed to expand uninsured loan was uncorrelated with these farmer access to basic insurance products. The fi rst is tar- characteristics. If one takes into account the lender’s geted at reducing markups by attenuating informa- perspective, a clearer picture emerges. For the lender, tion problems. For instance, the geospatial mapping weather insurance tends to be an attractive way to of weather risks can reduce uncertainty in calculating mitigate default risk. It can thus become an effective insurance premiums. Bundling insurance with other risk management tool with the potential of increasing fi nancial services can provide better client informa- access to credit in agriculture at lower prices. tion, allowing for more accurate risk assessments, In India, the popularity of rainfall insurance has which reduce the cost of providing insurance. The risen, although the growth in usage has been rather second approach relies on a new class of insurance limited. Giné, Menand, and others (2010) exam- contracts, index insurance. In contrast to traditional ine rainfall insurance in India’s Andhra Pradesh insurance contracts, payouts for this type of product using data from BASIX, an MFI and bank that sells are linked to a measurable index, such as the amount weather insurance policies in Andhra Pradesh on of rainfall over a given period or commodity prices at behalf of ICICI Lombard. Figure B2.4.1 plots the a given date. Index insurance represents a particularly growth in rainfall insurance and livestock insur- attractive alternative to financial innovations because ance (as a point of comparison) sold by BASIX since it eliminates problems of moral hazard (payouts occur 2005. Over this period, livestock insurance coverage according to a measurable index that is beyond the has risen fivefold, compared with an approximately control of the policyholder). Index insurance is also 50 percent increase in rainfall insurance cover- especially well suited to provide insurance against age. This is not simply due to a difference in value, adverse shocks that affect many members of informal because, as reported in fi gure B2.4.2, the payouts insurance networks at the same time. on rainfall insurance are greater relative to the pre- Analyzing specific index insurance products, miums than is the case in livestock insurance. Two Giné and Yang (2009) implemented a randomized facts are apparent from the figure. First, the average field experiment to examine whether the provision of payouts on rainfall insurance are much more vola- insurance against a major source of production risk tile, reflecting aggregate variation in the intensity of induces farmers to take out loans to adopt a new crop the monsoon. Second, average returns on the insur- technology. The study sample consisted of roughly ance product are quite high over this period, higher 800 maize and groundnut farmers in Malawi, where the dominant source of production risk is the level FIGURE B2.4.1 Growth in Livestock and Weather of rainfall. Half the farmers were randomly selected Microinsurance, India to receive credit to purchase high-yielding hybrid maize and groundnut seeds for planting; the other 600 half were offered a similar credit package, but were 500 also required to purchase (at actuarially fair rates) a Index, 2005 = 100 weather insurance policy that partially or fully for- 400 gave the loan in the event of poor rainfall. Surpris- 300 ingly, the take-up was lower by 13 percentage points 200 among farmers offered insurance with the loan. The take-up was 33 percent among the farmers who were 100 offered the uninsured loan. There is evidence that the 0 lower take-up of the insured loan arose because farm- 2005 2006 2007 2008 2009 ers already had implicit insurance through the limited Livestock Weather liability clause in the loan contract: insured loan take- up was positively correlated with farmer education, Source: BASIX administrative data. Note: Total nominal premiums (in Rs) paid on livestock and weather microinsur- income, and wealth, which may proxy for the indi- ance policies. Rainfall insurance data are for Andhra Pradesh only; livestock vidual’s default costs. By contrast, the take-up of the insurance data are for all states. (box continued next page) 72 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.4 Insurance: Designing Appropriate Products for Risk Management (continued) FIGURE B2.4.2 Payouts Relative to Premiums, Sygenta Foundation for Sustainable Agriculture and Rainfall and Livestock Insurance, India launched in partnership with Safaricom (the largest mobile network operator in Kenya) and UAP (a large 4.0 Kenya-based insurance company). Kilimo Salama is 3.5 notable because it is the fi rst microinsurance product Payouts divided by premiums 3.0 distributed and implemented over a mobile phone 2.5 network. Though there has been no formal evalu- 2.0 ation of the product, the rapid growth in the num- 1.5 ber of users suggests that the product seems to ben- efit farmers. The program was piloted in 2009, and 1.0 take-up grew from an initial 200 farmers to over 0.5 12,000 in 2010. Kilimo Salama is featured in one of 0 2003 2004 2005 2006 2007 2008 2009 the Product Design Case Studies of the International Finance Corporation.a Livestock Weather Recent research has supplied fresh insights on Source: BASIX administrative data. household participation in insurance markets, which Note: The figure shows the payouts relative to the total premiums paid for insurance policies sold by BASIX across all states. can help guide the design of new products and poli- cies. Using data from a fi eld experiment in India, Cole and others (2013) fi nd that lack of trust and liquidity constraints are significant nonprice fric- than actuarially fair based on a simple average of tions that constrain demand. There are several payout ratios across these years. This may reflect implications of this research for the design of insur- some unusual shocks over the past few years, par- ance products. First, products need to be designed ticularly the record drought in 2009. Alternatively, it to pay fairly often so as to engender trust in the user may reflect structural change in weather conditions, population. Also, an endorsement by a well-regarded such as a rise in the volatility of the monsoon. institution has been shown to increase client trust. A notable weather insurance product is Kilimo Second, because liquidity constraints matter, rapid Salama, an index-based insurance product that cov- payouts are important. Because of these constraints, ers the input of farmers in the event of drought or it may also be useful to bundle insurance with loans excessive rainfall. It was developed in Kenya by the for payment of the premiums. a. See “Product Design Case Studies,” International Finance Corporation, Washington, DC, http://www1.ifc.org/wps /wcm/connect/industry_ext_content/ifc_external_corporate_site/industries/fi nancial+markets/publications/product+ design+case+studies. of financial institutions that have targeted chain owned by the same mother company, lower-income segments of populations. These which has allowed the bank to use existing include BRAC and Grameen Bank in Bangla- customer information and collection tech- desh, Banco Sol in Bolivia, the Bradesco and nology (chapter 3, box 3.3). BBVA does not Caixa banks in Brazil, Equity Bank in Kenya, have such a retail company; it has relied on Banco Azteca and BBVA in Mexico, and the introduction of banking agents, allow- many others. ing commercial establishments to offer basic While all these institutions target the financial services on behalf of the bank. unbanked and underbanked, the emphasis Its strategy emphasizes a simple, low-cost of the individual business models differs sub- account targeted at low-income population stantially. For example, Banco Azteca’s strat- segments in Mexico. The simplified account egy has relied on synergies with a large retail opening process has boosted sales to 2 million GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 73 accounts (including 62 percent purchased in sent mobile ones, such as armored trucks with states with the highest poverty rates) since the a satellite dish on top and a bank manager launch of the process in March 2011 (Alonso inside. New standards of service created a and others 2013). These new account hold- large and loyal customer base. Since 2000, the ers benefit from BBVA’s alternate channel net- bank’s pretax profit has grown at an annual work, with most of transactions (82 million) average rate of 65 percent. Today, roughly done at ATMs and banking agents. The bank half of all bank accounts in Kenya are with is now launching a micro-life insurance prod- Equity. Nonperforming loans were only 1.3 uct, sold through banking agents. percent of the loan portfolio in 2011 (The The various institutions also differ in how Economist, December 8, 2012). they balance the trade-off between outreach A recent study on Equity Bank, based on (the ability to reach poorer and more remote household surveys and bank penetration data people) and sustainability (the ability to cover at the district level in 2006 and 2009, exam- operating costs and possibly create profits). ines the impact of Equity Bank on financial There is a wide variety of strategies, ranging inclusion in Kenya (Allen and others 2012a). from the for-profit (and profitable) operations The study finds that the bank’s presence has of commercial banks such as Banco Azteca, had a positive and significant impact on the BBVA, and Equity Bank to the clearly not-for- use by households of bank accounts and bank profit orientation of developmental financial credit, especially among Kenyans with low institutions such as BRAC. income, no salaried job, and lower educa- This section focuses in more detail on one tional attainment, and Kenyans who do not example of a financial institution that pur- own their own homes (figure 2.4). sues distinct branching strategies that target To what extent can Equity Bank’s business underserved areas and less-privileged house- model—providing financial services to popu- holds. The institution, Equity Bank, is a pri- lation segments typically ignored by tradi- vate commercial bank in Kenya that focuses tional commercial banks and generating sus- on microfinance.12 tainable profits in the process—help in solving In 1994, Equity Bank’s predecessor, Equity financial access challenges in other countries? Building Society, became technically insol- Equity has yet to replicate its Kenyan success vent. Because of a combination of economic abroad. For example, in the past four years, downturn and poor management, more than it has lost $359,000 on the $96 million it half its loan portfolio had gone bad, and its has invested to build an East African plat- accumulated losses were 10 times the avail- form. The return on equity in Uganda was able capital. The Central Bank of Kenya gave 18 percent in 2011, a positive number, but Equity time to convert its depositors into lower than the 34 percent the bank recorded shareholders. A new chief executive shifted in Kenya in the same period. It remains to be the organization’s focus from the competitive seen to what extent Equity’s model based on mortgage market to small loans. The back- banking for the unbankable is exportable. bone of the new strategy was to offer Kenya’s Though some financial service providers large unbanked population microloans from such as Equity Bank have developed business as little as K Sh 500 ($5.81); the average loan models and products that are specifically tar- amount was K Sh 16,000. geted at low-income individuals, the prod- In what followed, Equity bucked the trend uct designs and business models of financial of branch closures around the country. It institutions are often criticized, especially in waived property-ownership requirements and the developing world, because they are not allowed anyone with a national identity num- demand driven.13 A common complaint is ber to open an account. It was flexible about that most of the products offered are not forms of collateral, accepting marriage beds or tailored to the needs of low-income clients, personal belongings. Some customers repaid but rather look remarkably similar to prod- loans with cow’s milk. Where there were too ucts offered to high-end clients in more well- few customers for it to build branches, Equity developed markets. 74 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 2.4 Equity Bank’s Effect on Financial Inclusion 8 6.9 7 household having a bank account, % 5.8 6 Changes in the probability of a 5.3 5.2 5 4 3 2 1.1 1 0.5 0.6 0.2 0 Low asset High asset Does not Owns No secondary Has secondary Nonsalaried Salaried job score score own house house or tertiary or tertiary job education education Source: Based on Allen and others 2012a. Note: The figure shows estimates from a probit model of changes in the probability of a household having a bank account that are associated with the presence of Equity Bank. A number of factors explain the failure design and delivery. While regulators should of institutions to deliver products and adopt rightly be concerned with financial stability business models that are more conducive to and consumer protection, they should remain financial inclusion.14 First, institutions often open to offering support for innovative prod- do not pursue a human-centered design pro- ucts and approaches to financial inclusion. cess that requires them genuinely to engage They should build a regulatory framework with customers and understand their lives and that is proportionate with the risks involved in needs.15 innovative products and services and is based Second, a demand-driven approach to on an understanding of the gaps and barriers product design requires the entire organiza- in existing regulations (GPFI 2011a). Aside tion to be centered on the client, including from adopting the proper regulatory stance, governance, human resources, delivery chan- governments can also support innovations nels, processes, incentives, and core systems. in product design and delivery by collecting This implies a degree of commitment that is data on individual preferences and habits in often misunderstood by financial institutions. financial products that financial institutions Third, institutions need to be willing to try can use to tailor new products and business different ideas and be prepared for some to approaches (Group of 20 Financial Inclusion fail. In other words, institutions need to cre- Experts 2010). ate space for testing innovations outside the core business and for learning from failed FINANCIAL CAPABILITY products. Fourth, institutions need to feel that the As the use of financial services expands and regulatory environment supports innovation new products become available, it is crucial in product design and delivery. Hence, gov- that consumers be financially literate and ernments can play an important role in either capable. The global financial crisis has led to promoting or stifling innovations in product the insertion of financial literacy and financial GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 75 capability onto the global policy agenda of the concept of inflation relatively better than financial regulators out of recognition that one the concept of compound interest. This is of the contributing factors to the financial tur- especially true in upper- and middle-income moil was the fact that vulnerable consumers countries, where, in almost all cases, more (many of whom were termed subprime in the people correctly answered the inflation ques- United States) had been marketed loans they tion, and fewer people correctly answered the often did not understand and were unable to question on compound interest. The question service. Similar concerns were raised in emerg- on the concept of risk diversification tended ing markets such as India, where the rapid to be the one that most people in high- growth of microfinance through aggressive income countries answered incorrectly, per- marketing was seen as leading to overindebt- haps because it involves stocks, which many edness among the rural poor, with sometimes consumers, even in developed countries, do tragic consequences (chapter 1, box 1.5). not hold. Policy makers around the globe now recog- nize the importance of financial capability and Evidence on the financial mistakes of financial education.16 Comprehensive national consumers initiatives and programs funded by the World Bank and various development donors are Evidence on consumer credit card and mort- emerging all over the world. gage markets suggests that both consumer The term financial capability tends to lack of information and irrationality lead to encompass concepts ranging from financial substantial errors in financing choices that knowledge (including knowledge of finan- are accentuated by the design of products by cial products, institutions, and concepts), financial providers who want to exploit short- financial skills (such as the ability to calculate comings in understanding among consumers. compound interest payments), and financial Consumer credit can enhance household capability more generally (which includes all welfare by allowing for consumption smooth- the skills, attitudes, and behaviors that enable ing over time, but there are many reasons individuals to use financial services to their why growth in consumer credit may, at times, advantage). Financial knowledge does not be a concern (Bar-Gill and Warren 2008). For necessarily translate into wise financial behav- example, research has shown that consumers ior, that is, financial capability. In practice, are frequently ignorant about many of the these notions often overlap. features of the products they use; they do not always make good decisions; and credit pro- viders often exploit the tendency for consum- Basic financial knowledge around the ers to make mistakes. In particular, consumer world biases such as the exponential growth bias A broad review of survey data finds that basic and cognitive limitations such as the lack of financial knowledge is lacking in both devel- financial literacy lead to inefficient consumer oped and developing economies (table 2.1). credit market outcomes and overindebtedness On average, only about 55 percent of individu- (Lusardi and Tufano 2009; Stango and Zin- als demonstrate a basic understanding of com- man 2009). pound interest; 61 percent correctly answer a There is ample evidence of mistakes among basic question about the effect of inflation on consumers in the use of credit cards. For savings; and 49 percent respond correctly to example, Shui and Ausubel (2004) find that a a basic question on risk diversification. While majority of consumers who accept credit card there are substantial differences across coun- offers featuring low introductory rates do not tries, low-income countries tend to be at the switch to a new card with a new introductory bottom end of the performance rankings. rate after the expiration of the introductory This research indicates that basic financial period on the first card, even if their debt knowledge is weak. People tend to understand did not decline after the introductory period 76 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE 2.1 Financial Knowledge around the World Q1:a Compound Q2:b Q3:c Risk Survey Economy, year interest, % infl ation, % diversification, % sample, total Source High income British Virgin Islands, 2011 63 (20) 74 41 535 OECD survey Czech Republic, 2010 60 (32) 80 54 1,005 OECD survey Estonia, 2010 64 (31) 86 57 993 OECD survey Germany, 2009 82 78 62 1,059 Bucher-Koenen and Lusardi 2011 Germany, 2010 64 (47) 61 60 1,005 OECD survey Hungary, 2010 61 (46) 78 61 998 OECD survey Ireland, 2010 76 (29) 58 47 1,010 OECD survey Italy, 2006 40 60 45 3,992 Fornero and Monticone 2011 Japan, 2010 71 59 40 5,268 Sekita 2011 Netherlands, 2010 85 77 52 1,324 Alessie, Van Rooij, and Lusardi 2011 New Zealand, 2009 86 81 27 850 Crossan, Feslier, and Hurnard 2011 Norway, 2010 75 (54) 87 51 2,117 OECD survey Poland, 2010 60 (27) 77 55 1,008 OECD survey Sweden, 2010 35 60 68 1,302 Almenberg and Save-Soderbergh 2011 United Kingdom, 2010 61 (37) 61 55 1,579 OECD survey United States, 2009 65 64 52 1,488 Lusardi and Mitchell 2011a Upper-middle income Albania, 2011 40 (10) 61 63 1,008 OECD survey Azerbaijan, 2009 46 46 — 1,207 World Bank Bosnia and Herzegovina, 2011 22 58 — 1,036 World Bank Bulgaria, 2010 39 46 — 1,618 World Bank Chile, 2006 2 26 46 13,054 Behrman and others 2010 Colombia, 2012 26 69 — 1,526 World Bank Malaysia, 2010 54 (30) 62 43 1,046 OECD survey Mexico, 2012 31 56 — 2,022 World Bank Peru, 2010 40 (14) 63 51 2,254 OECD survey Romania, 2010 24 43 — 2,048 World Bank Russian Federation, 2009 36 51 13 1,366 Klapper and Panos 2011 South Africa, 2010 44 (21) 49 48 3,017 OECD survey Lower-middle income and lower income Armenia, 2010 53 (18) 83 59 1,545 OECD survey India, 2006 59 25 31 1,496 Cole, Sampson, and Zia 2011 Indonesia, 2007 78 61 28 3,360 Cole, Sampson, and Zia 2011 Mongolia, 2012 58 39 60 2,500 World Bank Tajikistan, 2012 56 17 52 1,000 World Bank West Bank and Gaza, 2011 51 64 — 2,022 World Bank Source: An expanded and updated version of Xu and Zia 2012. Note: Countries are ordered alphabetically. Additional information on World Bank surveys of financial capability can be found at Responsible Finance (database), World Bank, Washington, DC, http://responsiblefinance.worldbank.org/. OECD = Organisation for Economic Co-operation and Development. — = not available. The questions have been worded as indicated in the table notes below. The correct answers are shown in italics. a. Q1: “Suppose you had $100 in a savings account, and the interest rate was 2 percent per year. After five years, how much do you think you would have in the account if you left the money to grow? (More than $102 /Exactly $102/Less than $102/Do not know/Refuse to answer).” Respondents in Azerbaijan, Chile, Romania, Russia, and Sweden were asked a more difficult question. For example, in Sweden, the question was “Suppose you have SKr 200 in a savings account. The interest is 10 percent per year and is added into the same account. How much will you have in the account after two years?” The OECD data for compound interest in the table include correct responses to a question on the calculation of interest and principal and, in the parentheses, the correct responses to two questions on compound interest. b. Q2: “Imagine that the interest rate on your savings account were 1 percent per year, and inflation were 2 percent per year. After one year, how much would you be able to buy with the money in this account? (More than today/Exactly the same/Less than today/Do not know/Refuse to answer).” In Russia and in the West Bank and Gaza the question was “Let’s assume that, in 2010, your income is twice as much as now, and consumer prices also grow twofold. Do you think that, in 2010, you will be able to buy more, less, or the same amount of goods and services as today?” c. Q3: ”True or false: Buying a single company’s stock usually provides a safer return than a stock mutual fund. (True/False /Do not know/Refuse to answer).” Respondents in New Zealand were asked a more difficult question: “Which one of the following is generally considered to make you the most money over the next 15 to 20 years: a savings account, a range of shares, a range of fixed interest investments, or a checking account?” Respondents in Russia were asked: “Which is the riskier asset to invest in: shares in a single company stock, or shares in a unit fund, or are the risks identical in both cases?” In Indonesia and India: “Do you think the following statement is true or false? For farmers, planting one crop is usually safer than planting multiple crops.” The result for Italy is from a 2008 update of the survey, since a comparable question was not asked in 2006. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 77 ended. Gross and Souleles (2002) show that of consumers. Examples of such practices many consumers pay high interest rates on include low amortization rates, compound large credit balances, even if they are holding interest rates, and deceptive methods of bal- liquid assets in deposit accounts. Massoud, ance computation. Saunders, and Scholnick (2007) find simi- Unsuitable products can have significant lar results. Agarwal and others (2009), after distributive consequences. Because better edu- analyzing a representative random sample of cated consumers are less likely to make mis- about 128,000 credit card accounts between takes in their financing choices, financial January 2002 and December 2004, conclude institutions are more likely to target risky that more than 28 percent of consumers make and costly products at poor, less well edu- mistakes that trigger fees, including late fees cated consumers who are more likely to suffer and cash advance fees. from mistakes. Empirical evidence on the U.S. Mortgage loans, which are generally more mortgage market supports the notion that complex than other types of household credit, more well educated consumers are less likely provide larger opportunities for errors. In the to make mistakes (Van Order, Firestone, and context of the U.S. subprime crisis, the evi- Zorn 2007; Woodward 2003). Similar evi- dence indicates that a large percentage of bor- dence is available on the credit card market rowers took subprime mortgages when they (Massoud, Saunders, and Scholnick 2007). could have qualified for prime rate loans (Wil- lis 2006). In 2002, a study by the National Measuring financial capability Training and Information Center found that, at a minimum, 40 percent of those borrow- In recent years, progress has been made in ers who were issued subprime mortgages defining and measuring financial capabil- could have qualified for a prime market loan ity more precisely. An important part of this (involving a lower interest rate). A study based effort is the World Bank–Russia Trust Fund on 75,000 home equity loans made in 2002 multicountry survey of financial capability identified persistent consumer mistakes in (Holzmann, Mulaj, and Perotti 2013).17 The the estimation of home values in loan appli- survey was designed to reflect inputs from cations, which raised the loan-to-value ratio focus groups among low-income consumers and thus the interest on the loan (Agarwal and in developing economies. It covers three key others 2009). Many studies document that areas: behavior, attitudes, and motivations. consumers fail to exercise options to refinance The focus groups and the survey development their mortgages and end up with rates that are process began in 2011, and the surveys were higher than the market rates (Van Order, Fire- carried out in seven countries in 2012–13 stone, and Zorn 2007). (Armenia, Colombia, Lebanon, Mexico, Nige- Evidence on the behavior of providers of ria, Turkey, and Uruguay). Similar surveys, financial products reinforces the notion that with additions to measure financial inclusion consumers make poor choices, and providers and consumer protection, were conducted in exploit the imperfect information and irra- Mongolia and Tajikistan.18 tionality of consumers in designing financial The survey demonstrates the financial vul- products. For instance, evidence on the credit nerability of a significant number of people, card market indicates that, as credit card debt even in relatively wealthier countries. For among consumers grows and interest rates example, in most of the middle-income coun- become a more salient feature of credit cards tries surveyed, a quarter or more of the popu- in the view of consumers, issuers offer cards lation indicated that they borrowed for neces- with low interest rates but high fees to attract sities (figure 2.5). In most cases, people with and retain clients (Evans and Schmalensee lower educational attainment were more likely 1999). Other features of credit card con- to rely on borrowing to cover basic needs, but tracts are also designed to exploit the imper- even among people with some university edu- fect information and imperfect rationality cation, 20 percent or more indicated that they 78 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 2.5 Borrowing for Food and Other Essentials, by Level of Education 100 80 38 41 65 72 69 78 63 76 79 72 67 77 65 65 68 37 50 42 61 72 76 Share of all respondents, % 60 29 40 36 25 26 33 32 28 55 44 53 29 20 30 22 27 33 23 24 21 24 21 21 20 20 6 11 6 8 6 10 0 5 4 1 4 2 4 2 3 3 3 4 4 3 y y y y y y y y y y y y y y y y y y y y y ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar ar im nd rti im nd rti im nd rti im nd rti im nd rti im nd rti im nd rti Te Te Te Te Te Te Te Pr co Pr co Pr co Pr co Pr co Pr co Pr co Se Se Se Se Se Se Se Armenia Colombia Lebanon Mexico Nigeria Turkey Uruguay Regularly Sometimes No Sources: Based on data of the World Bank Survey of Financial Capability 2012; World Bank 2013a. Note: Primary, secondary, and tertiary refer to the respondent’s highest education. The chart shows responses to the following question: “Do you ever use credit or borrow money to buy food or essentials?” had sometimes been forced to borrow to make budgeting, living within means, monitoring ends meet. expenses, not overspending, using informa- There was a remarkable degree of consen- tion, covering unexpected expenses, saving, sus in the focus groups across the countries. attitudes toward the future, not being impul- Quite consistently, the focus group responses sive, and achievement orientation. suggested that good financial behavior is not The results of the financial capability sur- necessarily driven by financial knowledge: veys show that, while most respondents say individuals can have financial knowledge but they are relatively capable of day-to-day still make irrational financial decisions. Finan- money management (as indicated by the rela- cial capability is also not necessarily linked to tively high scores on “not overspending” and income (even though low income can be criti- “living within means” in figure 2.6), they lag cal in preventing people from planning for the in terms of self-discipline in making a bud- future). It is thus important to capture atti- get and then monitoring expenses and regu- tudes and motivations as well as experiences. larly saving. The survey also confirms that A major goal of the research was to mea- older people are more likely than younger sure financial capability in middle-income people to live within their means and not countries in a way that works both across overspend. Higher household income is asso- countries and across different subpopula- ciated with higher scores in most areas, but tions within the countries. A factor analysis it does not seem to matter for budgeting or of the data generated by the country surveys impulsiveness. was used to construct measures of compo- Overall, the survey highlights the impor- nents of financial capability that were aligned tance of the psychological traits and motiva- with the key concepts identified through the tions associated with financial capability, focus groups. These components included including one’s attitude toward the future, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 79 FIGURE 2.6 Survey Results on Financial Capability a. Not overspending b. Living within means 100 100 84 82 81 82 81 80 78 80 75 78 71 72 Average response, 0–100 Average response, 0–100 69 70 64 67 60 60 40 40 20 20 0 0 nia bia n ico ria y y nia bia n ico ria y y rke ua rke ua no no ge ge ex ex me lom me lom ug ug ba ba Tu Tu Ni Ni M M Ur Ur Ar Le Ar Le Co Co c. Monitoring d. Saving 100 100 80 80 Average response, 0–100 Average response, 0–100 60 60 60 55 55 48 49 47 44 45 43 42 39 40 35 39 40 23 20 20 0 0 nia bia n ico ria y y nia bia n ico ria y y rke ua rke ua no no ge ge ex ex me lom me lom ug ug ba ba Tu Tu Ni Ni M M Ur Ur Ar Le Ar Le Co Co Sources: World Bank Survey of Financial Capability 2012; Kempson, Perotti, and Scott 2013. Note: The figure illustrates the average among respondents, on a range from 0 to 100. The score for each component is a weighted combination of variables generated from the responses to several questions about different, but related behaviors or attitudes. impulsiveness, and goal orientation. The sur- and Lusardi 2011; Klapper and Panos 2011; vey findings corroborate earlier research on Xu and Zia 2012). There is broad evidence, financial capability in developed economies much of it from the United States and other such as the Netherlands, the United King- developed economies, that women score lower dom, and the United States.19 Across coun- on tests of financial literacy (Fonseca and oth- tries, a few key behavioral skills have been ers 2012; Van Rooij, Lusardi, and Alessie found to be crucial, in particular, money 2011). There is also evidence on links between management, budgeting, and the ability to levels of financial literacy and financial behav- save for the future. iors such as pension planning and investment There is widespread evidence that gender decisions (Dwyer, Gilkeson, and List 2002; influences the levels of financial literacy and Lusardi and Mitchell 2008, 2011b). capability, with only a few exceptions, such as The World Bank Survey of Financial Capa- eastern Germany and Russia (Bucher-Koenen bility provides new data and insights on the 80 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 role of gender in financial capability, using Solid general education is key data from developing economies. An analysis Solid general education, including numeracy, of the pooled World Bank data indicates that has a clear relationship with financial inclu- women achieved higher scores than men on a sion. According to Global Findex data for number of key financial behaviors, including 2011, only 37 percent of adults with primary budgeting, using information, and saving. or lower educational attainment had accounts Research recently published on farmers at formal financial institutions, compared in India shows no substantial gender bias in with 63 percent among adults with secondary financial capability if other characteristics educational attainment and 83 percent among (education, experience with financial prod- adults with tertiary or higher educational ucts, cognitive ability) are also measured attainment.23 These differences are massive (Gaurav and Singh 2012). The study points and are larger than the differences according to the fact that gender is often correlated with to other characteristics, such as gender, rural other factors that influence financial skills and versus urban residence, and income. This siz- opportunities (level of access to and use of able effect of education is confirmed by more financial services, educational attainment, for- rigorous, in-depth analysis. For example, mal employment). Disentangling the impact Allen, Demirgüç-Kunt, and others (2012) of gender from the impact of one’s environ- find that the probability of owning a bank ment, which is influenced or constrained by account is 12 percent lower among adults gender, is important in understanding the who have 0–8 years of education than among barriers faced by women in gaining finan- other adults after the authors control for age, cial capability and then using their skills and level of income, and many other factors. Simi- knowledge to improve financial outcomes. larly, Cole, Paulson, and Shastry (2012) show that the level of general educational attain- FINANCIAL EDUCATION ment has a strong effect on financial market PROGRAMS participation. The level of educational attainment also Interest in financial literacy has risen expo- has a clear, though imperfect link to financial nentially in recent years as evidenced, for capability. Measures of financial capability example, by the large number of new stud- tend to be positively correlated with educa- ies and literature reviews.20 There is growing tional levels (De Meza, Irlenbusch, and Rey- evidence that certain types of financial literacy niers 2008), and people with higher levels programs can improve financial knowledge of education perform better along a number and affect behavior. of dimensions, including budgeting, living Can people be effectively taught financial within means, attitudes toward the future, knowledge as well as the skills, attitudes, and, and impulse control (Kempson, Perotti, and ultimately, behavior that improve financial Scott 2013). outcomes? The diversity of existing studies limits the ability to draw conclusions.21 How- ever, with appropriate caveats, it is possible to Targeting people with lower levels of identify emerging lessons.22 In a study under- formal or financial education helps lying this report, Miller and others (2013) Financial literacy programs aimed at the gen- approach this question through a systematic eral population have not yet been shown to review of the impact evaluation literature on be effective, but programs focused on specific financial capability and construct a detailed segments of the population, especially those database that describes the range of interven- people with lower levels of formal or financial tions that have been studied and that permits education, can have substantial measurable a preliminary meta-analysis. The following effects. subsections provide a qualitative summary of The debate on expanding financial inclu- the recent evidence. sion has often focused on supply-side issues, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 81 such as the creation of financial services that incentives varied randomly from Rp 25,000 are appropriate for low-income consumers. to Rp 75,000 and Rp 125,000 (equivalent to Less attention has been paid to the demand $3, $8, and $14, respectively). side and to understanding why seemingly Financial literacy did not have a measur- attractive services, such as affordable basic able impact within the full sample, but larger bank accounts, have generated only limited monetary incentives did, raising the probabil- take-up. Using experimental evidence on ity of opening a bank account by 13.7 per- Indonesia, Cole, Sampson, and Zia (2011) centage points for the incentive of $14, more have evaluated the impact of a financial lit- than twice the impact of the $8 incentive (table eracy intervention and monetary incentives 2.2). In follow-up research carried out two on the decision to open a bank account. Only years after the initial study, Cole, Sampson, 41 percent of Indonesian households report and Zia (2011) found that households that they have a bank account, but 51 percent have had received the larger incentives continued to savings in a nonbank institution. When house- have a higher number of open accounts, and holds surveyed for the research were asked if approximately two-thirds of these households they would open a bank account if there were had used the accounts during the previous year. no fees, 58 percent responded that they would. On households without formal schooling, Additional evaluations indicated that financial financial literacy training had a significant literacy is a significant predictor of the use of effect and increased the probability of open- bank accounts, but less important than other ing an account. Also, the financial literacy factors such as household wealth. intervention had a greater impact among con- The researchers have tested the willingness sumers who had started out with low levels of of consumers to open a basic bank account financial knowledge, significantly increasing offered by Bank Rakyat Indonesia, the coun- their likelihood of opening an account (table try’s largest bank. This account required a 2.2). These results highlight the importance minimum deposit of Rp 5,000 ($0.53) and of targeting financial literacy efforts on more involved no fees for up to four transactions per disadvantaged consumers, as well as the rela- month. Deposits earned interest on balances tive importance of cost barriers, including the above Rp 10,000 ($1.06) and were covered possible role that incentives can play in boost- by the same deposit insurance used by other ing participation in financial markets. banks for deposits in Indonesia. Households were selected randomly and independently for Targeting the young can help, too a financial literacy intervention and a mon- etary incentive. The financial literacy course School-based interventions and programs tar- was a two-hour in-person classroom train- geting youth are among the most common ing session designed to teach unbanked indi- approaches to teaching financial literacy and viduals about bank accounts. The monetary capability. School-based programs benefit TABLE 2.2 Effects of Financial Literacy Interventions and Monetary Incentives, Indonesia Effect on the probability of opening a bank account, percentage points By fi nancial Indicator Full sample By education level literacy score Financial literacy training +2.9 −3.2 −4.9 Monetary incentive Rp 75,000 ($8) +6.6* +6.1** +6.0 Monetary incentive Rp 125,000 ($14) +13.7*** +9.9*** +10.0*** Financial literacy training and no formal schooling +15.5** Financial literacy training and below median financial literacy +10.0** Source: Based on Cole, Sampson, and Zia 2011. Note: Empty cells indicate the information was not included in the regression. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent 82 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 from access to a large captive population and was recently evaluated by Bruhn and others a skilled teaching staff (although educators are (2013). It considerably enhanced the finan- not necessarily comfortable teaching financial cial knowledge, attitudes, and behavior of concepts) and the ability to incorporate lessons students and is being rolled out on a national in a variety of ways to reach different kinds basis. In both follow-up evaluations, the stu- of students. The downside is that students dent test scores indicated that the average often have limited money or access to finan- level of financial proficiency was substantially cial products and services and thus cannot put higher in the treatment group than in the con- concepts to use. And, even if these programs trol group. The test scores of both low- and are effective over time, it is difficult to estab- high-achiever students rose across the distri- lish causality, and few researchers have tried bution. The distributional effects of the pro- to measure the long-run effects. While school- gram speak to the accessibility of the curricu- based financial education is a popular con- lum and highlight the benefits for students cept, and many countries have such initiatives, across a wide performance spectrum. Simi- few programs have been rigorously evaluated. larly, students in the treatment group boosted On the United States, the evidence on their savings and exhibited greatly improved financial education among youth seems spending behavior compared with the con- mixed. Bernheim, Garrett, and Maki (2001) trol group. In the treatment group, a higher show that personal finance courses in schools proportion of students saved at least some of are linked to higher rates of asset accumula- their income relative to the control group. The tion. However, Cole and Shastry (2010) find study shows that both the proportion of stu- no relationship between the two. Carpena dent savings and the actual amount of the sav- and others (2011) have extended this research ings per student increased (figure 2.7). and find weak evidence of an impact by sec- ondary-school financial education courses, Teachable moments have value but more promising results related to basic mathematics education on future financial Financial literacy programs aimed at poor habits, especially among girls. Research by households that do not have the leeway to the U.S. nonprofit Jump$tart also indicates change their savings behavior are bound to that courses on personal finance are not cor- produce small results. However, implement- related with improved performance in tests of ing such programs at critical life-decision financial literacy (Mandell 2008). However, points can be valuable. For example, a finan- more motivated students are more likely to cial literacy program in Indonesia for work- show a positive impact from such courses, ers who are about to migrate abroad (and indicating that student engagement with the obtain a major boost in income) has shown material may be a critical factor of success a relatively large impact on savings (Doi, (Mandell and Klein 2007). McKenzie, and Zia 2012). In Ghana, a study supported by Innova- The term “teachable moment” refers to tions for Poverty Action, a nonprofit, reported times when people may be especially moti- relatively limited results (Berry, Karlan, and vated to gain and use financial knowledge Pradhan 2013). While the financial education and skills and are able to put this knowledge program resulted in a small increase in youth to work. These moments include periods dur- savings rates and greater risk aversion, it had ing which specific financial decisions—such no measurable impact on the overall level of as the purchase of a financial product or ser- financial literacy, time preference, or financial vice or major life changes such as marriage, planning behavior of the youth involved. divorce, starting a job, retirement, the birth of A promising example of a comprehen- a child, or the death of a spouse—occur that sive financial education program is a three- may affect incomes or normal expenditures. semester secondary-school financial educa- Most impact studies focus on adult finan- tion scheme piloted in Brazil. The program cial education in developed-country markets GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 83 FIGURE 2.7 Effects of Secondary-School Financial Education, Brazil a. Students’ financial knowledge b. Students who save 65 50 49 62 46 Average score, scale 0–100 60 60 59 45 44 Students, % 56 40 55 40 50 35 Control Treatment Control Treatment Control Treatment Control Treatment Follow-up 1 Follow-up 2 Follow-up 1 Follow-up 2 Source: Bruhn and others 2013. Note: Students completed surveys that included a test of financial proficiency, with scores from 0 to 100. The intervention was launched in August 2010; the first follow-up survey was conducted in early December 2010; and the second follow-up survey was conducted in December 2011. (especially the United States) and are typically including access to captive populations at the linked to teachable moments. The teachable place of employment, which reduces the cost moments that are most often considered in of delivering the intervention, and incentive the literature on adult financial literacy are alignment between employers and employees retirement planning and participation in pen- (both benefit from employees making good sions (often provided by employers); training financial decisions), which may raise the cred- related to mortgage finance and home own- ibility of the information provided. ership; and credit or debt management, often targeted at people with credit problems. The Reinforcing messages through social ability to use immediately the new knowledge networks is helpful and skills in a particular financial decision or transaction also facilitates the impact analy- Several key studies have shown that it is sis of financial literacy interventions designed important to reinforce knowledge and behav- around teachable moments. Workplace finan- ior through social networks. Bruhn and cial education programs leverage teachable others (2013), in their above-mentioned moments that are common to a number of study of a secondary-school financial edu- employees, such as joining a pension fund or cation program in Brazil, have evaluated a retirement plan for new hires or deciding on the impact on student savings rates of the benefits such as insurance or stock options engagement of parents during a brief finan- when employer-provided programs change. cial workshop. Parents were randomly For example, Bayer, Bernheim, and Scholz assigned to various workshops, some focus- (2009) and Bernheim and Garrett (2003) pro- ing on health issues, others on the financial vide evidence of a positive impact of work- messages being taught to their children. The place financial education in the United States. savings rate among students increased by In both instances, the evidence is linked to 2.5 percentage points if the parents attended retirement savings rates. the financial workshop provided by the There are several commonsense reasons school. Thus, students whose parents had why workplace outreach may be effective, participated in health literacy workshops 84 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 reported they saved 13.5 percent of their to the enhancement of financial outcomes. income, compared with 16.0 percent among Small situational barriers, such as a “testy students whose parents had participated in bus ride, challenging hours, or the reluctance the financial literacy workshops. This dif- to face a contemptuous bank teller” (Ber- ference was statistically significant at the trand, Mullainathan, and Shafir 2004, 420), 5 percent level and indicates that parents can be a substantial hindrance to opening a were able to use their improved knowledge bank account despite the large benefit. In this to reinforce the financial messages their chil- context, alleviating several constraints simul- dren heard at school. taneously can provide a needed impetus for Research on financial literacy among change. migrant workers has found that both the Numerous studies have documented that desire to change savings behavior and the one-off interventions focusing only on one ability to put newly acquired knowledge constraint have little impact. For example, about financial issues into effect are much Gibson, McKenzie, and Zia (2012) have stud- greater if both the worker who sends a remit- ied the impact of teaching migrants in Aus- tance and the individual receiving the remit- tralia and New Zealand about various meth- tance have had financial literacy training. For ods and cost options for remitting money to example, Doi, McKenzie, and Zia (2012) their home countries and detected no change have assessed the impact of financial literacy in either the frequency or level of the remit- training among migrants who are remitting tances. Similarly, Seshan and Yang (2012) money and among the family members who conducted a financial literacy course on sav- are receiving the funds from the migrants. ings among migrants in Qatar and found no The training was delivered at the teachable significant impacts on savings or remittance moment when the migrant workers were levels. about to leave their home country for work In an innovative study, Carpena and oth- abroad. The training program emphasized ers (2013) tried to examine empirically the financial planning and management, sav- effect of broader interventions. They stud- ings, debt management, sending and receiv- ied the impact of the use of DVDs on sav- ing remittances, and understanding migrant ings, credit, insurance, and budgeting on insurance. The authors conclude that train- financial education in India. The DVDs were ing among both the migrants and the family produced specifically for the study. The pro- members has large and significant impacts on duction process incorporated several rounds financial knowledge, behavior (for instance, of feedback from focus groups to adapt the more careful financial planning and budget- content to a target audience in urban areas. ing), and savings. If only the family member The DVDs were shown over several weeks (the recipient of the remittances) is trained, to a randomly selected treatment group, and the effect is still positive, but smaller. Train- the outcomes were compared with the results ing among only the migrants had no effect on among a control group that watched DVDs the family members. Hence, there are comple- on health literacy. Despite the relatively long mentarities in treatment. intervention, adapted content, and attractive delivery, the influence on financial outcomes was limited. While the study found improve- Complementary interventions matter ments in attitudes toward and awareness Combining targeted financial literacy pro- of financial issues, budgeting and borrow- grams with other interventions is particularly ing were the only areas in which significant helpful among poor households, which often changes in behavior were detected. The link face multiple constraints on saving and other between financial knowledge and behavioral financial activities. Even if financial educa- outcomes thus appears rather weak in this tion can lead to better knowledge, individual case as well, in line with the studies on one- behavioral constraints are a significant barrier off interventions. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 85 To test the effectiveness of complemen- one-on-one counseling, are relatively expen- tary interventions, that is, interventions that sive because of the higher per unit costs. address multiple constraints at once, Carpena Technology-enabled solutions can help reduce and others (2013) combined their financial costs and target information and tools to spe- literacy treatment with goal setting and indi- cific consumer needs. vidualized financial counseling. The results “Entertainment education” or “edutain- of these integrated treatments were striking. ment” interventions can potentially reach First, goal setting and financial counseling large audiences, including individuals who have consequential effects on the take-up of may need to strengthen their financial capabil- financial products. While financial educa- ity, but who would not seek out or attend spe- tion alone does not induce individuals to cial training sessions. There is a growing lit- undertake informal or formal savings initia- erature on the use of entertainment education tives, the incorporation of goal setting and to impart financial literacy. Spader and others financial counseling in the financial educa- (2009) find that the awareness of key finan- tion program does accomplish this. Similar cial concepts increased among viewers of a results occurred with regard to opening bank television soap opera, “Nuestro Barrio,” pro- accounts and purchasing financial products duced for the U.S. Latino population. Tufano, such as insurance. Flacke, and Maynard (2010) have evalu- Second, the integrated treatments allowed ated the impact of video games produced by the respondents to plan their finances more Doorways to Dreams on a sample of 84 par- carefully; thus, respondents assigned to the ticipants and found a boost in financial skills, integrated treatment group were less likely to knowledge, and self-confidence. Di Maro and borrow for consumption purposes and more others (2013) have assessed the impact of the likely to understand the details of the interest use of a movie to encourage savings among rates associated with their loans. small entrepreneurs in Nigeria and find that, These results suggest that coupling finan- while viewing the film was helpful in motivat- cial literacy with individualized financial ing short-term behavioral change (opening counseling and a scheme of reminders can a savings account), it did not lead to longer- improve savings behavior. Hence, thinking term increases in savings. Berg and Zia (2013) more broadly about the constraints that indi- have analyzed the impact of including a finan- viduals and households face in poor settings cial literacy story line on debt management is crucial. into an ongoing commercial soap opera in South Africa (box 2.5). The delivery mode matters CONSUMER PROTECTION AND Unlike children, adults are often stubborn in MARKET CONDUCT their preferences, which are therefore difficult to change. Also, many people have a short Financial inclusion creates opportunities for attention span and become easily distracted if many consumers who have previously been they find lecture-based financial literacy pro- excluded from financial markets or who have grams boring, even if the programs are con- been underbanked. However, it is also accom- ducted by leading experts. Thus, innovating panied by risks. The emphasis on respon- the financial knowledge delivery mechanism sible financial inclusion in recent years is an matters. attempt to balance opportunity and innova- There are many ways that financial lit- tion in financial markets with safeguards to eracy and capability programs can be deliv- prevent abuse and to help consumers benefit ered, and the implications in terms of both from access, especially the most vulnerable. cost and effectiveness vary. Traditional forms Consumer protection and market conduct of outreach, such as classroom instruction, regulations are a key aspect of a respon- brochures, and other printed materials and sible finance agenda. The terms consumer 86 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.5 Behavior Change through Mass Media: A South Africa Example Household indebtedness has been a large and grow- episodes a week for eight years (more than 1,500 epi- ing problem in South Africa over the last decade. The sodes). It is broadcast on the second-most-watched ratio of household debt to disposable income was station in South Africa—the show reaches approxi- 76 percent in the second quarter of 2012, up from mately 3 million viewers—and has been especially 50 percent in 2002. a In June 2012, the National popular among low-income South Africans. The Credit Regulator reported that, of the roughly financial education story line stretched over two con- 20 million consumer borrowers on its books, 9 mil- secutive months, a time span deemed necessary by lion (47 percent) had poor credit records.b social marketing specialists for the viewers to con- The fi nancial woes in South Africa are not lim- nect emotionally with the characters, for the events to ited to indebtedness. The household savings rate is unfold, and for the financial literacy messages to sink low (1.7 percent in the second quarter of 2012). c in. The financial education messages in the story line According to FinScope, 67 percent of the popula- were tested through focus groups to ensure that they tion does not save at all, even though a majority of would be correctly understood by the target group, adults believe that saving is important. Moreover, of that the plot appeared realistic, and that the target those who save, only 22 percent save through formal group could identify with the characters and their channels.d problems. All changes suggested by the focus group A project was undertaken to assess the ability of members were included in the fi nal story line. The entertainment education to influence sound finan- development and production of the story line lasted cial management among individuals, with a focus for eight months and cost approximately $90,000. on managing debt. The project entailed the content To evaluate the impact of the project, a random- development, production, and impact evaluation of ized encouragement design methodology with the fi nancial capability story lines that were included in following key components was used. A financial a popular South African soap opera, “Scandal!” (fig- incentive of R 60 (about $7) was provided to half of ure B2.5.1). The soap opera has been running four the sample (the randomly selected treatment group) B2.5.1 Scandal! Cast Source: Ochre Media and e.tv. Used with permission; further permission required for reuse. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 87 BOX 2.5 Behavior Change through Mass Media: A South Africa Example (continued) to encourage the participants to watch “Scandal!” after an episode in which the association was intro- A fi nancial incentive of the same amount was pro- duced into the story line. The call volume jumped vided to the other half of the sample (the randomly from an average of 120 calls per day to over 500 per selected control group) to watch a soap opera that day, a more than 300 percent rise. A regression anal- is similar in terms of content and viewership profi le ysis confi rmed that there had been a large increase and that airs around the same time, but that did not in awareness of the existence of formal avenues have a financial literacy component. Two phone- for fi nancial advice: compared with an average of based surveys were conducted. (The fi nancial incen- 69 percent in the control group, nearly 80 percent tives were transferred in the form of airtime credits.) of the respondents in the treatment group stated A fi nal face-to-face survey probed the longer-term that they would seek fi nancial advice from a formal effects of the soap opera four months after the fi nan- source if they needed it. However, this effect dissi- cial literacy episodes had been broadcast. pated over the longer run, likely because the relevant T he i mpac t eva luat ion fou nd subst a nt ia l content only appeared in the soap opera for a brief improvements in content-specifi c fi nancial knowl- period and because reinforcement of the messages edge, a greater affi nity for formal borrowing, less affi nity for hire purchase agreements (which spread the cost of an expensive item over a period of time at FIGURE B2.5.2 Effects of Entertainment Education a high interest rate), and a decline in gambling (Berg 31 and Zia 2013). The possible negative consequences 30 26 of hire purchase and gambling were highlighted in Respondents, % 25 19 the soap opera story line: the main character ended 20 15 up in considerable fi nancial distress because of a hire 10 purchase transaction and gambling. 5 The results of the fi nal face-to-face survey (fi g- 0 ure B2.5.2) show that respondents in the treatment Has someone in the Has someone in the group were substantially less likely to have signed a household used hire purchase household gambled money hire purchase agreement or to have gambled during in the past 6 months? in the past 6 months? the previous six months (respectively, 19 percent and Treatment Control 31 percent in the control group vs. 15 percent and Source: Berg and Zia 2013. 26 percent in the treatment group). The heteroge- Note: “Hire purchase” refers to contracts whereby people pay for goods in neous effects were strongest among the respondents installments. with low initial financial literacy and low formal educational attainment. The complementary quali- through other interventions are needed to ensure tative analysis confi rmed these fi ndings and high- greater knowledge retention. lighted some key gender differences in the way men The study shows that entertainment media has and women think about borrowing: women gener- the power to capture the attention of individuals and ally take on debt as a last resort, while men are more thereby provide policy makers with an effective and willing to borrow for purchases. accessible vehicle to deliver carefully designed edu- In addition, daily call volume data were obtained cational messages that resonate with audiences and from the National Debt Mediation Association that influence fi nancial knowledge and behavior, at least showed an upsurge in incoming calls immediately in the short to medium term. a. Quarterly Bulletin 265 (September 2012), South African Reserve Bank. b. Credit Bureau Monitor (June 2012), National Credit Regulator. c. Quarterly Bulletin 265 (September 2012), South African Reserve Bank. d. FinScope (database), FinMark Trust, Randjespark, South Africa, http://www.fi nscope.co.za/. 88 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 protection and market conduct are sometimes marketplace. The greater transparency cre- used interchangeably, but, by consumer pro- ated in financial markets by consumer protec- tection, we mean a subset of market conduct. tion regulations can help shift demand away Transparency and disclosure, fair treatment, from products that offer poor value to con- and effective recourse mechanisms are the sumers and toward higher-quality products, three main components of consumer protec- resulting in more sustainable market growth. tion laws and regulations.24 Market conduct Proportionality is a widely accepted princi- includes these three aspects of consumer pro- ple of good regulation. It helps ensure that pro- tection, as well as broader regulatory respon- tection does not unnecessarily restrict access. sibilities related to entry and competition, According to this principle, the restrictions such as licensing institutions and market anal- imposed by regulators on industry must be ysis to identify instances of excessive market proportionate to the benefits that are expected power, new risks, or fraudulent behavior. to result from the restrictions (Lyman, Pick- Consumer protection is particularly useful ens, and Porteous 2008). An example that is if there are gaps in financial literacy and finan- relevant for financial inclusion is customer cial capability, especially as policies aimed at identification and documentation require- enhancing financial inclusion help open new ments, which can effectively limit the access market segments and as financial institutions to finance by low-income populations. Rather introduce new distribution channels. Legal than viewing the relaxation of these types of and regulatory protections can extend across requirements as a threat to financial stability the product life cycle by influencing the way and integrity, regulators today see financial products are designed and marketed to con- inclusion as part of an effective response to sumers, promoting the disclosure of fees and financial threats such as money laundering and interest when products are purchased, and terrorist financing. The Financial Action Task specifying mechanisms for redress and appro- Force, which aims to protect the international priate collection procedures if problems arise. financial system from misuse, recently affirmed the value of financial inclusion by declaring that financial exclusion can represent a risk by Proportionality, resources, and the boosting the use and prevalence of informal effectiveness of regulation providers who are not monitored.25 Consumer protection regulations should be On the credit side, well-designed gov- balanced with other goals of financial sec- ernment regulations can assist in ensuring tor policy. The G-20’s Global Partnership for that loans flow to creditworthy households, Financial Inclusion has defined four distinct thereby helping protect the stability of the financial policy goals—inclusion, stability, financial sector and avoiding the negative integrity, and protection—and highlighted the consequences of consumer overindebtedness. need for regulators to understand the effects Meanwhile, governments are sometimes sub- of changes in one area on the other three. The ject to incentives (political pressure) to expand links between consumer protection and the the access to credit, potentially even beyond stability and integrity of the financial system optimal levels. On the payment side, govern- are straightforward: measures to boost the ments can play a crucial role in retail payment ability of consumers to make informed deci- systems by addressing the potential market sions about financial products and services failures arising from coordination problems. and to seek redress if problems arise have a Streamlining these systems and facilitating direct bearing on market performance. The their interoperability can improve their effi- relationship with financial inclusion is more ciency and affordability. complex but no less critical. Consumer pro- Enhancing the transparency of financial tection can raise the confidence of consumers markets through disclosure requirements in financial products and services, increas- that make information available and easy to ing the willingness of consumers to enter the comprehend and use is another key aspect of GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 89 consumer protection laws and regulations. protection laws and regulations, including the In comparing products and services, even use of mystery shoppers and other tools that the most financially savvy consumers ben- provide accurate, unfiltered information on efit from standards on disclosure. Improve- how financial products and services are mar- ments in disclosure standards have been, for keted and sold. example, a major element of the recent U.S The inadequacy of existing disclosure financial regulatory reforms (Agarwal and regimes has been highlighted through World others 2009; Barr, Mullainathan, and Shafir Bank consumer protection diagnostic reviews 2008). Among other reforms, the 2009 Card (World Bank 2009a, 2009b, 2010). These Act in the United States requires that impor- reviews have found that at least half of the tant billing information be in plain language complaints submitted to supervisory authori- and plain sight. In addition to information ties are requests for additional background on the minimum payment, bills must include information to help consumers understand the the number of months needed to pay off the financial services they purchase. The poor con- debt if the consumer pays only the minimum ditions in low-access environments, including and provide the amount that would need to low levels of literacy and numeracy, regulatory be paid to retire the debt within 36 months. capacity constraints, the existence of unregu- A recent study shows that there have been lated informal providers, and overreliance on improvements because of these disclosure advice from loan officers, all pose obstacles to requirements, including higher minimum pay- the effective application of disclosure require- ments and, in some cases, payment amounts ments (Chien 2012). revised to approach the 36-month payoff Efficient consumer protection rules should amount. However, these gains have typically also cover the other communication channels not been retained, and people who pay off financial institutions use to present informa- higher debt amounts still tend to raise their tion to consumers beyond the disclosure pro- overall debt levels. vided at the point of sale. Along with informa- Consumer protection policies to address tion from family and friends, many consumers information failures and improve disclosures depend on advertising or comparisons in the are at least as challenging to implement in media (that may or may not be independent) low-access, low-income environments. Suc- as their primary sources of information about cessful implementation depends on the skill financial products. The reliance on advertis- mix of consumers and the local market struc- ing is even more pronounced in rural settings, ture (Beshears and others 2009; Gu and Wen- where consumers usually have less experi- zel 2011; Inderst and Ottaviani 2012). An ence with financial services and are thus more ongoing study on Mexico shows that loan susceptible to misleading information. For officers voluntarily provide little information example, World Bank studies in Azerbaijan to low-income clients, and, if probed, most and Romania find that more than one-third appear to be misinformed about key char- of rural consumers—34 percent and 38 per- acteristics of the products they offer (Giné, cent, respectively—rely on advertising for Martínez de Cuellar, and Mazer 2013). More their financial product information compared importantly, clients are never offered the with only 25 percent and 17 percent, respec- cheapest product that fits their needs, most tively, in urban areas. likely because institutions make more profit Once consumers have made a decision to by supplying more expensive products. This purchase a financial product or service, there illustrates the challenge of disclosure policies are numerous business practice issues that that run counter to the commercial inter- come into play. Key topics include ethical est of financial institutions (box 2.6). The staff behavior, selling practices, and the treat- Mexican study also highlights the potential ment of client data, as well as issues related value of regulators who are proactive in their to product design and marketing (Brix and approach to the supervision of consumer McKee 2010). For example, in various ways, 90 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.6 Case Study: New Financial Disclosure Requirements in Mexico The traditional approach to financial consumer staff. Shoppers had various planned scripts on the protection regulation has focused on providing kinds of accounts they were seeking (checking or information to consumers and assuming that ratio- fi xed term savings), their level of fi nancial literacy nal decisions will follow. Insights from behavioral or knowledge (neophytes or experienced), and their economics, however, indicate multiple reasons for awareness of the competition and other offers. departures from this classical model, including Giné, Martínez de Cuellar, and Mazer (2013) ana- present bias, loss aversion, difficulties dealing with lyzed the quality of the information obtained by the ambiguous information or too much information, shoppers along three dimensions: the conditions and and status quo bias. terms of the accounts, such as interest rates and type New research and practical examples from a vari- of interest; the fees and commissions, such as those ety of countries indicate ways that consumer protec- related to minimum balances and early withdraw- tion regulations and recourse mechanisms can be als; and account usage, including procedures for bal- crafted to be more effective by taking into account ance inquiries and money withdrawal. Staff provided how people actually react to disclosures on con- verbal information voluntarily (without prompting) tracts, information provided by lenders, and infor- on only one-third (6 of 18) of the information items mation on recourse mechanisms. being surveyed. The “experienced” shoppers were In Mexico in 2009, new requirements were provided more information on fees and on the usage approved on disclosure formats and pricing poli- of the accounts and were also more likely than the cies through the Law for Transparency and Regula- neophytes to be shown a loan contract. The ganan- tion of Financial Services. One part of the regula- cia anual total (total annual return), a measure of tion requires that consumers be presented with key annual interest that can be compared across insti- fi nancial terms. However, disclosure-related prob- tutions because it is specified by formula, was only lems persist in the Mexican financial market. To provided at the request of the experienced shoppers, understand how the regulations are being applied, and only in two cases were staff of fi nancial institu- researchers working with the National Commission tions able to explain this precisely. Printed materials, for the Protection of Users of Financial Services, the if they were supplied at all, offered little additional government agency charged with fi nancial consumer information that could help in evaluating products. protection, sent trained mystery shoppers to 19 The shoppers typically received product offers that fi nancial institutions in four towns near Mexico City were more expensive than products more suited to with predominantly low- to middle-income popula- their needs, reflecting the misalignment of incentives tions of between 30,000 and 50,000 people. The that exists between a fi nancial institution and its cus- mystery shoppers carried out 112 visits to fi nancial tomers. The research demonstrates the difficulty, in institutions, lasting, on average, 25 minutes, includ- practice, of regulating disclosure, transparency, and ing about 15 minutes of face-to-face encounters with the provision of information by fi nancial institutions. financial firms take advantage of informa- fair treatment in business practices can use tion asymmetries and limits on the ability incentives through regulation and by encour- of consumers to understand and use infor- aging competition and market transparency mation, including through pricing strategies (to identify good actors), as well as coer- and by highlighting irrelevant information cive tactics such as the regulation of specific or by developing redundant financial “inno- products and practices. However, enforc- vations” that target less-sophisticated con- ing fair treatment is complicated because, sumers (Carlin 2009; Choi, Laibson, and in low-access environments, the lack of for- Madrian 2010; Gabaix and Laibson 2005; mal financial options may lead customers to Henderson and Pearson 2011; Johnson and accept abusive or coercive treatment in the Kwak 2012). Policy makers concerned with formal or informal sector. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 91 The rules on the books vs. enforcement FIGURE 2.8 Consumer Protection Regulations and Enforcement Actions There is an important gap between enforce- ment and the consumer protection laws and 5 350 regulations on the books. According to the World Bank’s recent Bank Regulation and 300 Supervision Survey (Čihák and others 2012), 4 250 Actions taken per year many countries have consumer protection Instruments available regulations on the books, but only a small 3 200 fraction of these countries actually enforce the regulations (figure 2.8). In about two-thirds of 2 150 countries, banking regulators are responsible 100 for consumer protection. The share is higher 1 in Central and Eastern Europe (77 percent) 50 and lower in East Asia (63 percent) and Sub- 0 0 Saharan Africa (61 percent). Regulators report Low income Lower-middle Upper-middle High income they have a number of avenues of enforce- income income ment; the most common is the ability to issue Actions taken per year (right axis) a warning (64 percent of responding coun- Instruments available (left axis) tries), followed by the ability to impose fines and penalties (55 percent). The least common means of enforcement is the most drastic one, Sources: Bank Regulation and Supervision Survey (database), World Bank, Washington, DC, http:// go.worldbank.org/WFIEF81AP0; C ˇ ihák and others 2013. withdrawing the license of the provider. This Note: The numbers are per country averages for the respective country groups. step is available in 34 percent of countries in the sample. Most countries, with the excep- tion of countries in the Middle East and North a lack of adequate resources. In Brazil, for Africa, report they have several enforcement example, the Consumer Protection Secretary tools at their disposal; the median number of in the Ministry of Justice has responsibility tools is four. for all economic sectors. Most complaints and Enforcement actions (the orange columns issues originate in only a few key industries. in figure 2.8) are much more frequently taken In 2010, there were 26,000 complaints related in high-income countries than in other catego- to finance, representing 21.5 percent of the ries of countries. This may, to some extent, total (second place). Even so, no full-time reflect the bigger size of high-income country staff member is charged with dealing with financial systems, but analysis indicates that, financial complaints, and fewer than 30 full- even if the number of enforcement actions is time staff cover consumer protection issues scaled by banking system assets in each coun- nationwide in a country of 200 million.26 try, major differences persist, and the ratio Only about two-thirds of the regulators who of assets per enforcement action is signifi- are responsible for financial consumer protec- cantly higher in high-income countries than tion in countries have a dedicated team or unit in low-income countries. The distribution of working in this area (Čihák and others 2012). enforcement actions is also quite uneven in Some countries also have several separate sec- this sample: six countries account for about toral financial supervisors with varying con- 85 percent of the total number of enforce- sumer protection mandates, which may lead ment actions. The strongest sanction—revok- to uneven consumer protection, uneven atten- ing a license to operate—was available to tion to consumer issues, and regulatory arbi- 34 percent of the countries, but was only used in trage by financial groups seeking to limit the 4 percent of the countries in 2006–11. impact of consumer protection rules. Several factors may inhibit the effective- In countries where an agency with broad ness of consumer protection regimes. One is responsibility for consumer protection handles 92 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 consumer protection for finance, there may through the enforcement of consumer protec- also be problems because of a lack of knowl- tion regulations can be useful for prudential edge of financial sector issues in the agency. In supervision, and vice versa. Prudential regula- the case of Brazil, for example, the Consumer tors also tend to be well funded and attract Protection Secretary is supported at the local top-quality staff, providing a strong institu- level by consumer protection bureaus known tional framework for consumer protection as procons, which provide services directly to and market conduct activities. However, there citizens across all product and service catego- is also a potential downside to this approach. ries and through a variety of means, includ- This may include inefficiencies and disecono- ing by phone, in person at an office, and by mies of scale because of a lack of harmony mobile procon vans. While these bureaus are and poor task delegation across banking, important in making consumer protection insurance, and securities and the competition accessible to more Brazilians, they may lack of resources between the two regulatory func- the specific knowledge of finance necessary tions, which could result in circumstances in for effective action. On several key consumer which regulators choose to maintain financial protection issues, leading procons have taken sector stability at the expense of consumer positions against financial regulators and protection. arguably against the interests of consumers. For these reasons, some countries have For example, procons have criticized the law created a specific regulator for financial con- creating the positive data archive in the credit sumer protection and market conduct issues. reporting system (the cadastro positivo) on the Another reason for separating prudential and basis of concerns over data protection that are consumer protection regulation relates to valid, but have not balanced this by discussing the wide range of providers offering finan- the potential gains through the cadastro from cial products and services, many of which more transparent and competitive credit mar- are nonbank financial institutions. This kind kets. In the case of the payment card industry, of twin peaks approach is considered a way procons defend the right to claim nao sobre to limit regulatory gaps and ensure competi- preco (often translated as the law of one tive neutrality (Čihák and Podpiera 2008). price), arguing that consumers should not be Because of the relative stability of banking subject to higher costs for goods purchased systems in countries such as Australia and with credit cards instead of cash. However, Canada during the global financial crisis, this is a position that strengthens the hand of interest in the twin peaks model of supervi- card acquirers in their dealings with retailers sion has been growing in recent years. In the and may contribute to higher prices, which is United States, a new supervisory structure, not in the interest of consumers (Central Bank anchored by the Consumer Financial Protec- of Brazil, Ministry of Finance, and Ministry of tion Bureau, has been established to focus on Justice 2010). consumer protection and market conduct. In the United Kingdom, a similar change has been undertaken through the creation of the The organization of consumer protection Financial Conduct Authority. In some countries, including Malawi, Por- An alternative approach involves the tugal, and, until recently, the United States, establishment of a general consumer protec- financial regulators are responsible for pru- tion agency responsible for financial con- dential supervision, consumer protection, sumer protection. In the United States, for and market conduct. In countries with lim- example, prior to the creation of the Con- ited resources and institutional capacity and sumer Financial Protection Bureau, the Fed- a lack of professional staff able to perform eral Trade Commission handled some aspects financial market supervision, combining these of financial consumer protection, including two functions may be the best solution. There complaints related to credit reporting. The are also advantages to combined supervi- advantages of this approach include regula- sion given that the market intelligence gained tory independence because officials are not GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 93 FIGURE 2.9 Evolution of Business Conduct, by Financial Market Depth a. All economies b. High financial depth economies c. Low financial depth economies 60 70 45 40 50 60 35 50 Share of countries, % Share of countries, % Share of countries, % 40 30 40 25 30 30 20 20 15 20 10 10 10 5 0 0 0 20 9 00 01 20 2 20 3 20 4 05 06 07 20 8 20 9 10 20 9 00 01 20 2 20 3 20 4 05 06 07 20 8 20 9 10 20 9 00 01 20 2 20 3 20 4 05 06 07 20 8 20 9 10 9 0 0 0 0 0 9 0 0 0 0 0 9 0 0 0 0 0 19 20 20 20 20 20 19 20 20 20 20 20 19 20 20 20 20 20 Integrated business conduct Business conduct in place Twin peaks Source: Melecký and Podpiera 2012. working only with financial markets and have that can lead to inefficient outcomes. Asym- the ability to examine problems that involve metric information (which leads to adverse finance beyond supervised institutions in retail selection and moral hazard problems) and lending. However, there are often also poten- high transaction costs are two significant tial drawbacks, including a lack of resources obstacles that are commonly mentioned. and high-quality staff, a lack of knowledge Asymmetric information and high transac- of financial products and services, and lim- tion costs can generate first-mover dilemmas ited ability to influence powerful stakeholders and coordination problems that prevent the such as banks and other financial providers. expansion of financial services to certain seg- Whatever the specific institutional struc- ments of the population (de la Torre, Gozzi, ture, formal market conduct supervision has and Schmukler 2007). For example, a bank been catching on in both developed and devel- investing in a technology or business model oping economies. According to Melecký and that can reach underserved customers has to Podpiera (2012), over half of the 98 countries bear the risks and the initial costs if the exper- surveyed have employed some form of for- iment fails, but can quickly lose market share mal market conduct supervision (figure 2.9). if others follow its lead. Problems of this sort And, while specific mandates may differ, most can promote underinvestment in innovations countries share the same overall outcome that can mitigate asymmetric information and goals for the operations of the market conduct high transaction costs. At the same time, the supervisors: (1) fair treatment of consumers, departures of consumers from rational behav- (2) enhanced and informed participation in ior may also generate inefficient outcomes the financial system, and (3) sustained public and justify government action. confidence and trust in the financial system. The following subsections focus on three roles that the government can play in financial inclusion to mitigate the problems GOVERNMENT POLICIES FOR associated with asymmetric information, FINANCIAL INCLUSION high transaction costs, and consumer irra- A key role for the government in terms of tionality: (1) develop the legal and regula- financial inclusion is to deal with obstacles tory framework, (2) support the informa- in the supply or demand of financial services tion environment, and (3) subsidize access 94 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 or undertake other direct policies to expand for legal rights have deeper housing finance financial inclusion. systems. The efficiency of the legal system also mat- ters because this can affect the sectoral compo- Developing the legal and regulatory sition of lending. For instance, Costa and de framework Mello (2006) find that, in Brazil, banks pro- Legal institutions underpin the development vided payroll loans—the repayment of which of the financial sector (Djankov, McLiesh, was deducted from the employee’s payroll and Shleifer 2007; La Porta and others check—at lower rates than regular consumer 1997, 1998; Levine 1998, 1999). In particu- loans, which were subject to the inefficient lar, the protection of private property and procedures of the Brazilian legal system. Using the enforcement of shareholder and creditor databank-level survey data for over 20 tran- rights are cornerstones of developed financial sition economies, Haselmann and Wachtel sectors (for example, see Levine 1998, 1999). (2010) find that, if bankers have positive per- In environments with weak legal institutions, ceptions of the legal environment, they tend to contract writing and enforcement are prob- lend more to opaque borrowers such as house- lematic. As a result, financial institutions tend holds and small and medium enterprises. to resort to more costly business models (such The strength and enforcement of creditor as relationship lending or group monitor- rights can have implications for household ing) that might limit both the supply and the debt repayment behavior. Using data from demand for their services. the European Community Household Panel The government has a key part in enhanc- during 1994–2001, Duygan-Bump and Grant ing financial inclusion by introducing laws (2009) find that, if faced with adverse shocks, that protect property and creditor rights and households in countries with poor protection by making sure that these laws are adequately of creditor rights are more likely to delay their enforced. A number of studies find that both loan repayments. Hence, poorly designed and the quality of the laws and the efficiency of enforced creditor rights discourage lending enforcement of creditor rights affect the avail- and encourage households to default. ability and cost of credit to households.27 A key component of a modern collat- Meador (1982) finds that interest rates in eral framework is the existence of collateral the U.S. mortgage market are higher in those registries. The extensive research on the use states in which the cost and duration of judi- of property rights, land titles, and access to cial interventions to repossess collateral are finance suggests that property ownership is greater. Focusing on Europe, Freixas (1991) important in the attitudes, beliefs, and behav- shows that the cost and the duration of the iors that can have a considerable impact on a judicial process required to repossess collater- variety of social, income, and even environ- alized assets are inversely related to consumer mental factors (Di Tella, Galiani, and Schar- and home lending. Combining data on Ital- grodsky 2007; Goldstein and Udry 2008; ian households and the performance of Italian Jacoby, Li, and Rozelle 2002). However, the judicial districts, Fabbri and Padula (2004) traditional view of the importance of prop- find that an increment in the backlog of pend- erty rights in access to finance arising because ing trials has a statistically and economically property provides solid collateral is being significant positive effect on the probability challenged by new findings. For example, of loan rejections among households. Japelli, Deininger and Goyal (2012) evaluate the Pagano, and Bianco (2005) find similar evi- impact of modernizing land title registries in dence using aggregate credit data across 95 Andhra Pradesh, India, and find significant, Italian provinces. Using data on mortgage but modest rises in access to credit only in debt outstanding in 62 countries during urban areas. Galiani and Schargrodsky (2010) 2001–05, Warnock and Warnock (2008) study the impact of the acquisition of clear find that countries with stronger protections land titles on access to finance and on other GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 95 measures of well-being in a poor suburban The government can enhance financial area of Buenos Aires. They find that house- inclusion by facilitating the access of banks hold welfare improves, but not through the to borrower information either by passing credit channel; rather, long-term investments laws and regulations that enable banks to in housing and human capital formation made share information or by directly setting up without access to formal credit account for the public credit registries. These are databases changes. Galiani and Schargrodsky (2011a) established and managed by central banks or discuss the difficulty that low-income home- financial supervisors that capture informa- owners face in maintaining formal land titles tion on both individual and firm borrowers over time because of the relatively high cost and their credit. Private credit bureaus refer, of the administrative procedures. Galiani and meanwhile, to information-sharing arrange- Schargrodsky (2011b) find evidence of links ments spontaneously created and maintained between long-term investments and stronger by private financial institutions. property rights in rural areas, but none of the The current evidence on the relation- links are attributable to the use of land for col- ship between information sharing and credit lateral in formal credit markets. These stud- market performance relies mostly on cross- ies point to the importance of strong property country comparisons using aggregate or firm- rights, but call into question policies focused level data. Jappelli and Pagano (2002) and narrowly on providing land titles. Djankov, McLiesh, and Shleifer (2007) show that aggregate bank credit to the private sec- tor is more substantial in countries in which Developing the information environment information sharing is more well developed Information asymmetries between people (figure 2.10). Analyses of firm-level survey who demand and people who supply finan- data indicate that access to bank credit is cial services can lead to adverse selection and easier in countries in which there are credit moral hazard. For example, in credit mar- bureaus or registries (Brown, Jappelli, and kets, adverse selection arises when informa- Pagano 2009; Galindo and Miller 2001; Love tion about the borrower’s characteristics is unknown to the lender. Moral hazard refers to a situation whereby the lender’s inability FIGURE 2.10 Credit Information Sharing and Per Capita Income to observe a borrower’s actions that affect repayment might lead to opportunistic behav- ior on the part of the borrower. In both cases, 5.0 asymmetric information leads to rationing (a situation where the supply falls short of the 4.5 demand). Log, GDP per capita, US$ 4.0 y = 1.8096x + 3.2741 Theoretical models suggest that informa- R 2 = 0.1765 tion sharing can reduce adverse selection in 3.5 markets in which borrowers approach dif- ferent lenders sequentially (Pagano and Jap- 3.0 pelli 1993). Moreover, information sharing can also have a strong disciplining effect on 2.5 borrowers (Padilla and Pagano 2000). The model of Diamond (1984) indicates infor- 2.0 mation sharing can motivate borrowers to 0 0.2 0.4 0.6 0.8 choose agreed projects. Other models show Coverage of credit reporting system, % of population that information sharing can discipline bor- rowers into exerting substantial effort in Sources: Global Financial Development Report (database), World Bank, Washington, DC, http://www.worldbank.org/financialdevelopment; International Financial Statistics (database), projects and in repaying loans (Klein 1992; International Monetary Fund, Washington, DC, http://elibrary-data.imf.org/FindDataReports Padilla and Pagano 2000; Vercammen 1995). .aspx?d=33061&e=169393. 96 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 and Mylenko 2003). Though fewer stud- Subsidies and debt relief programs ies exist that focus specifically on the impact As in other areas of development, the use of of credit registries on consumer lending, the public funds is easy to justify in the interest of evidence is generally positive. For example, improving access and thereby promoting pro- using a unique data set on credit card appli- poor growth. Such subsidies should be evalu- cations and decisions from a leading bank in ated against the many alternative uses of the China, Cheng and Degryse (2010) analyze donor funds or scarce public funds, not least how the introduction of information shar- of which are alternative subsidies to meet the ing via a public credit registry affects bank education, health, and other priority needs lending decisions. They find that borrowers among the poor. In practice, such a cost- on whom there was extra information (those benefit calculation is rarely made. Indeed, the on whom information was shared with the scale of subsidies is often unmeasured. bank by other institutions) received a more Furthermore, as with financial sector taxa- substantial line of credit on their credit cards tion, subsidies can be more liable to dead- than those borrowers on whom the informa- weight costs in finance than in many other tion came only from the bank. sectors (Honohan 2003). It is often especially Because participation in public credit reg- difficult to ensure that finance-related subsi- istries is typically mandatory, they can jump- dies reach the target group or have the hoped- start credit reporting in countries with no for effect. private credit bureaus (Jappelli and Pagano An even more serious problem is the pos- 2002). However, to be effective, public credit sible chilling effect of subsidies on the com- registries need to provide timely and sufficient mercial provision of competing and poten- data on borrowers and their creditworthiness tially better services for the poor. Subsidizing (Maddedu 2010). Basic borrower informa- finance is likely to undermine the motivation tion (such as full name, unique identification and incentive for market-driven financial number, and location) is necessary, along with firms to innovate and deliver. It is this dan- the corresponding information on the credit ger—that subsidies will inhibit the viability of extended (such as credit type, outstanding sustainable financial innovation—that can be amount, days past due, and the date of origi- the decisive argument against some forms of nation) and on any risk mitigation measures subsidy. securing the credit (such as collateral and Note that it is not subsidization of the poor guarantors). The coverage of the public credit that should be questioned: the poor need help registry should be comprehensive—it should and subsidies in many dimensions. Subsidies include not only large, but also small debts— to cover fixed costs (for example, in payment and include as many financial intermediaries systems, especially if these generate network as possible; in other words, it should include externalities) may be less subject to this chill- not only banks, but also MFIs, cooperatives, ing effect than subsidies that operate to cover and so on. Ultimately, the extent, accuracy, marginal costs. Each case must be assessed on and availability of the information collected the merits. by the government will determine the useful- Microfinance is the area of financial access ness of the public credit registry as part of the in which subsidies have been most highly toolkit to expand access to credit. debated. Many well-intentioned people have While public credit registries can be impor- sought to make credit affordable for the poor tant in the early stages of financial develop- by means of subsidies. As a result, a major- ment, they can also reduce the attractiveness ity of MFIs today—though fewer of the larg- of private bureaus. In countries in which the est ones—operate on a subsidized basis (Cull, government decides to retain a public credit Demirgüç-Kunt, and Morduch 2009). Some of registry, ensuring fair competition should be a these subsidies are for overhead, and the MFIs major objective of regulation (box 2.7). do not think of them as subsidies on interest GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 97 BOX 2.7 Monopoly Rents, Bank Concentration, and Private Credit Reporting The existence of a comprehensive credit reporting the data also show that higher bank concentration is system is beneficial for the financial market as a associated with the less-extensive coverage and qual- whole, but individual lenders may profit from shar- ity of the information being distributed by credit ing only limited information with other market par- bureaus. These findings suggest that market fail- ticipants. If only one lender has credit information ures can prevent the development of effective credit- on a fi rm or individual, this lender faces less com- sharing systems, implying that the state may have to petition in lending to these borrowers because other intervene to help overcome these obstacles. institutions may be reluctant to offer them credit. In economic terms, a lender can capture monopoly rents by not sharing information. This issue may be FIGURE B2.7.1 Credit Bureaus and Registries Are particularly pronounced if the market for credit is Less Likely if Banks Are Powerful dominated by a few large banks. These banks would 1.0 each already have a broad customer base and may 0.92 Probability of existence of a credit reporting try to maintain their large market share by holding 0.80 onto information. Not making information available 0.8 can also prevent entry by new banks. Bruhn, Farazi, and Kanz (2012) examine the 0.56 0.6 relationship between bank concentration and the 0.53 emergence of private credit reporting. Using data for 0.37 0.39 close to 130 countries, the authors fi nd that bank 0.4 concentration is negatively associated with the prob- ability that a credit bureau emerges. Figure B2.7.1 0.2 illustrates that 80 percent of countries with low bank concentration have a credit bureau, whereas only 39 percent of countries with high bank concentration 0 have a credit bureau. This difference is smaller for Credit registry Credit bureau Any credit reporting credit registries (56 percent vs. 37 percent), which Low bank concentration High bank concentration may reflect the fact that banks are required to report Source: Based on Bruhn, Farazi, and Kanz 2012. to a credit registry, while participation in a credit Note: The figure shows the percentage of countries with private (credit bureau is often voluntary. bureau), public (credit registry), or any credit reporting institutions. The countries are distinguished by high or low bank concentration (above or below This result is robust if one controls for confound- the sample mean). Bank concentration is the asset share of a country’s three ing factors that could bias the analysis. In addition, largest banks. rates. Many currently subsidized MFIs aspire were lower. Many would agree with Morduch to reach a break-even point and ultimately (1999) that the prospects of reaching many of become fully profitable. Others, including the the poor with unsubsidized credit are limited. well-known Grameen Bank, consciously apply But even if subsidized microcredit can be used subsidies to keep interest rates down. MFIs to broaden financial inclusion, credit is only that operate group-lending schemes and thus one among several financial service needs of that are more focused on the poor, rely on the the poor. Ensuring sustainable access to credit relatively largest amount of subsidies. markets requires greater access to other finan- While many borrowers are able and will- cial services, including savings and insurance ing to service interest rates at levels that allow products. efficient MFIs to be fully profitable, there is Policies that support access to credit at no doubt that demand and borrower sur- below market rates may also be part of gov- pluses would be even greater if interest rates ernment programs used to gain political sup- 98 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 port. Interest rate caps, interest rate subsidies, that the individuals with the accounts will and debt restructuring programs are common cite lack of funds as a barrier. This suggests around the world, and especially widespread that government policies to enhance inclu- in rural credit, where they can help in reach- sion can increase the likelihood that individu- ing geographically dispersed constituencies. als perceive that financial services are within While such programs may offer short-term their reach. welfare gains for borrowers, they can also There is growing evidence that postal lead to severe credit market distortions that networks can play a powerful role in finan- may hinder financial inclusion in the longer cial inclusion. According to some observers, run. As a case study of such a policy, box 2.8, postal networks around the world have gone examines the impacts of a government bailout or are going through a transformation from for heavily indebted rural households in India. mail-centered bureaucracies to diversified commercial enterprises with a social mission to serve the whole population and reduce Other direct interventions to address financial exclusion (Berthaud and Davico financial inclusion 2013; El-Zoghbi and Martinez 2012). The Several other direct interventions for financial arrival of the Global Findex data provides inclusion have attracted attention in recent some new cross-country empirical evidence years. These include government-to-person on the subject, suggesting where postal finan- (G2P) payments, the use of state-owned cial inclusion may be relevant and appli- banks, the use of government postal services cable.28 The data show that postal financial for financial inclusion, and explicit financial inclusion reaches the poorer, older, less well inclusion strategies. educated, and unemployed people at the bot- In the payments and savings arena, gov- tom of the pyramid. ernments can potentially play a direct role in enhancing financial inclusion by using Financial inclusion strategies G2P payments to raise the demand for bank accounts. These payments, which could A growing number of countries have adopted include social transfers and wage and pen- formal national financial inclusion strate- sion payments, can facilitate access through gies. These are public documents, developed existing government infrastructure, such as through a consultative process that typically post office networks. If well designed, these involves a variety of public sector bodies (the payments have the potential to become a ministry of finance or economy, the central vehicle for extending financial inclusion, and bank, consumer protection agencies, the min- some observers have noted that providing istry of justice, social protection agencies, and poor G2P recipients with financial services so on) as well as private sector firms (commer- could strengthen the development impact of cial banks, insurance firms, nonbank financial G2P payments (Bold, Porteous, and Rotman institutions, telecommunications firms), and 2012; Pickens, Porteous, and Rotman 2009). civil society partners (such as microfinance Allen, Demirgüç-Kunt, and others (2012) organizations and nongovernmental organi- examine the effects of G2P payments as zations in financial education). In many coun- part of a broader examination of the factors tries, the financial inclusion strategy is led by driving financial inclusion. Specifically, they the central bank. In Kenya, for example, the include in their models of account penetra- central bank was aided by Financial Sector tion a dummy variable for whether the gov- Deepening Trusts, supported by the United ernment reported it had encouraged or man- Kingdom’s Department for International dated the payment of government transfers Development, and surveys and research con- or social payments through bank accounts. ducted by Finmark Trust. In Colombia, a They find that the existence of accounts to government-created independent trust fund receive G2P payments lowers the likelihood has led efforts on financial access. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 99 BOX 2.8 Exiting the Debt Trap: Can Borrower Bailouts Restore Access to Finance? Many households around the world are excluded the access to credit at the household level. India’s from access to formal fi nancial services because of experiment with debt relief provides an excellent high levels of household debt. This is particularly chance to bring some empirical evidence to bear on true of rural households that are exposed to a vari- this debate. ety of recurring economic shocks (drought, export Two recent studies use the natural experiment price volatility), but that may lack the tools to insure arising from India’s large-scale borrower bailout against the resulting income volatility. The poten- to examine the effect of debt relief at the house- tially far-reaching aggregate implications of extreme hold and economy level. Kanz (2012) uses a sur- household indebtedness have motivated a number of vey of 2,897 households that were benefi ciaries of large government-led debt relief and debt restructur- India’s debt relief program to gather evidence on the ing programs that aim to bring recipients back into impact of debt relief on the subsequent economic the fold of the formal credit market. decisions of these households. One important inten- In India, where overindebtedness among rural tion of the program was to provide households with households has been a particularly substantial a “fresh start” by clearing pledged collateral and problem, the government has resorted to an unusu- enabling new borrowing that would allow house- ally bold policy response. Prompted by a wave of holds to fi nance productive investment. The results defaults among microfi nance borrowers and a highly suggest that the program was successful along only visible rise in farmer suicides in poorer regions of one of these dimensions. Debt relief led to a substan- the country, the government embarked on what was tial and persistent reduction of household debt. Even perhaps the largest household-level debt relief pro- one year after the program, beneficiaries of uncon- gram in history. Announced in February 2008, the ditional debt relief were approximately Rs 25,000 Agricultural Debt Waiver and Debt Relief Scheme ($464 or 50 percent of median annual household (ADWDRS) for Small and Marginal Farmers can- income) less indebted than households in a con- celled the outstanding debt of more than 40 million trol group. However, the program did not manage rural households across the country; this represented to reintegrate the recipient households into formal approximately 1.7 percent of India’s GDP. lending relationships. Despite the fact that banks What was the effect of this large-scale interven- were required to make bailout recipients eligible tion on access to fi nance, credit supply, and loan for fresh loans, a large fraction of the households performance? While the benefit of debt relief pro- did not use their freed-up collateral to take on grams for individual households can be substantial, new loans. This is directly refl ected in household the merit of such programs as a tool to improve investment and productivity: beneficiary house- household welfare in the long run remains highly holds reduce their investment in agricultural inputs controversial. Proponents of debt relief argue that (which tend to be largely credit fi nanced) and suffer extreme levels of household debt are likely to dis- a corresponding decline in agricultural productiv- tort investment and production decisions, and thus ity. Taken together, these household-level results debt relief holds the promise of improving the pro- suggest that merely writing off debt might be a nec- ductivity of benefi ciary households. Critics of debt essary, but not sufficient policy to help households relief, on the other hand, worry that it is diffi cult gain access to new fi nancing and engage in produc- to “write off loans without also writing off a cul- tive investment. ture of prudent borrowing and repayment.”a If debt This means that, if debt relief is to encourage new relief changes expectations and borrower behavior, investment, debt forgiveness should be accompanied they argue, bailouts may, in fact, lead to widespread by measures to restore banking relationships. By moral hazard and more severe credit rationing the same token, the results indicate that, among the in the future. Although both views can appeal to sources of constraints to produce investment, expec- a foundation in economic theory, there exists sur- tations about future access to credit are as important prisingly little evidence on how indebtedness and, as the disincentives generated by the debt overhang hence, debt relief affect economic decisions and arising from high levels of inherited debt. Hence, (box continued next page) 100 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 2.8 Exiting the Debt Trap: Can Borrower Bailouts Restore Access to Finance? (continued) a bailout designed to encourage new investment is most strongly affected regions has rebounded in the likely to be effective only if it is implemented in a years since the bailout, the results suggest that the way that credibly improves the long-term expecta- program has had a negative long-run impact on the tions of households about their access to finance. fi nancial access of marginal borrowers. This under- While the microevidence gathered from household- scores that a major challenge in designing effective level studies can tell us something about the effect of debt relief and restructuring programs is to mini- debt relief on individual households, it does not mize the adverse effect on repayment incentives and allow us to draw inferences about the larger eco- credit market discipline. nomic effects of such programs. For example, an Are there alternatives to borrower bailout pro- important caveat about large-scale debt relief or grams in settings in which overindebtedness is as restructuring programs is that they might generate widespread as in the case of India? Some countries significant moral hazard and do lasting damage to a have experimented with debt restructuring that country’s credit culture. This, in turn, might lead makes relief conditional on borrower behavior. banks to change the way they allocate credit that While this approach is less likely to distort repay- might make it even more difficult for marginal ment incentives, it has often been more diffi cult to households to obtain credit in the future. Are these establish eligibility and enforce compliance with concerns justified? conditional debt relief programs. Perhaps the most To explore these questions, Giné and Kanz (2013) efficient policy solution to the problem of excessive use district-level data on credit supply and loan household debt is to strengthen provisions for per- performance for a panel of Indian districts for the sonal bankruptcy settlements. This would allow for period between 2001 and 2012, that is, before and the orderly discharge of household debt in a way that after ADWDRS. The study shows that the bailout avoids incentive distortions and deals with the prob- led to a reallocation of new credit away from high lem before it becomes so severe as to require large- bailout districts, an overall slowdown in new loans, scale policy interventions into the credit market that and signifi cant moral hazard in the post program are prone to mismanagement and political capture. period (figure B2.8.1). Although credit supply to the FIGURE B2.8.1 Bailouts and Moral Hazard a. Credit growth b. Loan performance 70 15 6,000 Program Program launched 60 launched 5,000 Agricultural credit, Rs million 13 Nonperforming loans, % 50 4,000 Loans ['000] 40 11 3,000 30 9 2,000 20 1,000 7 10 0 0 5 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2006 2007 2008 2009 2010 2011 2012 New credit (amount) New loans (number) Nonperforming agricultural loans/loans outstanding Source: Based on Giné and Kanz 2013. Note: Rs = Indian rupees. a. “Waiving, Not Drowning: India Writes Off Farm Loans. Has It Also Written Off the Rural Credit Culture?” The Economist, July 3, 2008. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 101 The main stated objective of these strate- and initiatives, greater knowledge of the sector, gies is to increase the access to finance. A and clearer commitments to good practices. typical strategy would highlight a headline Evidence on the effectiveness of the target. For example, Nigeria’s strategy, pub- national financial inclusion strategies is still lished in October 2012, aims to reduce finan- only emerging. Duflos and Glisovic-Mézières cial exclusion in the country from 46 percent (2008) discuss national microfinance strate- to 20 percent by 2020. Beyond the headline gies in some 30 countries, mostly in Africa, target, the strategy defines financial inclu- pointing out that usually the strategies have sion, identifies key stakeholders, outlines been initiated at the request of donors. They their roles and responsibilities, reports on the find little evidence that the strategies have status of financial inclusion in Nigeria rela- influenced access to finance in the countries tive to international benchmarks, discusses where they have been adopted. They identify barriers to financial inclusion, introduces a some challenges, such as weak diagnostics, range of targets, presents a range of strate- inadequate government leadership, and unre- gies for achieving financial inclusion targets, alistic action plans. The paper offers sugges- discusses the implications for regulation and tions to donors on the development of the policy in Nigeria, introduces regular moni- strategies. These include investing in compre- toring and evaluation, and puts in place an hensive sector diagnostics (so that the strategy organizational framework for the institution- is based on solid data), analyzing the political alization of the strategy. Notably, the strat- climate, ensuring local ownership, evaluating egy sets rather specific targets for payments, results, and being open to changing course. A savings, credit, insurance, pensions, branches recently published framework document for of deposit banks and microfinance banks, financial inclusion strategies describes how and so on. The key initiatives in the strategy selected countries have approached this task include a tiered approach to spreading aware- and provides general guidance for nations ness of customer rules, agent banking, mobile that may be contemplating such strategy exer- payments, cashless policies, the financial lit- cises (World Bank 2012i). eracy framework, consumer protection, and While it is still too early to perform a rigor- the implementation of credit enhancement ous examination of these strategies, there are schemes and programs. some initial signs that improvements can be The Alliance for Financial Inclusion, a achieved through shared public and private global network of financial policy makers sector commitments to financial inclusion. from developing and emerging countries, For example, the South Africa Financial Sec- has begun a Policy Champion Program to tor Charter helped raise the percentage of promote dissemination of good practices. In banked adults from 46 to 64 percent in four 2012, alliance member states issued the Maya years, and six million basic bank accounts Declaration, the first global and measurable (called Mzansi accounts) were opened. In set of commitments by developing and emerg- the United Kingdom, a Financial Inclusion ing country governments to unlock the eco- Taskforce contributed to halving the number nomic and social potential of the 2.5 billion of unbanked adults through e-money regula- unbanked people worldwide through greater tion, G2P payments linked to bank accounts, financial inclusion. More than 40 countries and access to financial services through post have signed the declaration.29 offices. The government of Brazil imple- An important perceived benefit of these mented regulatory reforms that enabled the strategy documents is better coordination of financial sector to respond with innovative multiple stakeholders and initiatives. Indeed, products and enhance access, leading to a most country authorities report that the con- dramatic expansion in financial service access sultative process involved in preparing the points. As a result, every municipality in the strategies leads to improved dialogue and bet- country, including in remote rural areas, is ter coordination among multiple stakeholders now covered. 102 FINANCIAL INCLUSION FOR INDIVIDUALS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 NOTES ity. This is useful in the context of a push of microlenders who are trying to move toward 1. There are a few variants of debit cards, individual liability, especially on bigger loans. including delayed debit cards, whereby the 11. See “Bank On San Francisco,” San Francisco payment instruction resulting from the use of Office of Financial Empowerment, San Fran- the debit card for a payment results in placing cisco, http://sfofe.org/programs/bank-on. See a hold on the funds in the underlying account, also Department of the Treasury (2011). as opposed to resulting directly in a debit. 12. Another relevant example, Mexico’s Banco 2. While the terms prepaid card and stored-value Azteca, is discussed in chapter 3, box 3.3. card are frequently used interchangeably, 13. See “Product Design Case Studies,” Inter- differences exist between the two products. national Finance Corporation, Washing- Prepaid cards are generally issued to persons ton, DC, http://www.ifc.org/wps/wcm who deposit funds into an account of the card /connect/industry_ext_content/ifc_external issuer. During the prefunding of the account, _corporate_site/industries/financial+markets most issuers establish an account and obtain /publications/product+design+case+studies. identifying data from the purchaser (name, 14. See “Product Development,” CGAP Topics, phone number, and so on). Stored-value cards Consultative Group to Assist the Poor, World do not typically involve a deposit of funds Bank, Washington, DC, http://www.cgap into an account because the prepaid value is stored directly on the cards. .org/topics/product-development. 3. Data of World Development Indicators (data- 15. Human-centered design refers to a process base), World Bank, Washington, DC, http:// whereby products are designed taking into data.worldbank.org/data-catalog/world account the needs of the people and commu- -development-indicators; Global Financial nities for which the products are intended. Inclusion (Global Findex) Database, World See “Human-Centered Design Allows Us Bank, Washington, DC, http://www.worldbank to Create and Deliver Solutions Based on .org/globalfindex. See also Demirgüç-Kunt People’s Needs,” HCD Connect, http://www and Klapper (2012). .hcdconnect.org/toolkit/en. 4. The focus here is on mobile technology 16. The same factors that have focused more because of its growth and its promise to attention on financial capability are also increase inclusion, but, to put matters in per- behind a growing emphasis on financial spective, the bulk of the agent-based transac- consumer protection regimes. This includes tions mentioned in this paragraph can be and activities by the Financial Stability Board, the still are carried out through payment cards G-20, and the Organisation for Economic rather than mobile phones. Co-operation and Development (OECD). 5. Data of the Reserve Bank of India, at http:// 17. The World Bank has also organized Financial www.rbi.org.in. Capability and Consumer Protection surveys, 6. Global Financial Inclusion (Global Findex) which build on the World Bank–Russia Trust Database, World Bank, Washington, DC, Fund questionnaire and complement it with http://www.worldbank.org/globalfindex. financial knowledge and consumer protection 7. See Markswebb Rank & Report website, at questions. The results of this expanded survey http://markswebb.ru/ [in Russian]. are not yet available. 8. See “e-Commerce in Russia: What Brands, 18. Financial literacy surveys using different ques- Entrepreneurs, and Investors Need to Know tionnaires were previously undertaken by the to Succeed in One of the World’s Hottest World Bank in countries such as Azerbaijan, Markets,” East-West Digital News, http:// Bosnia and Herzegovina, Bulgaria, Romania, www.ewdn.com/e-commerce/insights.pdf. and Russia. 9. See “Product Design Case Studies,” Inter- 19. There is extensive research finding that national Finance Corporation, Washing- financially capable behavior is influenced ton, DC, http://www1.ifc.org/wps/wcm by psychological traits and not merely by /connect/industry_ext_content/ifc_external the individual’s financial knowledge. On the _corporate_site/industries/financial+markets United Kingdom, De Meza, Irlenbusch, and /publications/product+design+case+studies. Reyniers (2008) analyze a survey of financial 10. Carpena and others (2011) examine a micro- capability and find that psychological dif- lender’s shift from individual to group liabil- ferences (such as the propensity to procras- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR INDIVIDUALS 103 tinate) rather than informational differences an impact of financial education on defaults. explain much of the variation in financial Caution is required in interpreting this result capability. On the United States, Mandell and because the studies on which the meta- Klein (2011) analyze data from the Jump$tart analysis is based are diverse in many charac- national survey of high school students and teristics, including the countries in which the find that motivation is a key determinant of interventions occurred and the intensity and financial literacy levels. type of the interventions. 20. These include Atkinson (2008); Fernandes, 23. See the Global Financial Inclusion (Global Lynch, and Netemeyer (forthcoming); Gale Findex) Database, World Bank, Washington, and Levine (2010); Hastings, Madrian, and DC, http://www.worldbank.org/globalfindex. Skimmyhorn (2012); Hathaway and Khati- 24. For a comprehensive review of this topic, see wada (2008); Lusardi and others (2010); World Bank (2012a). Martin (2007); and Xu and Zia (2012). The 25. See Cull, Demirgüç-Kunt, and Lyman (2012) review suggests that financial knowledge is and “Ministers Renew the Mandate of the linked to higher levels of retirement plan- Financial Action Task Force until 2020,” ning and savings (Alessie, Van Rooij, and Financial Action Task Force, Paris, April 20, Lusardi 2011; Behrman and others 2010; 2012, http://www.fatf-gafi.org/topics/fatf Bucher-Koenen and Lusardi 2011; Lusardi general/documents/ministersrenewtheman and Mitchell 2007, 2011b), the quality of dateofthefinancialactiontaskforceuntil2020 investment decisions (Abreu and Mendes .html. 2010; Van Rooij, Lusardi, and Alessie 2011), 26. Based on interviews at the Consumer Pro- credit management and satisfaction (Akin tection Department during the March 2012 and others 2012), and mortgage performance Financial Sector Assessment Program mission. (Ding, Quercia, and Ratcliffe 2008; Gerardi, 27. Other studies find similar results in consider- Goette, and Meier 2010; Quercia and Spader ing aggregate bank credit as opposed to lend- 2008). ing to households (see Bae and Goyal 2009; 21. In particular, the design, quality, and inten- Qian and Strahan 2007). siveness of interventions vary, making com- 28. Global Financial Inclusion (Global Findex) parisons difficult; until recently, most studies Database, World Bank, Washington, DC, focused on developed economies; few studies http://www.worldbank.org/globalfindex. evaluate long-term or sustainable changes; 29. Under the declaration, each country makes and few provide a cost-benefit analysis of the measurable commitments in four areas: intervention compared with other possible (1) creating an enabling environment to har- policies. ness new technology that increases access 22. Based on a meta-analysis of over 100 stud- to and lowers the costs of financial services; ies, Miller and others (2013) find that, while (2) implementing a proportional framework some of these papers cannot reject the null that advances synergies in financial inclusion, hypothesis that financial education has no integrity, and stability; (3) integrating con- effect on savings, these papers, taken together, sumer protection and empowerment as a key do supply evidence of an impact on savings. pillar of financial inclusion; and (4) utilizing The study also exposes positive evidence of an data for informed policy making and track- impact on recordkeeping, but no evidence of ing results. CHAPTER 3: KEY MESSAGES • One of the key channels through which finance affects economic growth is the provision of credit to the most promising fi rms. Micro and small fi rms and young fi rms face the great- est constraints to access to fi nance because of market imperfections such as information asymmetries and transaction costs. Finance also has a crucial role in supporting innovation. Research shows that the use of external finance is associated with greater innovation by pri- vate small, medium, and large enterprises. To promote job creation and economic growth, it is also important to address the fi nancing challenges associated with new entry and with young firms that are financially constrained because of a lack of information and collateral. • Recent research concludes that the provision of microcredit is not enough to spur microen- terprise investment and firm growth in part because borrowers have heterogeneous growth prospects, lack managerial capital, and face regulatory barriers and psychological or behav- ioral constraints. And, in part, debt may not be the appropriate instrument because repay- ment requirements without grace periods and joint liability discourage risky investment. Theoretically, equity-like contracts may overcome this problem, but they may not emerge because of moral hazard and a weak legal environment. While there is little evidence on the effectiveness of microequity, recent studies suggest that savings products and microinsurance can spur microenterprise investment and growth. • Financial inclusion can be considerably enhanced by improvements in fi nancial sector infra- structure. Mounting evidence indicates that movable collateral frameworks and registries as well as credit information systems can boost lending to small and medium enterprises (SMEs) by overcoming information problems. Encouraging greater competition in the fi nan- cial sector is critical in promoting innovation among financial institutions and the adoption of technologies that help cover underserved segments such as SMEs. • Business models and product design matter. Leveraging relationships can be vital in lending to micro and small firms. Novel mechanisms deliver credit through retail chains or large sup- pliers, while relying on payment histories to make loan decisions and lowering costs by using existing distribution networks. • Financial and business training in an effort to enhance financial management leads to greater expertise, although the impacts on business practices and performance tend to be small, depend on context and gender, and show mixed results. The content of training also matters: simple rule-of-thumb training may be more effective than standard business and accounting training. • Direct government interventions in the credit market, such as directed lending programs and risk-sharing arrangements, can have positive effects on the access of SMEs to fi nance and growth, but it is a challenge to design and manage the interventions properly. This prob- lem is even greater in weak institutional environments where good governance is difficult to establish. • Woman-owned firms and agricultural firms face particular challenges in gaining access to fi nance. Woman-owned fi rms tend to be smaller than fi rms owned by men and grow at a slower rate partly because women have less access to finance. • Agricultural firms are constrained because of geography as well as high seasonal variability and weather-related risks in agricultural production. Financial products, such as insurance, and government programs can help mitigate these risks and promote investment and produc- tivity among agricultural firms. FINANCIAL INCLUSION FOR FIRMS 3 Financial Inclusion for Firms O ne of the core functions of financial systems is the allocation of funds to the most productive uses. These productive may access through banks or other forms of finance, such as factoring and leasing (see below). The policy interventions targeted at uses may be available in firms that have good microenterprises and SMEs may thus also be growth opportunities and that need addi- different.3 tional funds to finance working capital and Research indicates that microenterprises fixed asset investments. However, some of enjoy high returns to capital. This chapter these firms may be excluded from accessing addresses the question whether lack of access external finance because of principal-agent to microcredit, savings, and insurance repre- problems and transaction costs (see chap- sents a barrier to the purchase by microenter- ter 1, box 1.1). prises of more capital to realize these returns. This chapter examines the challenges and The chapter summarizes the evidence from a possible solutions in enhancing the financial range of impact evaluations and concludes inclusion of firms. The first part of the chapter that microcredit is not sufficient to support focuses on microenterprises, followed by an microenterprise investment and firm growth. examination of the situation among small and One reason for this finding appears to be medium enterprises (SMEs) and young firms that stiff repayment requirements and joint because principal-agent problems and trans- liability can discourage investment. Equity- action costs that restrict access to finance are like contracts may overcome this problem, particularly relevant for these firms.1 Micro- but, currently, there is a lack of evidence on enterprises, defined here as firms with fewer whether equity investments can be used suc- than 5 employees, and SMEs, defined as firms cessfully to support microenterprise growth. with 5–99 employees, are discussed separately Some recent studies find, however, that sav- because their financing needs tend to be dif- ings products and insurance can promote ferent.2 Microenterprises often need relatively investment in microenterprises. small loans that are provided by microfinance The chapter then argues that a sizable institutions (MFIs). SMEs, on the other hand, fraction of SMEs in developing countries are require larger amounts of credit that they credit constrained because of principal-agent GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 105 106 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 problems and lack of the financial infrastruc- for young firms because they take on risk and ture necessary to mitigate these problems. also contribute business experience through Within-country studies show that relieving their mentorship. Currently, angel investor credit constraints can foster SME growth. groups are less common in developing econo- Governments can take a range of actions to mies than in developed economies, but finan- mitigate the credit constraints affecting SMEs. cial and technical assistance can help support Direct interventions in the credit market, such the creation of angel investor networks. as directed lending programs and risk-sharing The chapter examines the relationship arrangements, can have positive effects on between firm finance and innovation to SME access to finance and growth, but it is a understand the channel through which finan- challenge to get the design and management cial inclusion can promote economic growth. right. These concerns are even greater in weak Innovative projects are especially difficult to institutional environments where good gover- finance with external funds because they are nance is difficult to establish and SMEs face often risky, and firms are reluctant to disclose particularly severe financial constraints. much information about inventions that may Policy makers can take other, more be imitated by others. Governments can step market-oriented actions to encourage SME in by sponsoring matching grant programs lending. These actions aim to improve the and business plan competitions. Evidence ability of financial sector infrastructure to suggests that matching grants can improve mitigate the principal-agent problems faced firm productivity and employment growth, by SMEs. Mounting evidence shows that legal but these grants are also difficult to imple- frameworks and registries for movable col- ment well. Research on the impact of business lateral, as well as credit information systems, plan competitions is under way, but not yet can increase lending to SMEs. Factoring and completed. leasing are two ways to support financing to The chapter then turns to two topics SMEs, but there is currently little evidence involved in firm financial inclusion that have documenting the impact of these two financial recently received much policy and research instruments on SME investment and growth. notice. First, it asks whether gender matters The chapter also points out that SMEs in access to financial services. Several stud- with risky, but potentially lucrative invest- ies document that women have less access ment opportunities may not be able to obtain to credit because women tend to own fewer bank loans to fund these projects. Private assets than men, have lower levels of educa- equity investors can step in to share the risk tional attainment, are restricted by societal and potential rewards. However, these types norms, or are sometimes subject to taste of investments rely on sound legal systems discrimination. Evidence also shows that for contract enforcement and previous busi- woman-owned firms tend to be smaller than ness expertise, which may be lacking in firms owned by men and grow at slower rates. developing countries. International agencies The chapter concludes that getting more have launched projects to help develop local credit to women can promote the growth of private equity industries. More research is woman-owned firms if policies also address needed to document the extent to which these the underlying factors that lead to the gender projects promote SME growth. gap in access to finance in the first place. Much policy and research attention has Second, the chapter discusses the particular been focused on the financial inclusion of challenges and financing needs of agricultural SMEs, but young firms also face financial firms. Agricultural production is subject to constraints. Relieving these constraints can seasonal variability and risks, such as pests, potentially foster productivity growth and job livestock diseases, and adverse weather con- creation, although more evidence is needed ditions. As a result, agricultural firms may to document this relationship. Angel inves- decide to forgo riskier investments to guar- tors provide a promising source of finance antee a lower, but more certain stream of GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 107 income. Lenders may also be reluctant to pro- FIGURE 3.1 Percentage of Micro, Very Small, Small, vide financing for risky projects. The chapter and Medium Firms reviews how financial products, such as insur- ance, and government programs can help mit- 100 2 2 1 3 igate these risks and promote investment and 8 6 12 90 15 productivity among agricultural firms. 80 17 23 18 70 23 MICROENTERPRISES 60 Firms, % 50 The vast majority of firms around the 40 world are microenterprises. The Interna- 74 69 69 tional Finance Corporation (IFC) Enterprise 30 59 Finance Gap Database shows that about 20 three-fourths of formal microenterprises and 10 SMEs in developing economies are microen- 0 terprises, defined in this data set as enterprises Low Lower middle Upper middle High with 1–4 employees.4 Even in developed Country income level economies, 59 percent of all microenterprises Micro firms Very small firms l firms Small Medium firms and SMEs are microenterprises (figure 3.1). About 80 percent of microenterprises and Source: IFC Enterprise Finance Gap Database, SME Finance Forum, http://smefinanceforum.org. Note: The data cover 177 countries. The observation year varies from 2003 to 2010 by country. Firm SMEs are informal.5 The data do not pro- size is based on the number of employees (1–4: micro; 5–9: very small; 10–49: small; and 50–250: vide a separate number for microenterprises medium). The data set does not include large firms (typically less numerous than medium firms). only, but informality rates tend to be highest among the smallest firms (de Mel, McKenzie, and Woodruff, forthcoming). According to FIGURE 3.2 Biggest Obstacles Affecting the Operations of the World Bank informal enterprise surveys, Informal Firms most firms in the informal sector report that lack of access to finance is the biggest obstacle 40 36 they face (figure 3.2).6 Box 3.1 summarizes a 30 Firms, % 22 study by Farazi (2013) that looks at financ- 20 16 ing issues among microenterprises and small 13 9 10 4 firms in the informal sector. 0 Limited Crime, Other Restricted Poor Corruption High returns to capital for access to theft, and access to public finance disorder land infrastructure microenterprises Theory and evidence suggest that microen- Source: World Bank informal enterprise surveys. terprises can obtain high returns to capital. Note: The figure covers microenterprises (0–5 employees) and small firms (6–20 employees) in 13 countries. The bars represent the percentage of firms identifying a given option as a major From a theoretical standpoint, returns to cap- obstacle to business. “Other” includes difficult business registration procedures, the workforce, ital at enterprises that operate at a small scale limited demand for products or services, and political instability. may be high (and notably higher relative to larger firms) if production functions display decreasing returns. Studies in Ghana, Mexico, and Woodruff (2008) use a field experiment and Sri Lanka have provided empirical evi- to illustrate the high returns to capital among dence that returns to capital among micro- microenterprises (box 3.2). enterprises are indeed sizable, ranging from Given the evidence on the high returns to 5 to 60 percent per month (de Mel, McKen- capital, a natural question is evoked: why do zie, and Woodruff 2008; Khandker, Samad, microentrepreneurs not purchase additional and Ali 2013; McKenzie and Woodruff 2008; capital assets (for example, machinery or tools) Udry and Anagol 2006). De Mel, McKenzie, to realize these returns? In other words, can 108 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.1 Financial Inclusion of Informal Firms: Cross-Country Evidence Lack of access to fi nance is identified as the biggest owners have jobs in formal businesses, implying that operational challenge by informal fi rms according to the informal business is their main source of income. the World Bank informal enterprise surveys. a This Informal firms report low use of loans and bank box is based on an analysis of issues in access to accounts, and a significant majority finance their fi nance among informal fi rms. The analysis was car- operations through sources other than financial ried out by Farazi (2013) using World Bank surveys institutions, including internal funds, moneylend- of approximately 2,500 fi rms across 13 countries in ers, family, and friends. A majority of the respon- Latin America and Sub-Saharan Africa. dents would like their fi rms to become formal (that A majority of informal firms are microenter- is, to register), but do not do so because registration prises, that is, they have no more than five employ- requires them to pay taxes. They state that the ease ees, and started operations less than 10 years ago. of access to fi nance would be the most important Most informal fi rm owners have received some type benefit they could obtain from registering (table of educational or vocational training. Few of the B3.1.1). TABLE B3.1.1 Snapshot of Informal Firms Indicator Share of firms, % Size: microenterprise (0–5 employees) 89 Age: 10 years or less 74 Level of education: no education 6 Largest owner has a job in a formal business 8 Use of accounts 23 Use of loans 11 Working capital finance: internal funds, family, and moneylenders 80 Investment finance: internal funds, family, and moneylenders 84 Would like to register 59 Main reason for not registering: taxes 26 Most important benefit of registering: access to finance 52 Source: Farazi 2013. The use of finance by informal firms is signifi- Recent research indicates that lowering ini- cantly associated with (1) fi rm size, (2) the owner’s tial registration costs and providing information level of educational attainment, and (3) the owner on registration procedures have only small effects having a job in the formal sector. The results of the on firm formalization (Bruhn 2013; de Andrade, regression analysis conducted by Farazi (2013) sug- Bruhn, and McKenzie 2013; De Giorgi and Rah- gest that, relative to small enterprises, microenter- man 2013). In ongoing research for the World prises show lower rates of use of bank accounts Bank, Francisco Campos, Markus Goldstein, and and rely less on banks and more on MFIs for working David McKenzie fi nd that the variable costs associ- capital fi nancing. Size is not significantly associated ated with becoming formal, such as tax payments, with the use of loans. The owner’s level of education may be comparatively more important for infor- and the owner’s having a job in the formal sector are mal firms. Unless these firms grow and become positively associated with the use of bank accounts suffi ciently profitable to cover such costs, it would and negatively associated with the use of internal be difficult for them to enter the formal sector. funds and fi nancing from family and moneylenders Enhancing the fi nancial inclusion of informal fi rms for working capital. The higher educational levels of can potentially help them grow and pave their path owners are also positively associated with the use of toward formalization (Campos, Goldstein, and loans by informal firms (figure B3.1.1). McKenzie 2013). (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 109 BOX 3.1 Financial Inclusion of Informal Firms: Cross-Country Evidence (continued) FIGURE B3.1.1 Use of Finance by Informal Firms 15 0.10 10 probability of use of finance 0.06 0.06 5 0.04 Regression estimates, 0.02 0 –5 –0.03 –0.06 –0.07 –0.07 –10 –15 –0.16 –20 Size Education Job Education Education Job Size Job Size Education Accounts Loans Financing Sources: Internal funds, etc. Banks MFIs Regressions Source: Farazi 2013. Note: Variables listed above the line are independent variables, while those listed below the line are dependent variables. The orange bars are regression estimates for accounts as a dependent variable (equals 1 if a firm has a bank account and 0 otherwise); the blue bar is a regression estimate for loans as a dependent variable (equals 1 if a firm has a loan and 0 otherwise); and the red bars are estimates derived from separate regressions for different sources for the financing of working capital (outcome variables equal 1 if firms use a given financing source and 0 otherwise). Each regression controls for firm-level vari- ables (age, microsize, sector, owner’s gender, educational level, and job in the formal sector) and country fixed effects. The results are robust if country-level variables (proxies for financial sector development and the quality of institutions) are used instead of country fixed effects. A probit model is used; estimates show marginal effects. MFI = microfinance institution. a. For additional information on the informal surveys, see “Enterprise Surveys Data,” World Bank, Washington, DC, http://www.enterprisesurveys.org/Data. greater access to financial products, including Credit for microenterprises credit, savings, and insurance, promote micro- Commercial banks often do not lend to micro- enterprise investment and growth? Note that, enterprises because the operational costs of apart from lack of financial access, there are lending to these firms are high relative to the many other potential reasons why microenter- revenue generated by the small loan amounts. prises do not invest more, including regulatory Microfirms also typically lack sufficient col- barriers, lack of skills or qualified employees, lateral to pledge against commercial bank will to remain informal, and psychological loans. A study in Sri Lanka finds that only or behavioral constraints. A detailed review about 10 percent of microenterprises receive of these issues goes beyond the scope of this bank loans (de Mel, McKenzie, and Wood- chapter. They are, however, discussed again ruff 2011). Many more would like loans, but in this section in so far as they influence the are constrained by their inability to provide access of microenterprises to finance or the collateral or personal guarantors or by other impact of finance on firm growth. bureaucratic procedures. The study also finds 110 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.2 Returns to Capital in Microenterprises: Evidence from a Field Experiment The rapid increase in development funding directed capital stock and to measure the impact of the toward MFIs raises a central question for policy additional capital on business profits. The interven- makers: do small and informal fi rms have the poten- tion consisted of one of four grants: SL Rs 10,000 tial to grow, or do they only represent a source of worth of equipment or inventories for their business, subsistence income for low-productivity individuals SL Rs 20,000 worth of equipment or inventories, unable to fi nd alternative work? De Mel, McKenzie, SL Rs 10,000 in cash, and SL Rs 20,000 in cash. and Woodruff (2008) attempt to answer this ques- The SL Rs 10,000 treatment was equivalent to about tion using randomized grants to a set of Sri Lankan three months of median profit reported by the fi rms microenterprises. They carried out a baseline inquiry in the baseline survey, and the larger treatment was on microenterprises in April 2005 as the fi rst wave equivalent to six months of median profits. For in- of the Sri Lanka Microenterprise Survey; eight addi- kind treatments, the amount spent on inventories tional waves of the panel survey were conducted at and equipment by entrepreneurs sometimes differed quarterly intervals through April 2007. To ensure the from the amount offered under the intervention. grant intervention would be a sufficiently large shock Entrepreneurs usually contributed funds of their to business capital, the survey organizers included own to purchase larger items. In the case of the cash only fi rms with invested capital of SL Rs 100,000 grants, on average, 58 percent of the cash treatments (about $1,000) or less. Of the 659 enterprises that were invested in the business between the time of the were surveyed, fi rms directly affected by the Decem- treatment and the subsequent survey. ber 2004 Indian Ocean tsunami were excluded from Figure B3.2.1 shows estimates of the treatment the analysis, leaving a sample of 408 fi rms. effect for the whole sample, on average, and for sce- The aim of the intervention was to provide ran- narios taking into account heterogeneous charac- domly selected fi rms with a positive shock to their teristics among the enterprise owners. For the aver- FIGURE B3.2.1 Estimated Returns to Capital 6 5.41 5.29 5 3.80 4 Regression estimates 3 2 1.56 1 –2.43 0 –1 –2 –3 Average returns Average returns Change in returns Change in returns Change in returns to capital to capital with for higher value for higher value for higher value controls of household of years of of digit span recall asset index education Source: De Mel, McKenzie, and Woodruff 2008. Note: The returns are shown in percent per month. The orange bar shows the average returns to capital for the whole sample. The blue bar shows the estimated returns based on regressions that control for indicators of ability (including years of schooling and digit span recall) and household assets. The household asset index is the first principal component of variables representing the ownership of 17 household durables; digit span recall is the number of digits the owner was able to repeat from memory 10 seconds after viewing a card showing the numbers (ranging from 3 to 11). The three red bars show how much returns changed relative to the blue bar at a higher household asset index, more years of education, or longer digit span recall, respectively. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 111 BOX 3.2 Returns to Capital in Microenterprises: Evidence from a Field Experiment (continued) age enterprise in the sample, the authors estimate other hand, returns to capital do not vary signifi - the real return to capital at about 5.4 percent per cantly with measures of risk aversion or uncertainty month (65 percent per year), which is substantially in sales and profits. The observed heterogeneity of higher than market interest rates. By examining the returns seems to suggest that the high returns are heterogeneity of treatment effects, the authors inves- more closely associated with missing credit markets tigate the importance of imperfect credit and insur- than missing insurance markets. ance markets. They claim that, within a context of The high returns at low levels of capital stock imperfect markets, returns to shocks to capital stock estimated by the authors indicate that entrepreneurs should be greater among entrepreneurs who are starting out with suboptimal capital stocks would be more constrained and more risk averse. The authors able to grow by reinvesting profits. Individuals might fi nd that returns vary substantially based on the abil- remain inefficiently small for some time, but would ity and household wealth of the entrepreneurs. The not be permanently disadvantaged. The authors fi nd, returns to shocks to capital stock are higher among however, such high levels of returns somewhat puz- more constrained entrepreneurs (those entrepreneurs zling. It is unclear what prevents fi rms from growing with fewer household assets, that is, less wealth) incrementally by reinvesting profits. The results point and entrepreneurs with higher ability as measured to the need for a better understanding of how these by years of education and digit span recall. On the microentrepreneurs make investment decisions. that a local bank allocates credit to microen- Brazil, Banco Santander has been promoting terprises with more household assets, that is, entrepreneurship and encouraging the growth collateral, and not to enterprises exhibiting of small businesses by making microloans to particularly high returns. informal microfirms that are unable to obtain Another reason why microenterprises may loans otherwise. The majority of their loans go have difficulty obtaining bank loans is that to businesses run by women. In Chile, Banco they are often opaque given their usually inad- Estado, through its subsidiary, BancoEstado equate documentation on formal accounts. Microempresas, targets segments of the popu- MFIs use lending techniques that allow them lation that are generally not served by com- to provide credit to these small, information- mercial banks. Relying on its broad country- ally opaque firms. They employ intensive wide capacity, the bank provides financial screening technologies to collect information services to microenterprises and low-income on potential borrowers, including visits to households. the borrower’s house or firm. These screening technologies imply high costs, meaning that Does microcredit promote investment? these institutions typically charge higher inter- est rates than banks. However, given that the Several recent studies have used rigorous returns to capital are also high among micro- research designs to examine whether micro- enterprises, credit from MFIs could potentially credit promotes investment and firm growth, lead to more investment and growth in these but have come up with mixed results. Most firms (see, for example, Khandker, Samad, papers tend to find positive effects on some and Ali 2013). business outcomes, but not on others, raising Despite the difficulties associated with additional questions. financing microenterprises, there are some A study conducted in Mexico City shows examples of banks that have successfully that improved access to microloans led to catered to microenterprises. For example, in increases in inventory investment and fixed 112 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 assets for very small retail enterprises (Cotler because these variables tend to have large and Woodruff 2007). Sales and profits also variances across firms and across time. went up, but the effects are not statistically Another reason why some studies find significant. positive impacts of microcredit on enterprise Similarly, a paper on India finds that growth while others do not may be that micro- microcredit allows households with an exist- finance borrowers are heterogeneous in terms ing business to invest more in durable goods, of their growth prospects. For example, de that is, to expand the business (Banerjee and Mel, McKenzie, and Woodruff (2010) show others 2010). However, microcredit has no that only 30 percent of microenterprise own- clear impact on the nondurable consump- ers in Sri Lanka have personal characteristics tion of these households, meaning that exist- similar to those of large firm owners, whereas ing businesses may or may not become more 70 percent have characteristics more akin to profitable if they scale up. those of wage workers, that is, a large share of In Bosnia and Herzegovina, individuals microenterprise owners may be running their who received a loan from an MFI as part of business to make a living while they are look- a randomized experiment were more likely to ing for a wage job and may not have plans for be self-employed or own a business relative to expanding the business (see also Bruhn 2013). individuals who did not receive a loan (Augs- In addition, the microenterprise owners burg and others 2012). The study also finds who would like to expand their business may evidence of increased business investment. lack the necessary managerial capital to do However, this greater investment did not so (Bruhn, Karlan, and Schoar 2010). Some necessarily translate into substantially higher MFIs provide financial and business train- business profits. ing to their clients in an effort to improve the In contrast, a large microcredit initiative in financial management and use of microloans Thailand had no measurable effect on busi- by these clients.7 While many studies find ness investment, but business income did rise improvements in knowledge deriving from because of the expansion in credit (Kaboski the training, the impacts on business practices and Townsend 2012). A possible explanation and performance are relatively small. Alterna- for these findings is that the businesses used tively, other studies find benefits of the train- the microcredit to finance working capital. ing among the MFIs, such as changes in the An impact evaluation in the Philippines likelihood that clients will be retained or in shows that loans going to microentrepreneurs the characteristics of the clients who apply reduced the number of business activities and for loans. Other studies find that the train- employees in these businesses and that sub- ing is most effective among certain groups in jective well-being declined slightly (Karlan different contexts, such as women in Bosnia and Zinman 2011). However, the microloans and Herzegovina or men in Pakistan (Bruhn raised the ability to cope with risk, strength- and Zia 2013; Giné and Mansuri 2012). ened community ties, and boosted the access One study shows that the content of train- to informal credit. ing may matter: simple rule-of-thumb train- Overall, these findings point to a positive ing leads to considerable improvements in impact of credit on microenterprises, although business practices, while standard accounting this impact is not always reflected in greater training does not. However, neither type of investment and growth. One caveat here is training has a strong effect on business sales that most research papers examine only the (Drexler, Fischer, and Schoar 2011). Overall, short-term impact of microcredit; it may take it appears that financial and business train- more time for loans to translate into increased ing does not substantially amplify the impact profit or growth. In addition, the study sam- of microloans on microenterprise investment ples may not always be large enough to gener- and growth across the board, although it can ate sufficient statistical power to detect signifi- have positive effects among certain borrowers cant effects on investment, sales, and profits in various settings. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 113 It is also possible that microloans are not access to bank accounts increases savings and the proper financial instrument for encour- productive investment among women market aging investment and growth. Most micro- vendors, but not among men bicycle-taxi driv- loan contracts require that repayment begin ers (Dupas and Robinson 2013). One reason immediately after loan disbursement. A study women benefited more from formal savings conducted in India shows that this repay- accounts could be that they may face regu- ment requirement can discourage risky illiq- lar demands on their incomes from relatives, uid investment and thereby limit the impact neighbors, or husbands (Ashraf, Karlan, and of microcredit on firm growth (Field and Yin 2010). More broadly, savings accounts others, forthcoming). Similarly, joint liability may also promote discipline among individu- for microloans may discourage investment als with present-biased preferences who are because group members have to pay more if tempted to spend the cash they hold (Gul and a fellow borrower makes a risky investment Pesendorfer 2004; Laibson 1997). Further- that goes bad, but they do not enjoy a share more, the graduation model supported by the of the profits if the investment yields returns. Consultative Group to Assist the Poor (CGAP) Equity-like financing, in which investors share and the Ford Foundation emphasizes that sav- both the benefits and risks of more profitable ing regularly in a formal way can help poor projects, may be a solution to these incen- entrepreneurs build financial discipline and tive problems (Fischer forthcoming). There become familiar with financial service provid- is little evidence on the feasibility and impact ers (Hashemi and de Montesquiou 2011). of equity financing among microenterprises. Innovative commercial banks can reach However, a research team with World Bank out to microentrepreneurs by providing both participation is starting to implement and credit and savings accounts. For example, evaluate a microequity scheme in South Asia. Banco Azteca in Mexico is able to make The team will partner with local chambers of microloans with low documentation require- commerce and MFIs to select firms that will ments by relying on synergies with a large be offered a microequity contract. Under the retail chain owned by the same mother com- contract, firms will receive funds to invest in pany, Grupo Elektra. Most branches are machinery and equipment. Firms will have a located inside retail stores and share the costs choice between debt-like or equity-like repay- of the distribution network. Banco Azteca ment plans, that is, firms will agree to repay also draws on a large database that the retail a share of the invested amount, and they will chain has accumulated on customer repay- also pay the investor a share of firm revenue. ment behavior on installment loans. An In the first phase of the project, a range of impact evaluation has shown that the open- contracts will be available that will shed light ing of Banco Azteca promoted the survival on whether firms favor debt-like or equity- of informal businesses and led to expansion like contracts. in labor market opportunities and earnings among low-income individuals (box 3.3).8 Another feature of Banco Azteca is that it The role of savings in promoting offers consumption loans. Thus, in addition microenterprise growth to providing financial services to microentre- Savings accounts are an important financial preneurs, it boosts the purchasing power of instrument that may aid microenterprise own- potential microenterprise clients, which may ers to accumulate resources to purchase addi- promote firm growth. tional capital. These accounts may be par- ticularly beneficial for individuals who have Insurance among microenterprises difficulty saving at home because of demands on their cash holdings by family members or Another financial constraint among micro- because of behavioral biases. An impact evalu- enterprises may be the inability to insure ation in Kenya has documented that gaining income against the risk posed by volatile 114 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.3 The Effect of Financial Inclusion on Business Survival, the Labor Market, and Earnings In October 2002, Banco Azteca simultaneously tions of Banco Azteca branches to identify the causal opened more than 800 branches in Mexico in all impact of the opening of the branches on economic of the existing stores of its parent company, a large activity. They compare the changes in the employ- retailer of consumer goods, Grupo Elektra. From ment choices and income levels of individuals before the start, Banco Azteca catered to low- and middle- and after the branch openings across municipalities income borrowers who had mostly been excluded with or without Grupo Elektra stores at the time of from the commercial banking sector. It capitalized the branch openings. They control for the possibility on Grupo Elektra’s experience in making small that time trends in outcome variables may be differ- installment loans for the purchase of merchandise ent in municipalities that had Grupo Elektra stores and relied on the parent company’s rich data, estab- and those that did not have these stores. lished information, and collection technology. Banco The results show that the branch openings led Azteca was thus uniquely positioned to target the rel- to a 7.6 percent rise in the proportion of individu- evant segment of the population, which it estimated als who run informal businesses, but to no change in at more than 70 percent of all households. Many of formal businesses. This is consistent with anecdotal these households were part of the informal economy, evidence suggesting that Azteca targets lower-income operating small informal businesses that lacked the individuals, as well as with Azteca’s flexible docu- documentation necessary to obtain traditional bank mentation requirements. In contrast, formal business loans. Banco Azteca requires less documentation owners have easier access to commercial bank credit than traditional commercial banks, often accepting and likely prefer it because of the higher interest rates collateral and cosigners instead of the documents. charged by Azteca. Figure B3.3.1 illustrates that the Through a difference-in-difference strategy, proportion of informal business owners followed Bruhn and Love (2013) use the predetermined loca- a similar pattern across municipalities and across FIGURE B3.3.1 Individuals Who Work as Informal Business Owners in Municipalities with and without Banco Azteca over Time 15.0 9.0 owners (municipalities without Banco Azteca), % Banco Azteca opening owners (municipalities with Banco Azteca), % Individuals who are informal business Individuals who are informal business 14.5 8.8 8.6 14.0 8.4 13.5 8.2 13.0 8.0 12.5 7.8 12.0 7.6 2000 2001 2002 2003 2004 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th Quarter Quarter Quarter Quarter Quarter Municipalities without Banco Azteca Municipalities with Banco Azteca Source: Bruhn and Love 2013. (box continued next page) GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 115 BOX 3.3 The Effect of Financial Inclusion on Business Survival, the Labor Market, and Earnings (continued) time before the branch openings. After the branches (measured by demographic data on bank branch opened, the proportion of informal business owners penetration). These results provide additional evi- rose substantially in municipalities with branches, dence that the channel through which Banco Azteca but, on average, remained similar to the initial levels exerts an impact on economic activity is the expan- in municipalities without the new branches. Bruhn sion in access to financial services among low- and Love also fi nd that the branch openings led to income individuals. an increase in employment by 1.4 percent and an Banco Azteca also offers other fi nancial services, increase in income levels by 7.0 percent. including savings accounts, business loans, and con- The measured impacts of Banco Azteca are larger sumer loans. The measured rise in informal entre- among individuals with below median incomes and preneurial activity, employment, and income may in municipalities that were relatively underserved thus be associated with the greater access to a com- by the formal banking sector before Azteca opened bination of these products. entrepreneurial returns, that is, individuals lack of rainfall measured at weather stations, may be reluctant to start or expand a micro- as discussed in the context of agricultural enterprise because they do not know if they finance below. will succeed and if their investment will pay Overall, the evidence suggests that micro- off.9 A recent study in Mexico suggests that credit is not sufficient to promote microen- this constraint is at least as important as the terprise investment and firm growth. One constraint represented by the lack of liquid- reason for this finding appears to be that ity for capital investment and that microen- repayment requirements and joint liability terprises may benefit from insurance products can discourage investment. Equity-like con- (Bianchi and Bobba, forthcoming). The study tracts may overcome this problem, but there finds that the existence of a conditional cash is no evidence that equity investments can transfer program has a positive impact on the be successfully used to promote microenter- likelihood of entry into microentrepreneur- prise growth. Mounting evidence indicates, ship. The transfer effectively insures entre- however, that savings products and insurance preneurs against income risk because they can promote investment in microenterprises. can rely on the transfer in case their business Ongoing research will shed more light on this income is low. The setting in this study is thus question within the next couple of years. specific, and it is not clear to what extent the findings would carry over to an economy in SME FINANCING: which insurance products are sold to micro- WHY IS IT IMPORTANT? enterprises by an insurance company. These AND WHAT TO DO ABOUT IT? insurance products may be difficult to design because of moral hazard issues. Moral hazard SMEs employ a large share of the workers in is particularly salient among informationally developing economies. Ayyagari, Demirgüç- opaque enterprises that lack formal and reli- Kunt, and Maksimovic (2011a) report that able records because the insurance company formal SMEs account for about 50 percent will experience difficulty verifying whether of employees in developing countries (fig- and why the enterprises show low incomes. ure 3.3). They also find that SMEs create a In some industries, such as agriculture, moral greater share of net jobs relative to large firms hazard is mitigated if insurance payouts can even after they account for job destruction.10 be tied to publicly observable events, such as At the same time, the contribution of SMEs 116 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 3.3 Employment Shares of SMEs vs. Large Firms slower output growth, while other reported constraints are not as robustly associated with growth (Ayyagari, Demirgüç-Kunt, and 100 Maksimovic 2008). Within-country evidence Percentage of employees working in . . . 90 also points to credit constraints on SMEs. 80 45 50 An impact evaluation in India exploits varia- 70 60 55 tions in access to a targeted lending program 60 and finds that many SMEs are credit con- 50 strained and that providing additional credit 40 to SMEs can accelerate their sales and profit 30 55 growth (Banerjee and Duflo 2012). In addi- 50 20 40 45 tion, research in Pakistan shows that a drop 10 in subsidized export credit led to a substantial 0 decline in exports among small firms, but not Low Lower middle Upper middle High among large firms. Large firms were able to Country income level replace subsidized credit with credit at market Small and medium enterprises Large firms interest rates, but this was not true of small firms, indicating that small firms were, in fact, Source: Calculations based on Ayyagari, Demirgüç-Kunt, and Maksimovic 2011a. credit constrained (Zia 2008).12 Note: The year of observation varies by country, ranging from 2006 to 2010. Firm size categories are defined based on the number of employees: 5–99 refers to small and medium enterprises In examining how widespread credit con- (SMEs), and 100+ refers to large firms. Shown are the mean employment shares across countries straints are among SMEs, one should distin- within each income group. The data do not cover employment in firms with less than 5 employees or employment in informal firms. guish between firms that are voluntarily ver- sus involuntarily excluded from using loans. Figure 3.4 relies on data from the World Bank to productivity growth in developing econo- enterprise surveys to shed light on this issue. mies is not as high as that of large firms (for It shows the percentage of SMEs in low-, example, see Ayyagari, Demirgüç-Kunt, and middle-, and high-income countries that did or Maksimovic 2011a). Moreover, cross-country did not apply for a loan during the past year. research suggests that the existence of a large For SMEs that did not apply, it lists whether SME sector does not promote growth in per firms report they did not need a loan or were capita gross domestic product (GDP) (Beck, involuntarily excluded from applying for a Demirgüç-Kunt, and Levine 2005). One rea- loan. Among SMEs, 44 percent in low-income son for these findings may be that weak legal countries, 28 percent in middle-income coun- and financial institutions in developing coun- tries, and 20 percent in high-income countries tries prevent SMEs from growing into large were involuntarily excluded from applying for firms, and, so, a large SME sector coincides a loan. An even higher share of SMEs may with weak institutions that directly under- not have obtained a loan because firms that mine growth (Beck, Demirgüç-Kunt, and applied for a loan sometimes had their appli- Maksimovic 2005). It is thus crucial to take a cations rejected.13 The numbers for involun- closer look at the constraints that SMEs face. tary exclusion are thus a lower bound. Figure 3.5 provides the corresponding statistics on large firms. Compared with SMEs, a smaller Are SMEs credit constrained? share of large firms are involuntarily excluded Several cross-country studies find that the from applying for a loan: 25 percent in low- lack of access to finance is a key constraint to income countries and 14 percent in middle- SME growth in developing economies (Beck, and high-income countries. Demirgüç-Kunt, and Maksimovic 2005; Beck The reasons for involuntary exclusion from and others 2006).11 Cross-country research applying for a loan vary. Some SMEs may sim- analyzing 10,000 firms in 80 countries shows ply not benefit from investment opportunities that financing constraints are associated with that are sufficiently profitable to pay market GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 117 interest rates, or their projects may be too FIGURE 3.4 Voluntary vs. Involuntary Exclusion from Loan risky for banks to finance. Other SMEs may Applications, SMEs be credit constrained because they are subject to principal-agent problems (adverse selection and moral hazard) that are less salient among All small and medium enterprises large firms. SMEs often do not have adequate 100% records and accounts to document firm per- formance to apply successfully for a loan. Applied Did not apply for a loan This problem is compounded by the fact that Low: 25% Low: 75% Middle: 33% Middle: 67% some SMEs operate informally. Thus, they High: 33% High: 67% may not register with the government, or, if No need Involuntary exclusion they are registered, they may not declare all for a loan Low: 44% their revenue, implying that they have no or Low: 31% Middle: 40% Middle: 28% inadequate official proof of income. This lack High: 46% High: 20% of documentation may be one reason why Interest Application Collateral Other 12 percent of SMEs in low-income countries rates procedures requirements Low: 12% Low: 13% Low: 12% Low: 7% Middle: 9% state they did not apply for a loan because of Middle: 9% Middle: 5% Middle: 4% High: 10% High: 4% High: 2% High: 4% complex application procedures, that is, the application requires information that they cannot easily provide (see figure 3.4). To fill in Source: 2006–12 data from the Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. for missing information, the lender may ask Note: The figure includes data on 120 countries and relies on the most recent enterprise survey for additional collateral, but SMEs also often for each country. It covers only small and medium enterprises ( SMEs), that is, firms with 5–99 employees. The last row lists the main reason firms did not apply for loans. “Other” reasons for lack adequate collateral. In low-income coun- involuntary exclusion include “Did not think would be approved”; “Size of loan and maturity are tries, 7 percent of SMEs state that the major insufficient”; “It is necessary to make side payments”; and unspecified reasons. “Low, Middle, High” refer to low, middle, and high-income economies, respectively. reason they have not applied for a loan is high collateral requirements (see figure 3.4). Appli- FIGURE 3.5 Voluntary vs. Involuntary Exclusion from Loan cation procedures and collateral requirements Applications, Large Firms appear to be less of an obstacle for large firms (see figure 3.5). Only 5 percent of large firms in low-income countries report they did not apply for a loan because of complex applica- All large firms 100% tion procedures (and 3 percent say it was due to collateral requirements). Applied Did not apply In addition, high transaction costs can for a loan Low: 59% restrict access to credit among SMEs because Low: 41% Middle: 47% Middle: 53% the high fixed costs of financial transactions High: 47% High: 53% render lending to small borrowers unprofit- No need Involuntary exclusion able. The higher costs of lending to SMEs and for a loan Low: 25% Low: 34% the greater risks involved are often reflected in Middle: 39% Middle: 14% High: 39% High: 14% higher interest rates and fees for SMEs rela- tive to larger firms, particularly in developing Interest Application Collateral Other rates procedures requirements countries (Beck, Demirgüç-Kunt, and Mar- Low: 8% Low: 5% Low: 3% Low: 10% Middle: 6% Middle: 4% Middle: 2% Middle: 2% tínez Pería 2008a, 2008b). Figures 3.4 and High: 2% High: 1% High: 2% High: 9% 3.5 show that, compared with large firms, a greater percentage of SMEs did not apply for Source: 2006–12 data from the Enterprise Surveys (database), International Finance Corporation a loan because of high interest rates. and World Bank, Washington, DC, http://www.enterprisesurveys.org. Note: The figure includes data on 120 countries and relies on the most recent enterprise survey Overall, figure 3.4 illustrates that involun- for each country. It covers only large firms, that is, firms with 100+ employees. The last row lists tary exclusion for all reasons—high interest the main reason firms did not apply for loans. “Other” reasons for involuntary exclusion include “Did not think would be approved”; “Size of loan and maturity are insufficient”; “It is necessary to rates, difficult application procedures, and make side payments”; and unspecified reasons. “Low, Middle, High” refer to low, middle, and high- substantial collateral requirements—tends to income economies, respectively. 118 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 3.6 The Depth of Credit Information the lack of financial infrastructure to miti- gate these problems). Country studies show 4.5 4.2 that relieving the credit constraint can foster SME growth. The following subsections dis- 4 3.7 cuss private and public sector actions that can Depth of credit information index 3.5 relieve the credit constraint on SMEs. 2.9 3 2.5 Private sector initiatives to expand 2 SME lending 1.5 1.3 In some countries, banks have developed 1 innovative strategies for lending to SMEs. 0.5 In Argentina and Chile, for example, banks 0 often seek out creditworthy SMEs through Low Lower middle Upper middle High client relationships with large firms. They ask Country income level their large clients for references on their most dependable buyers and suppliers, which, in Source: 2012 data from Doing Business (database), International Finance Corporation and World many cases, are SMEs. Banks in Argentina Bank, Washington, DC, http://www.doingbusiness.org/data. and Chile perceive the SME sector as large, Note: The figure shows average values across all countries in each income group. The depth of credit information runs from 0 to 6; the higher values indicate the availability of more comprehen- unsaturated, and possessing good prospects sive credit information. The index measures rules and practices affecting the coverage, scope, and (de la Torre, Martínez Pería, and Schmukler accessibility of the credit information available through either a public credit registry or a private credit bureau. 2010b). This interest in SME lending seems to be at least in part an outcome of strong be more prevalent in low- and middle-income competition in other market segments, such than in high-income countries. This pattern as corporate and retail lending, that is, banks may reflect the fact that high-income coun- may make more of an effort to overcome tries have more financial sector infrastructure market failures related to SME lending if they to mitigate the principal-agent problems that are pushed to seek out new markets because SMEs face. For example, the World Bank’s of competition in existing markets.14 Doing Business database shows that coun- An interesting recent development is the tries with higher income levels have more fact that banks in emerging markets are comprehensive credit information available increasingly providing nonfinancial services through a credit reporting system (figure 3.6). to SMEs (IFC 2012b). For example, banks Other reasons why involuntary exclusion var- offer training and consulting services that can ies across countries include the structure of improve recordkeeping among SMEs, thus economies, competition in the banking sec- permitting banks to assess more easily the tor, and the extent of government borrowing. creditworthiness of these SMEs. In addition, An analysis of the obstacles to bank financing a majority of these nonfinancial services are among SMEs in African countries highlights managed by SME account managers, allow- that the share of SME lending in bank port- ing them to obtain detailed information about folios varies between 5 and 20 percent and the business, financial situation, and bank- that contributing factors are the structure and ing needs of the SMEs. This business model size of the economy, the extent of government mitigates information asymmetry problems borrowing, the degree of innovation intro- as banks gain more accurate information on duced by entrants to financial sectors, the SME loan applicants. For instance, the Turk- state of the financial sector infrastructure, and ish Bank TürkEkonomiBankası has success- the enabling environment (box 3.4). fully used nonfinancial services to expand In sum, a sizable portion of SMEs in its SME lending. Starting in 2005, the bank developing economies are credit constrained developed and implemented training, con- because of principal-agent problems (and sulting, and information-sharing services for GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 119 BOX 3.4 Financing SMEs in Africa: Competition, Innovation, and Governments Access to finance has been identified as the most a). According to the commercial bank surveys, the binding constraint on SME growth in Africa in share of SME lending in the overall loan portfolios the World Bank enterprise surveys. a Research on of banks ranges between 5 and 20 percent. While the the fi nancing of SMEs in Kenya, Nigeria, Rwanda, defi nitions of SMEs certainly differ across countries, South Africa, and Tanzania has sought to explore banks in Kenya, Rwanda, and even Tanzania seem this issue, analyzing both the supply and demand to be more involved with SMEs in terms of the share sides of SME fi nance (Berg and Fuchs 2013). The of loans going to SMEs relative to the corresponding supply-side studies consist of surveys of commer- shares at banks in Nigeria and South Africa (figure cial banks and other formal fi nancial institutions B3.4.1, panel b). This can partly be explained by to understand their involvement with SMEs and the size of bank lending portfolios, which is largest their business models for serving this segment. The in South Africa, implying that a lower share of the supply-side studies are comparable to previous World loan portfolio is still a sizable investment in SME Bank surveys (such as Beck and others 2008; de la lending. Apart from this, the reasons for the differ- Torre, Martínez Pería, and Schmukler 2010a; Rocha ences across countries vary, but key contributing fac- and others 2010; Stephanou and Rodriguez 2008; tors are the structure of the economy, the extent of World Bank 2007b, 2007c). In each country, banks government borrowing, the degree of innovation in were surveyed and interviews held with a majority SME lending models as practiced by domestic fi nan- of the respondents. For the demand-side surveys, cial intermediaries or foreign entrants, and the state either new data were collected to construct a panel of fi nancial sector infrastructure and the enabling of SMEs from the latest enterprise survey (Nigeria environment. and South Africa), or data from the enterprise sur- The five countries in which the SME finance stud- vey were used if recent data were available or still ies were undertaken—Kenya, Nigeria, Rwanda, being collected (Kenya, Rwanda, and Tanzania). South Africa, and Tanzania—differ in economic pro- Data from the demand-side surveys show that the fi le. South Africa is the largest economy in Africa. use of bank fi nancing among SMEs varies consid- Nigeria depends heavily on the oil and gas sectors. erably among the countries examined, though the Kenya has a reasonably well-diversified economy data refer to different years (figure B3.4.1, panel and a well-established layer of medium and larger FIGURE B3.4.1 Financing Small and Medium Enterprises in Africa a. Share of SMEs with loans b. Share of SME loans in total loans SMEs with bank loan or line of credit, % 70 18 17 17 59 16 60 14 48 14 % of all bank lending 50 12 SME lending, 40 10 8 30 8 23 6 5 20 10 12 4 10 2 0 0 Kenya Nigeria Rwanda South Africa Tanzania Kenya Nigeria Rwanda South Africa Tanzania Sources: Panel a: Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org; additional small and medium enterprise (SME) surveys (see the text). Panel b: SME finance surveys (see the text). (box continued next page) 120 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.4 Financing SMEs in Africa: Competition, Innovation, and Governments (continued) companies. Rwanda and Tanzania have considerably In South Africa, in contrast, four big banks domi- smaller economies. Commercial banks in Nigeria nate the fi nancial sector, collectively holding about focus their lending on the oil, gas, and telecommu- 80 percent of bank assets and covering the market nications sectors and the associated value chains, for (cheap) retail deposits. The market concentration while, in Rwanda and Tanzania, banks need to lend has affected innovation because competitors face to SMEs simply because of a lack of alternatives. difficulty growing in this market. The extent of government borrowing has led to Evidence from the five SME finance studies the crowding out of the private sector, especially in above suggests that competition, especially through Nigeria, but also in Tanzania, where banks continue innovators, can have a large impact on the frontier to hold a sizable proportion of their balance sheets lending of banks. While competition and innova- in government securities. Particularly in smaller tion cannot be introduced by government directive, fi nancial systems with weaker legal and regulatory encouraging innovation (for example, by allowing structures and capacity, there is a strong relationship agency banking) or a supportive regulatory environ- between the willingness of banks to lend to private ment for MFIs (to boost competition from within enterprises and the availability and yields of invest- the lower end of the market) is within the realm ment opportunities perceived as safer, such as gov- of possibility of a regulatory authority. Competi- ernment securities. tion is required to move commercial banks out of A strong determinant of the involvement of banks their comfort zone in countries such as Nigeria, with SMEs is the degree of competition in the mar- where high interest rates on government securities ket and the innovation introduced either by domes- provide a disincentive to intensify lending to SMEs. tic institutions or foreign entrants. Competition in In countries with lower yields on government secu- the SME market is strongest in Kenya, where a large rities, the incentives are much greater for banks to number of commercial banks target different market expand their lending to SMEs, and, if knowledge is segments. The difference between Kenya and most transferred in such circumstances, as was the case in other African countries is that innovation in Kenya Rwanda, SME lending expands. started through a combination of microfinance- While there seems to be a role for government to rooted institutions scaling up to become commercial encourage lending to SMEs in markets where that banks and innovative lending models and technology development has not yet taken place, providing an in the retail banking segment, most notably Equity environment conducive to lending seems to be cru- Bank. The innovations pioneered by Kenyan banks cial. Ensuring that an effective credit bureau is in have spread to other countries as banks expand operation and that the securitization and realization their footprint in the East African Community and of (movable) collateral are effi cient is a fundamen- beyond. The studies outlined above show that com- tal challenge in a number of countries. The reforms petition among banks for SME clients and retail in fi nancial infrastructure in Rwanda, for instance, customers has increased in Rwanda because of the have been appreciated as a positive development by entrance of Kenya-based banks in the market, while the banking sector and have led to an increase in the the availability of innovative distribution chan- use of movable assets to secure SME loans and to a nels such as agency banking has expanded as well. general expansion in lending to the sector. a. Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. SMEs with the goal of building a client base of of the bank’s SME clients rose from 20,000 healthy businesses, gaining new SME clients, in 2005 to 700,000 in 2011, and the share promoting customer loyalty, and reducing of SME loans in the total loans of the bank the credit risk in the SME sector. The number grew from 25 percent in 2006 to 44 percent GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 121 in 2011, while loan delinquency rates in the survey and concludes that state banks have bank’s SME portfolio declined.15 Driven by generally been inefficient in allocating credit the success in Turkey, BNP Paribas (one of because they often serve political interests the Turkish bank’s larger shareholders) repli- (World Bank 2012a). Focusing on the gover- cated some of the bank’s nonfinancial services nance of state-owned banks may help policy in Algeria and is now also seeking to replicate makers address the inefficiencies associated the model in European markets.16 with these institutions. However, governance Some banks take advantage of cross-selling reforms are particularly challenging in weak opportunities with SMEs by providing check- institutional environments. In addition, the ing accounts, transaction banking services, bulk of the empirical evidence suggests that and cash management services. Through the government ownership of banks in devel- these services, the banks develop a relation- oping economies has had negative conse- ship with the SMEs and learn about them, quences for long-run financial and economic which can facilitate lending to these SMEs in development. the future. For example, ICICI Bank in India Policy makers can encourage banks to lend currently derives most of its SME revenues to SMEs by taking on some of the credit risk through deposits and other nonlending prod- through guarantees either for a portfolio of ucts. However, its lending revenues are grow- loans or for individual loans. Both govern- ing quickly as the deposit-only clients begin to ments and international organizations offer take out loans (IFC 2010a). such risk-sharing arrangements. IFC provides risk-sharing facilities whereby IFC reimburses a bank for a portion of the principal losses Direct state interventions incurred on a portfolio of SME loans.17 For Despite the promising results in some coun- example, as part of the Global SME Finance tries, private sector action may not always Initiative, IFC set up a $22 million risk- be sufficient to lessen the credit constraint on sharing facility with Access Bank in Nigeria in SMEs, and policy makers often pursue SME December 2012. The project aims to increase financing policies to promote firm growth access to finance among SME distributors of and employment creation. Commonly used Coca-Cola’s Nigeria bottler. Eligibility for the policies include direct state intervention in risk-sharing facility is based on agreed crite- the form of directed credit, subsidies, or state- ria, and IFC does not review individual loans bank lending to SMEs. The success of these at origination. Other schemes guarantee indi- programs tends to be rare, but exceptions vidual loans. In these schemes, the guarantor, exist. For example, an impact evaluation in for instance, the government, pledges to repay India has analyzed a program that requires a fixed percentage of individual loan amounts banks to lend a share of their credit to small to the relevant bank in case of borrower firms. The study (Banerjee and Duflo 2012) default. The scheme administrator typically has found positive effects on the sales and reviews and approves individual credit guar- profit growth of the firms, namely, Re 1 of antee applications on a case-by-case basis.18 directed lending boosted profits before inter- Risk-sharing arrangements can increase est payments by Re 0.89. However, it is not lending to SMEs by lowering the amount of clear to what extent this finding applies to collateral that an SME needs to pledge to other contexts because the study examined receive a loan because the guarantor provides only one bank, which has been consistently part of the collateral. Similarly, for a given rated among the top five public sector banks amount of collateral, a credit guarantee can by a major business magazine. The Global allow more risky borrowers to receive a loan Financial Development Report 2013 dis- because the guarantee lowers the risk of the cusses the issue of state ownership in banks loan. and other state interventions more broadly In practice, a concern is that risk-sharing by conducting a comprehensive literature arrangements may not lead to additional 122 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 lending. Instead, banks may use guarantees Research and practitioner experience sug- to lower the risk on loans that they would gest that best practices for credit guarantee have issued even in the absence of the guar- schemes include (1) leaving credit assessments antees. Rigorous impact evaluations in Chile and decision making to the private sector, and France have assessed the extent to which (2) capping coverage ratios and delaying the risk-sharing arrangements lead to additional payout of the guarantee until recovery actions SME lending and whether this promotes firm are taken by the lender so as to minimize growth.19 moral hazard problems, (3) pricing guaran- In Chile, the Fondo de Garantía para tees to take into account the need for finan- Pequeños Empresarios (the guarantee fund for cial sustainability and risk minimization, and small businesses) is managed by a large public (4) encouraging the use of risk management bank (BancoEstado) that provides guarantees tools.20 However, many existing schemes for loans to small firms. The fund does not do not follow best practices, which likely evaluate guaranteed loans on a case-by-case explains their limited effectiveness in promot- basis, but sets broad eligibility criteria and ing SME lending. Saadani, Arvai, and Rocha lets participating banks decide which loans to (2011) review the design of credit guaran- guarantee. Two separate studies find that the tee schemes in the Middle East and North fund has generated additional loans for new Africa. They find that guarantee schemes in and existing bank clients (Cowan, Drexler, the region look financially sound, but tend to and Yañez 2009; Larraín and Quiroz 2006). focus on larger loans and are not yet reach- Moreover, the additional loans seem to have ing smaller firms. Most schemes have room to led to higher sales and profit growth among grow, but this growth should be accompanied the firms receiving the guarantees (Larraín by an improvement in key design and man- and Quiroz 2006). However, another study agement features, as well as the introduction questions whether the fund truly leads to of systematic impact evaluations. additional lending (Benavente, Galetovic, and In summary, directed lending programs Sanhueza 2006). It points out that approxi- and risk-sharing arrangements can have posi- mately 80 percent of the firms that participate tive effects on the access of SMEs to finance in the fund had had bank loans in the past and growth, but it is a challenge to design and that many of these firms had previously and manage such initiatives. These concerns received guarantees. are even greater in weak institutional environ- Lelarge, Sraer, and Thesmar (2008) study ments where good governance is difficult to the impact of the SOFARIS credit guaran- establish and SMEs face particularly severe tee scheme in France that reviews and cov- financial constraints. ers individual loans. The authors find that obtaining a loan guarantee lowers the cost Market-oriented government policies of capital among firms and helps the firms grow more rapidly. However, the loan guar- Policy makers can take other, more market- antee also increases the probability of default. oriented actions to promote SME lending. Higher default may be a reflection of the Such actions aim to improve financial sec- riskier, but potentially more profitable invest- tor infrastructure to mitigate the principal- ments made by the firms because of the loans. agent problems faced by SMEs. They include The investments may be beneficial from a (1) putting in place adequate movable col- policy perspective and could justify the loss of lateral laws and registries; (2) fostering the some guarantee amounts. On the other hand, availability of credit information by improv- guarantee schemes also lower the repayment ing corporate accounting and by supporting incentives of firms, and the schemes need to information sharing among various actors, be designed carefully and managed effectively including banks, utility companies, and sup- to prevent large-scale losses. pliers; and (3) strengthening the legal, regula- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 123 tory, and institutional infrastructure for fac- can verify their priority through an electronic toring and leasing. archive of security filings (Fleisig, Safavian, and de la Peña 2006).22 Priority rules work best if a country has Movable collateral laws and registries a single registry for pledges of collateral so To compensate for missing information on the that prospective lenders can easily establish creditworthiness of SMEs, lenders may ask whether there is a prior claim on an asset for collateral to guarantee a loan. Data from (Fleisig, Safavian, and de la Peña 2006). A the World Bank enterprise surveys show that recent study using enterprise surveys for 73 about 79 percent of loans or lines of credit countries finds that introducing movable col- require some form of collateral.21 Movable lateral registries increases the access of firms assets, as opposed to fixed assets such as land to finance (box 3.5). There is also some evi- or buildings, often account for most of the dence that this effect is larger among smaller capital stock of firms, particularly SMEs. For firms. Overall, sound collateral laws and reg- example, in the developing world, 78 percent istries can allow firms to use their own assets of the capital stock of businesses is typically to guarantee loans and may reduce the need in movable assets such as machinery, equip- for publicly sponsored guarantee schemes. ment, or receivables, and only 22 percent is in immovable property (Alvarez de la Campa Credit information 2011). Yet, banks are often reluctant to accept movable assets as collateral because of non- An additional crucial component of the finan- existent or outdated secured transaction laws cial infrastructure that can support SME and collateral registries. Many legal systems financial inclusion is credit information. Such place unnecessary restrictions on creating col- information encompasses any data that can lateral, leaving lenders unsure whether a loan help a lender decide whether a firm is credit- agreement will be enforced by the courts. For worthy. It can be used to generate credit scores example, about 90 percent of the movable predicting repayment on the basis of borrower property that could serve as collateral for a characteristics. In the United States, the use of loan in the United States would likely be unac- credit-scoring technology for small business ceptable to a lender in Nigeria (Fleisig, Safa- loans has led to an expansion in the availabil- vian, and de la Peña 2006). ity of loans for small and riskier firms even by Reforming the movable collateral frame- larger banks that would otherwise have shied work may enable firms to leverage their assets away from this segment (Berger, Frame, and to obtain credit. Canada, New Zealand, and Miller 2005). Some countries, such as India, the United States were the first to reform the have introduced credit-rating agencies that legal system governing the use of movable focus specifically on small firms (GPFI 2011f). property as collateral and now have among The SME Rating Agency of India Limited pro- the most advanced systems. Some developing vides credit ratings for microenterprise and countries have also successfully reformed these SME loan applicants for a fee (equivalent to systems, including Afghanistan, Albania, Bos- roughly $900, with variations depending on nia and Herzegovina, China, Ghana, Mexico, turnover). The Indian government covers a Romania, and Vietnam. The reformed systems large part of this fee to subsidize the use of have three common features: (1) laws do not credit scoring. impose limits on what can serve as collateral; Credit information can be supplied from (2) creditors can seize and sell collateral pri- numerous sources, including firm financial vately or through summary proceedings, dra- statements, credit registries or bureaus, and matically reducing the time it takes to enforce supplier networks. Firm financial statements a collateral agreement; and (3) secured credi- and official documentation are essential parts tors have first priority to their collateral and of loan applications at many banks, but the 124 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.5 Collateral Registries Can Spur the Access of Firms to Finance To reduce the asymmetric information problems asso- movable assets against three control groups: fi rms ciated with extending credit and to increase the prob- in all countries that did not implement collateral ability of loan repayment, banks typically require reform, fi rms in a sample of countries matched by collateral from their borrowers. Movable assets are location and income per capita with the countries the main type of collateral that firms, especially that introduced movable collateral registries (fig- those in developing countries, can pledge to obtain ure B3.5.1), and fi rms in countries that introduced fi nancing. However, lenders in developing countries other types of collateral reforms, but did not set up are usually reluctant to accept movable assets as col- registries of movable collateral. Overall, they fi nd lateral because of the inadequate legal and regulatory that the introduction of movable collateral regis- environment in which banks and fi rms coexist. Spe- tries increases the access of fi rms to fi nance. There cifically, three conditions are required before banks is also some evidence that this effect is larger among are able to accept movable assets as collateral: the smaller fi rms. creation of security interest, the perfection of secu- rity interest, and the enforcement of security interest (Fleisig, Safavian, and de la Peña 2006). The movable FIGURE B3.5.1 Effect of Collateral Registry Reforms collateral registry is a necessary component because on Access to Finance it allows for the perfection of security interest, mean- 80 73 ing that security interest becomes enforceable with 70 Firms with access to finance, % regard to other creditors or third parties rather than merely between the lender and the borrower. Specifi- 60 54 50 cally, the registry fulfi lls two essential functions: to 50 41 notify parties about the existence of a security inter- 40 est in movable property (existing liens) and to estab- 30 lish the priority of creditors with respect to third par- ties (Alvarez de la Campa 2011). Therefore, without 20 a well-functioning registry of movable assets, even 10 the best secured transaction laws could become com- 0 pletely ineffective. Prereform Postreform Using fi rm-level surveys for up to 73 countries, Registry reformers Nonreformers Love, Martínez Pería, and Singh (2013) explore the (matched by region and income) impact of the introduction of collateral registries Source: Doing Business (database), International Finance Corporation and of movable assets on the access of fi rms to fi nance. World Bank, Washington, DC, http://www.doingbusiness.org/data; Enter- prise Surveys (database), International Finance Corporation and World Bank, They compare the access of fi rms to fi nance in seven Washington, DC, http://www.enterprisesurveys.org; calculations by Love, countries that introduced collateral registries of Martínez Pería, and Singh 2013. quality and reliability of these statements vary fewer resources and capabilities than larger across countries and firms. In an effort to stan- firms for the preparation of financial state- dardize financial reporting, the IFRS Founda- ments (IASB 2012). About 60 countries have tion develops rigorous international financial adopted the SME standards. However, some reporting standards. Recognizing that the full countries have adopted a simpler set of oblig- standards, as well as many national generally atory standards because they have concluded accepted accounting principles are complex, that the SME standards of the IFRS Founda- the foundation issued self-contained global tion are too costly and burdensome for local standards for SMEs in 2009. These SME stan- and small firms (GPFI 2011f). SMEs often dards focus on concepts that are most relevant lack sufficient technical knowledge and capac- for helping SMEs gain access to capital, such ity to prepare sound financial statements. as cash flows, liquidity, and solvency. They Nonetheless, business development services also take into account that SMEs often have may help to build capacity in this area. Two GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 125 recent studies show that management con- Another valuable source of credit infor- sulting services can improve accounting and mation is buyer-supplier relationships. Buyer recordkeeping among SMEs (Bloom and oth- or supplier credit, also called trade credit, is ers 2013; Bruhn, Karlan, and Schoar 2013). an important channel of external financing Regulatory reforms that encourage informal for firms (Demirgüç-Kunt and Maksimovic firms to register with the authorities can also 2001). For example, suppliers often allow lead to better information on SMEs. buyers to pay for goods a number of months Credit registries and credit bureaus pro- after delivery. In this way, large suppliers with vide records of past and current loans taken sufficient liquidity and access to loans can act out by firms.23 These records can help lenders as a financial intermediary for firms that are observe whether loans have been repaid suc- small or credit constrained (Marotta 2005; cessfully and also whether firms have other McMillan and Woodruff 1999). Trade credit liabilities that may make them risky borrow- has been shown to act as a substitute for bank ers. Cross-country evidence confirms that the credit during periods of monetary tightening availability of detailed information about the or financial crisis (Choi and Kim 2005; Love, borrowing and repayment behavior of pro- Preve, and Sartia-Allende 2007). spective clients places banks in a better posi- Suppliers often have detailed records on tion to assess default risk, counter adverse how diligent buyers are in repaying trade selection, and monitor institutional exposure credit. An innovative credit information to credit risk (Jappelli and Pagano 2002; scheme in Peru uses a large technology plat- Miller 2003; Pagano and Jappelli 1993). form to process and analyze repayment data Cross-country research also shows that the held by suppliers. The platform generates presence of credit bureaus is associated with payment reports—similar to those provided lower financing constraints and a higher by a credit bureau, with proof of historical share of bank financing among SMEs, as fulfillment of payments—on the firms that well as with a higher ratio of private credit to participate in the scheme. Firms can then use gross domestic product (Djankov, McLiesh, these independent payment reports when they and Shleifer 2007; Love and Mylenko 2003). approach banks, improving their chances of In addition, using firm-level survey data from receiving bank credit.25 In addition to pro- 24 transition economies, Brown, Jappelli, and viding repayment histories, the project also Pagano (2009) find that information shar- encompasses a factoring scheme that can ing by banks is associated with the enhanced channel more supplier credit to firms. availability and lower cost of credit to firms. This correlation is stronger for opaque firms Insolvency regimes than transparent ones and stronger in coun- tries with weak legal environments than in Insolvency regimes are a key aspect of finan- those with strong legal environments. Finally, cial infrastructure. They can promote access Beck, Lin, and Ma (2010) show that more to finance among SMEs by supporting pre- effective credit information sharing is associ- dictability in credit markets. An effective ated with lower tax evasion (and thus infor- insolvency framework can help regulate effi- mality), particularly among smaller firms cient exits from the market, ensure fair treat- and firms requiring more external finance. ment through the orderly resolution of debts The Credit Reporting Knowledge Guide, incurred by debtors in financial distress, and “The General Principles for Credit Report- provide opportunities for recovery by bank- ing,” and the Global Financial Development rupt entities and their creditors (Cirmizi, Report 2013, the Credit Reporting Knowl- Klapper, and Uttamchandani 2012). edge Guide, provide detailed information on Many countries have substantial legal gaps credit reporting institutions and actions that such that insolvency frameworks are unable governments can take to foster the develop- to deal with SMEs effectively (IFC 2010b). ment of these institutions (IFC 2012a; World If legal frameworks are absent, SMEs can be Bank 2011a, 2012a).24 negatively affected in a number of ways. First, 126 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 SMEs that are fundamentally viable, but face porting payment services for SMEs has gar- short-term liquidity crises have no safety net. nered growing interest. Among individuals, The SME facing financial distress cannot seek payments are an entry point into the formal temporary protection from its creditors, can- economy and potentially regulated finan- not propose a plan of reorganization, and cial services. Firms have specific needs that cannot compromise debt to achieve greater should be taken into account in the design returns to all creditors. Second, in the event of payment systems and instruments, and of liquidation, SMEs would find it difficult to they depend closely on banks for payment go through an orderly and transparent pro- solutions. An interesting question that has cess to repay creditors and return productive emerged in the early policy discussions is assets into the economy as quickly as possible. whether this makes SMEs a captive market Third, if an SME fails, its outstanding obliga- or whether they can seek alternative solutions tions will be the obligations of the individual elsewhere, for example, in software compa- entrepreneur, in perpetuity, unless specifically nies. The migration to electronic payments is forgiven by creditors. The absence of effective an established trend in the corporate world. exit mechanisms can thus lock the produc- Key benefits seem to be standardization (that tive assets of SMEs in a legal limbo, making is, adopting common formats that can be pro- reentry into the marketplace problematic and cessed automatically to execute instructions inhibiting entrepreneurship. and provide ancillary data), greater efficiency, An efficient and modern framework for and savings in time and resources (according SME insolvency should include fast-track, to evidence in the World Bank’s Doing Busi- expedited bankruptcy provisions in unified ness database).26 However, SMEs are lagging or corporate bankruptcy laws. It should in the paperless trend. Scale, cultural, and provide alternative dispute resolution frame- cost factors could limit the use of electronic works, such as mediation and conciliation, solutions by SMEs, as well as regulatory to improve efficiency. Legislation on SME aspects. In some cases (Italy is an example), insolvency should specify a clear and trans- government legislation plays an active role in parent process that entrepreneurs can use promoting the adoption of new technologies to rescue their troubled businesses. This can such as e-invoicing. include stays on proceedings by creditors and the ability of SMEs to propose restructuring Factoring plans to creditors. In the event of business failure, the legislation should set out a clear Take the case of an SME that provides goods method for liquidating the business, repaying or services to a large buyer, but the large creditors in a timely manner, and discharging buyer only pays 30 to 90 days after deliv- the remaining debt. There should be estab- ery, while the SME needs working capital lished punishments for fraudulent activities, throughout the production cycle. Reverse such as the act of dissipating the assets of factoring schemes, such as NAFIN (Nacional a business so that creditors cannot recover Financiera) in Mexico, allow SMEs to address their claims and for negligently incurring this problem by transferring outstanding bills obligations from creditors if the entrepreneur to a factor. The factor pays the SME immedi- knew, or should have known, that the busi- ately (at a discount) and later collects the full ness was insolvent. These critical protections outstanding amount directly from the large for creditors can help ensure that SMEs con- buyer (Klapper 2006). Because, in this case, tinue to be able to access credit at reasonable the large buyer is the liable party, the factor rates (IFC 2010b). can issue credit at better terms than it would grant if the more risky SME were the bor- rower. The scheme in Peru described in box Payment services 3.6 works differently. There, a large supplier Although data are lacking, and discussion on pools the invoices from many small buyers. this topic is at an early stage, the issue of sup- Risk pooling allows the package of bills to be GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 127 BOX 3.6 Case Study: Factoring in Peru The World Bank is currently implementing a fac- ogy platform to purchase accounts receivable from toring scheme in Peru (fi gure B3.6.1). Factoring is large companies that supply many micro and small a fi nancial transaction in which a fi rm sells its cred- enterprises (MSEs). Doing so frees up working capi- itworthy accounts receivable to a third party, the tal that the suppliers can use to extend more credit factor, at a discount (equal to interest, plus service to MSEs, helping to solve the problem of access fees) and receives immediate cash. The World Bank to fi nance. It also benefi ts the suppliers and other scheme uses a financial structure and a technol- stakeholders. FIGURE B3.6.1 Actors and Links in the Financing Scheme, Peru Goods and services Financing Cash Development banks MSEs Suppliers Fiduciary agents Capital markets Accounts Discounted accounts Financial receivable receivable instrument Technology platform Because of the close relationship between large build credit histories that are valuable when they suppliers and their MSE customers, the suppliers can approach other financial intermediaries. Large provide a substantial amount of credit to MSEs at suppliers benefit from transferring a portion of relatively low risk and low cost. The accounts receiv- their accounts receivable portfolio to a third party, able portfolios of these large suppliers are diversified improving their fi nancial ratios, that is, factoring and carry low risk, qualities that are the building improves the liquidity of suppliers by substituting blocks of the fi nancing scheme. cash for accounts receivable. The suppliers can use The scheme, which has been developed with this additional (off-balance-sheet) fi nancing to raise fi nancing from FIRST (Financial Sector Reform and the sales of their products and services by provid- Strengthening Initiative), involves the creation of a ing additional credit to their MSE customers with- fiduciary agent, such as a trust, to purchase accounts out negatively affecting working capital. In addition, receivable from corporate suppliers on a revolving because of the guarantee structure, the financing basis. During the initial implementation in Peru, cost implicit in the factoring scheme will usually be the local development bank, COFIDE, will fund lower than traditional bank fi nancing. Moreover, the trust. Once the scheme is implemented, the fidu- reducing the cost of funding will boost the rate of ciary agent can raise funds on capital markets. The return on assets among suppliers. Finally, the system scheme minimizes the risks involved in the transac- will help suppliers achieve better risk management tion through its fi nancial structure and the technol- of their client MSEs. ogy platform used to manage the flow of receivables. The World Bank, COFIDE, and Capital Tool Cor- Invoice factoring has several benefits for MSE poration, working directly with structuring, legal, customers. First, participating MSEs receive more and tax advisers, have implemented a pilot version financing from suppliers, which they can use to of this factoring scheme in Peru. In the transaction, increase sales. In addition, the scheme helps MSEs COFIDE and Axur (a local MSE supplier)—through obtain better credit terms elsewhere because they a fiduciary agent—created a special-purpose vehicle, (box continued next page) 128 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.6 Case Study: Factoring in Peru (continued) which issued a term note that COFIDE purchased, 25,000 client MSEs, for a total funding amount of providing $5 million in capacity for fi nancing MSEs. $30 million–$40 million. The vehicle used the proceeds to purchase prese- This factoring scheme was one of 14 winners lected accounts receivable from Axur on a revolv- of the G-20 SME challenge award for fi nding new ing basis. The transaction would extend fi nancing ways to fi nance MSEs and received a grant to ex- to approximately 1,000 MSEs, discounting invoices pand the scheme. The solution also attracted interest of $500, on average, and 21 days maturity. In sub- from other multilateral lending agencies. The Inter- sequent transactions, the rating agencies will de- American Development Bank has approved a grant termine the risk of the financial instrument to be to expand the scheme by incorporating the portfolio issued by the vehicle. This rating would allow the of local MFIs. Colombia and Paraguay have also vehicle to sell participations in the financing to local expressed interest in the scheme. institutional investors. At that time, it could cover associated with financing at better terms than agreed rate of interest. Leasing thus focuses each individual bill, providing additional liq- on the ability of firms to generate cash flows uidly to the large supplier that can be passed from business operations to service leasing on to the small buyers. payments, rather than on the credit history Factoring is used in developed and devel- of firms or their ability to pledge collateral oping countries around the world, but it (Fletcher and others 2005). The ownership of requires an appropriate legal framework the equipment is often transferred to the firms (GPFI 2011f; Klapper 2006). For instance, at the end of the lease period. the law should allow firms to transfer their Brown, Chavis, and Klapper (2010) show receivables to factors, giving factors the right that close to 34 percent of firms in high- to enforce payment without consent of the income countries use leasing, compared with firm. The Legislative Guide on Secured Trans- only 6 percent in low-income countries. They actions of the United Nations Commission on also find that a strong institutional environ- International Trade Law (UNCITRAL 2010) ment is associated with the greater use of includes detailed recommendations on how to leasing. Fletcher and others (2005) discuss set up a legal framework that is amenable to different variations of leasing and provide factoring transactions. a manual on leasing legislation, regulation, and supervision based on international best practices and IFC’s technical assistance expe- Leasing rience (see also GPFI 2011f for more infor- Another financial product that can improve mation on standards, guidelines, and good access to finance among SMEs is leasing practices). (Berger and Udell 2006). Leasing provides In summary, mounting evidence indicates financing for assets, such as equipment and that movable collateral frameworks and reg- vehicles, rather than direct capital. Leas- istries, as well as credit information systems, ing institutions purchase the equipment and can increase lending to SMEs. Factoring and provide it to firms for a set amount of time. leasing are two alternative ways of channel- During this time, the firms make periodic ing financing to SMEs. Currently, there is lit- payments to the leasing institution, typically tle evidence documenting the impact of these covering the cost of the equipment and an two financial instruments on SME invest- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 129 ment and growth. More research is needed FIGURE 3.7 Average Loan Term in this area. 70 61 60 60 Dealing with risk 50 44 42 42 So far, the discussion has largely focused on 37 38 40 Months bank credit for SMEs, but some investments 29 30 25 26 that SMEs make may require other sources of 21 17 financing, such as private equity. In particu- 20 lar, banks may be reluctant to finance risky 10 investments. This reluctance may not arise 0 from the creditworthiness of the SMEs, but Low Lower middle Upper middle High from the probability of failure of the risky, Country income level but potentially profitable investments. For example, banks may be reluctant to lend to Small firms Medium-size firms Large firms a firm that considers an investment with an 80 percent chance of success and a 20 per- Source: Data for 2006, 2007, and 2009 from Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. cent chance of failure, especially if the firm Note: The data cover 23 countries. It shows results based on the most recent enterprise survey for has limited liability. If the firm does not have each country. limited liability, the owner will be reluctant to invest because the owner will lose personal assets in the event the project fails. This prob- age loan maturity is only 17 months for small lem may be salient in countries in which other firms in low-income countries, compared mechanisms for dealing with risk, such as with 37 months for large firms in low-income bankruptcy protection, are weak. Innovative countries, and 61 months for small firms in insurance products may provide one way of high-income countries. mitigating the investment risk affecting firms. Despite the potential benefits, private equity investment and risk-sharing arrange- ments may not become more common because Private equity for SMEs they can also lead to moral hazard. With risk Private equity financing can encourage firms sharing, the SME owner has less incentive to to make risky investments because it allows ensure that investments are successful because the firm to share the risk with a private equity the SME’s stake in the investment is less than investor. Another advantage of private equity the full amount. In fact, in some developing is that it can provide financing that is longer countries, equity investments in SMEs are term relative to loans, particularly for riskier supplied by friends and family, who may have and more opaque borrowers, because banks good information about the SME owner’s often offer such borrowers shorter-term loans, intentions and actions and who can rely on which then need to be renewed or renegoti- the personal relationship as an enforcement ated. Demirgüç-Kunt and Maksimovic (1999) mechanism (for example, see Allen and oth- use data from 30 developed and developing ers 2006). However, not all SMEs have access countries during 1980–91 to show that small to sufficient funding from friends and family. firms have less long-term debt as a propor- This is a case where private equity investors tion of total assets and total debt compared can step in. To mitigate moral hazard, private with larger firms. They also find that firms equity investors rely on contracting contin- in developed countries hold a greater pro- gencies and securities that shift control rights portion of their total debt as long-term debt depending on the performance of the invest- compared with firms in developing countries. ment. Because these contracts require sound Data from the World Bank enterprise surveys legal enforcement, private equity financing is show a similar pattern (figure 3.7). The aver- more likely to flourish in countries with strong 130 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 enforcement mechanisms (Lerner and Schoar investments in private equity funds in emerg- 2005). Governments can thus promote the ing markets in the form of senior secured formation of a private equity industry by put- loans. These types of loans are senior to all ting in place a strong legal framework that other claims against the borrower, which allows for efficient contract enforcement. means that, in case the borrower goes bank- Private equity investors often supply more rupt, the senior secured loan is the first to than funding to the firms they support. They be repaid before all other interested parties rely on extensive industry experience to offer receive repayment. Since 1987, the corpora- market knowledge and back-office services tion has committed $4.4 billion to 63 pri- that can help firms make large changes in vate equity funds in emerging markets. These their business. This market knowledge and funds have invested $5.6 billion in more than expertise may be scarcer in developing coun- 570 firms across 65 countries. tries, which is another reason why private IFC recently launched the SME Ventures equity investment may be more difficult to Project to provide risk capital and support to find in these economies. SMEs in International Development Associa- Some governments and international orga- tion countries. Under the program, IFC pro- nizations have taken steps to address the vides private equity funds that are managed shortfalls in private equity investment in through independent investment managers, developing countries. For example, a major who are selected on a competitive basis. The attempt to stimulate the private equity indus- program thereby helps develop the capacity try in Nigeria was the Small and Medium of investment managers to invest risk capi- Industry Equity Investment Scheme, which tal successfully in small businesses in these required banks to set aside 10 percent of their countries. Capacity building is a crucial com- pretax profits for equity investments in SMEs. ponent of the project. IFC provides financing Recently, the Nigerian government also and technical assistance to fund managers in changed pension regulations to allow up to areas such as partial support for start-up and 5 percent of pension assets to be invested in operational costs, legal structuring and reg- private equity. However, some of these funds istration, and capacity building among new have remained uninvested in part because of a staff. SME Ventures also offers advisory ser- high prevalence of fraud and a lack of SME vices to the SME business community. SMEs capacity and transparency. In case studies selected for private equity investments receive conducted by the World Bank, a private tailored business support to prepare them for equity fund manager pointed out that his staff the investment, as well as during the life of visits the businesses they invest in at least the investment. These services include busi- once a week and calls them every day to stay ness planning, market research, governance, as involved in the business as possible and management information and accounting sys- guard against fraud (Berg and others 2012). It tems, and upgraded environmental and social was also mentioned that most Nigerian entre- standards. preneurs do not understand private equity as One of the funds created through SME a source of financing for growing their busi- Ventures is the West Africa Fund for Liberia ness, indicating a need for training and capac- and Sierra Leone. It currently has an approved ity building. A case study in Rwanda uncov- investment portfolio of 19 projects worth ered similar challenges and also highlighted $7.4 million in different sectors, including that many entrepreneurs are reluctant to cede food processing, transportation, construction, ownership to external private equity investors health, and light manufacturing. To help man- (Abdel Aziz and Berg 2012). agers identify investment opportunities, the Another example from a developed econ- IFC advisory services team conducted market omy is the U.S. Overseas Private Investment surveys and identified more than 240 high- Corporation, which is the U.S. government’s potential SMEs in Liberia and Sierra Leone. development finance institution.27 It makes In a first phase, 60 of these SMEs developed GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 131 business plans, and the 20 best plans were finance is not catching up to the access experi- submitted to the West Africa Fund, facilitat- enced by larger firms. ing five investment appraisals. In a second One reason why more SMEs do not list on phase, the remaining high-potential SMEs are stock markets is that the fixed costs of an ini- also likely to develop business plans, which tial public offering (IPO), as well as the costs can lead to more appraisals and investments. implied by ongoing reporting requirements, In conclusion, SMEs that are looking to can be high. In some cases, stringent regula- make risky investments with potential high tions that are intended to safeguard the inter- returns may not be able to obtain bank loans ests of investors may even debar SMEs from to fund these projects. Private equity inves- listing on large stock exchanges altogether tors can step in to share the risk and poten- (Nair and Kaicker 2009). Some countries tial rewards. However, these types of invest- have created secondary trading exchanges ments rely on sound legal systems for contract with regulations and requirements specifi- enforcement and previous business expertise, cally adapted to smaller firms, such as AIM which may be lacking in developing countries. in London, NASDAQ in New York, and International agencies have launched projects AltX in South Africa. For example, AIM was to help develop local private equity industries. launched in 1996, and more than 3,000 firms This is another area where research can help from across the globe have since listed on the to document the extent to which these proj- exchange. A crucial component of AIM is the ects promote SME growth. use of nominated advisers (Nomads), com- panies that specialize in corporate finance. Each firm seeking admission to AIM must Can stock markets alleviate the work with a Nomad throughout the period financing constraints on SMEs? of listing on AIM. Nomads provide advice Stock markets can be a potential source of and guidance to firms. They also have a qual- financing for some SMEs, but they are asso- ity control function: they must periodically ciated with a number of challenges. Accord- report information on the firm to AIM. AltX ing to a 2010 survey of the World Federation in South Africa uses a similar system. Each of Exchanges, 44 percent of all listed com- company that lists on the exchange needs to panies are microcaps, which are defined as obtain a designated adviser who advises the companies with market capitalization below company on its responsibilities during the list- $65 million, but these companies account ing process and on its duty to maintain its sta- for only 1 percent of total market capitaliza- tus once listed. tion (WFE 2011). Compared with the num- There are also instances in which SME ber of SMEs in the economy, the number of exchanges have not been successful. The Over listed microcaps is small. For example, in the the Counter Exchange of India was estab- European Union in 2008, there were 20 mil- lished in 1992 as a platform where SMEs lion SMEs (defined here as enterprises with could generate equity capital, but it only had fewer than 250 employees), of which about 60 listed companies as of March 31, 2012. 220,000 were medium enterprises (between One factor that has prevented the growth 100 and 250 employees). The number of of the exchange is the lack of institutional listed microcaps in the European Union in participation (Nair and Kaicker 2009). Insti- 2010 was less than 4,000, representing only tutional investors, such as pension funds, about 1 percent of medium enterprises and a hold a large fraction of shares in many stock tiny fraction of all SMEs. Between 2007 and exchanges, and they tend to invest in rela- 2010, the number of microcaps changed at tively large firms. Even AIM is dominated by rates that were similar to the change in the institutional investors, who hold more than number of larger listings (increasing in the 60 percent of AIM-listed companies and also Americas and declining in Europe), suggest- focus on the largest companies listed on AIM, ing that the access of SMEs to stock market that is, the demand of the investors for stocks 132 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 in small companies (microcaps) may be more gests that (1) start-ups and surviving young limited than their demand for stocks in large businesses are critical for job creation, and companies (Nair and Kaicker 2009). In fact, (2) there is no systematic relationship between microcaps tend to be illiquid, meaning that firm size and employment growth after one the number of trades per listed company is controls for firm age. relatively small compared with larger caps. These results do not necessarily apply in Calculations based on data of the World Fed- developing economies, where firms, especially eration of Exchanges suggest that enterprises small ones, face many institutional constraints. with market capitalization below $200 mil- Thus, research shows that firm dynamics lion made up 64 percent of the world’s listed in India and Mexico are different from the companies in 2011, but accounted for only dynamics in the United States: firms grow 14 percent of individual stock market trades much more slowly as they age in India and and 4 percent of share trading volume. Simi- Mexico compared with firms in the United larly, a study of listed firms in China and States (Hsieh and Klenow 2012). Ayyagari, India shows that larger firms are more likely Demirgüç-Kunt, and Maksimovic (2011a) than smaller firms to issue new equity, and study the relationships across firm size, age, the top 10 issuing firms capture a large frac- and growth using data for 99 economies from tion of the total amount raised through these the World Bank enterprise surveys. They find issues (Didier and Schmukler 2013). that SMEs and young firms exhibit higher job Overall, the extent to which stock mar- creation rates than large and mature firms, kets can be a viable and sustained source of showing that size is still a good predictor of finance for SMEs is thus not clear. In addi- employment growth after they have controlled tion, even exchanges such as AIM tend to for age. However, large firms and young firms focus on the most rapidly growing SMEs so exhibit higher productivity growth. Overall, that stock markets may not be a financing these findings suggest that it is important to solution for a broad range of SMEs. How- focus on promoting finance among both SMEs ever, stock markets can potentially have posi- and young firms in developing countries.29 tive spillover effects on SMEs; for example, if more large firms obtain financing through The financing challenges faced by stock markets, they may need less financing young firms from banks, and banks may expand their SME lending operations. Similar to microenterprises and SMEs, young firms may be credit constrained because of principal-agent problems. Because young YOUNG FIRMS: CAN FINANCE firms have not been in the market for long, PROMOTE ENTREPRENEURSHIP? there is little information on their perfor- While a good deal of policy and research mance or creditworthiness. Data from the attention has been directed toward the finan- World Bank enterprise surveys on more than cial inclusion of SMEs, greater emphasis is 70,000 firms in more than 100 countries show needed on the financing available to young that, in all countries, younger firms rely less on firms. There are good reasons to focus on bank financing and more on informal financ- SMEs: the evidence shows that a sizable share ing from family and friends (figure 3.8). Only of the SMEs in developing countries are credit 18 percent of firms that are 1 or 2 years old constrained; alleviating these constraints may used bank financing, compared with 39 per- allow these firms to grow. Policy makers are cent of firms that are 13 or more years old. also often concerned with job creation, and In contrast, 31 percent of firms that are 1 or SMEs have traditionally driven job creation in 2 years old rely on informal sources of finance, developed countries.28 However, more recent such as family and friends, whereas only empirical work, such as the research by Halti- 10 percent of firms that are 13 or more years wanger, Jarmin, and Miranda (2010), sug- old obtain financing from informal sources GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 133 (Ayyagari, Demirgüç-Kunt, and Maksimovic FIGURE 3.8 Financing Patterns by Firm Age 2011a). Potential problems with financing from friends and family include that it might 45 working capital or new investments, % Firms that use the financing source for be unreliable, untimely, or bear significant 40 39 nonfinancial costs (Djankov, McLiesh, and 35 31 31 32 Shleifer 2007). For example, a study of Chi- 29 30 26 nese firms finds that more firms use informal 25 21 20 credit than bank credit, but only bank credit 20 18 16 16 15 is associated with higher growth rates (Ayya- 15 12 10 gari, Demirgüç-Kunt, and Maksimovic 2010). 10 A study in the United States estimates the 5 impact of formal credit on young firm survival 0 by comparing start-ups that have received a 1–2 3–4 5–6 7–8 9–10 11–12 13+ loan from a financial institution as a result of Firm age, years falling slightly above a threshold in the appli- Bank financing Informal finance cation criteria with start-ups that fell slightly below this threshold (Fracassi and others Source: Chavis, Klapper, and Love 2011 based on 1999–2006 data on 170 cross-sectional surveys 2013). The results show that the start-ups that in 104 countries in Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. have obtained a loan are 40 percentage points Note: The figure shows the share of firms that use the financing source for working capital or new more likely to survive, enjoy higher revenues, investment. Informal sources of finance include family and friends. and create more jobs. Policies and innovative approaches to preneurs. For example, entrepreneurs place a support lending to young firms higher value on work, are happier, and per- Some of the interventions discussed in the sec- ceive themselves as more successful. tion on SME financing (see above) can also The Entrepreneurial Finance Lab has be used to improve the access of young firms developed a tool for screening entrepreneurs to bank loans. Adequate collateral laws and that can potentially help lenders provide registries can allow young firms to leverage financing to borrowers who lack financial the assets they have to obtain credit, although documentation or credit histories. The tool they may have accumulated comparatively uses questions on character, abilities, and fewer assets than older firms. Credit infor- attitude to identify high-potential, low-risk mation systems can document the creditwor- entrepreneurs. The answers to these ques- thiness of young firms and can reduce the tions are combined into a psychometric credit reliance of young firms on informal finance score. Entrepreneurs who scored in the top 30 (Chavis, Klapper, and Love 2011). percent of this score had a 72 percent lower Accessing finance can be particularly chal- default rate than entrepreneurs who scored in lenging for entrepreneurs who have a business the middle 40 percent.30 Those with the low- idea, but have not yet started their company, est scores had a 94 percent higher default rate because they do not possess business assets than entrepreneurs in the middle range. The that can be used as collateral, and they do Entrepreneurial Finance Lab is a G-20 SME not have financial statements. One poten- Finance Challenge winner. Its screening tool is tial avenue for fostering financial inclusion already being introduced by some banks, for among these entrepreneurs is to screen them example, in Kenya and South Africa. based on their personal characteristics and attitudes. Djankov and others (2006) use sur- Venture capital and angel investors vey data from Brazil, China, and the Russian Federation to show that these characteristics Venture capitalists or angel investors can also can distinguish entrepreneurs from nonentre- provide financing for start-up activities. Ven- 134 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ture capital firms raise funds that they invest cial determinant of venture commitments. in early stage companies. Angel investors, on However, the presence of an IPO market is the other hand, invest their personal funds, mainly correlated with venture capital invest- tend to operate more locally in industries ment in existing firms, not start-ups. The lack where they have direct experience, and are of an IPO market can explain why venture often less visible than venture capital firms. capital funding is not common in developed These investors address informational asym- economies dominated by banks, such as Ger- metries in young firms by intensively scruti- many and Japan (Hall and Lerner 2010). In nizing firms before providing capital and then addition, cross-country evidence from 39 monitoring them afterwards. Venture capital countries suggests that the legal framework, firms differ from banks in that they typically legal origin, and accounting standards may have more specialized skills to evaluate proj- influence the characteristics of the venture ects that have few assets that may act as col- capital industry. For example, better laws— lateral and carry significant risk. In addition, approximated by a combination of measures high-powered compensation structures give of the efficiency of the judicial system, the venture capitalists incentives to monitor firms rule of law, corruption, risk of expropria- closely. Banks sponsoring venture funds with- tion, risk of contract repudiation, and share- out high-powered incentives have found it holder rights—are associated with more rapid difficult to retain personnel (Hall and Lerner deal screening and origination (Cumming, 2010). Schmidt, and Walz 2010). Research in the United States suggests that In the United States, the venture capital the supply of venture capital promotes firm industry seems to have moved away from start-ups, as well as employment and income funding start-ups in recent years. Instead, (Samila and Sorenson 2011). Angel investors venture capital firms seek to invest large sums also have positive effects on the performance in more well established and less-risky firms. of the start-ups they support (Kerr, Lerner, To fill this growing gap, angel investors have and Schoar, forthcoming). However, some played an increasingly important role in pro- of this effect may arise because of personal viding financing to early-stage companies involvement and advice that angel investors (Fishback and others 2007). A growing num- contribute to start-ups, in addition to funding. ber of angel investors have formed groups that Venture capital firms have been active in conduct screening and due diligence, allow the United States, as well as in Canada, Israel, individual angels to diversify their holdings, New Zealand, and the United Kingdom, but collect knowledge from investors with varied they tend to be less common in other coun- industry experience, and pool capital (Shane tries (Hall and Lerner 2010). Among develop- 2005). Angel groups have also been active in ing countries, Brazil and India boast advanced other developed economies, including Canada, venture capital industries with several venture France, Germany, Spain, and the United King- capital firms that are privately or government dom, with more than 30 angel groups each in financed and that focus on innovative and 2009.31 Outside Europe and North America, high-technology industries (Zavatta 2008). angel networks tend to be less developed, but A key reason why venture capital firms are in the past few years, Angel investment has particularly active in the United States may be become much more visible in East Asia and the existence of a robust IPO market because the Pacific, South and Southeast Asia, Israel IPOs allow venture capitalists to transfer and Latin America (OECD 2011a). IFC’s control back to the entrepreneur (Black and infoDev initiative has been providing technical Gilson 1998). Jeng and Wells (2000) exam- assistance to support the formation of angel ine the factors that influence venture capital networks in developing economies. To date, funding in a sample of 21 countries and find infoDev has assisted in the creation of angel that the strength of the IPO market is a cru- networks in eight economies (Belarus, Chile, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 135 BOX 3.7 Case Study: Angel Investment in the Middle East and North Africa The IFC’s infoDev initiative is launching a project ing facility. It will be piloted in the Middle East to support the creation of angel investor networks and North Africa, starting with the Arab Republic in developing economies through cofi nancing and of Egypt, Jordan, Lebanon, Morocco, and Tuni- technical assistance. infoDev is a global partner- sia. The pilot version will consist of a $50 million ship program within the World Bank. Since 2002, coinvestment fi nancing facility managed by a third- infoDev has developed a network of more than 400 party fi nancing facility manager and a $20 million small-business incubators—all locally owned and technical assistance facility managed by infoDev. operated—in more than 100 developing countries. The coinvestment facility aims to make investments A business incubator is a facility with a program in the range of $50,000–$1 million each in about to help small companies have a better chance of 200 start-up companies, investing jointly with angel survival through the start-up phase. An incubator investors who are organized around an angel net- may offer services such as offi ce space at a reduced work or angel syndicate. The technical assistance rate, shared offi ce services, entrepreneurial advice facility seeks to spur the creation of about eight and mentoring, business planning, contacts, and angel networks and to strengthen the performance networking. infoDev’s incubator network has sup- of about 50 incubators to ensure a steady stream ported about 25,000 start-up enterprises that have of investable projects. The project also proposes to reportedly created approximately 270,000 jobs. build a body of knowledge around the needs of high- Building on its network of incubators, infoDev growth entrepreneurs in the region and to contribute is implementing an early-stage innovation fi nanc- to the long-term capacity to support these needs. Jordan, Moldova, Nigeria, Senegal, South DOES ACCESS TO FINANCE Africa, and Trinidad). In each instance, local LEAD TO INNOVATION? high-net-worth individuals have been identi- fied, convened, and advised on global best Recent research has studied the relationship practices. infoDev is now also launching an between finance and innovation to understand early stage innovation financing facility that the channel through which access to finance will provide financing and technical assistance can promote economic growth. Many econo- to support angel networks in developing econ- mists have argued that innovation is essential omies, leveraging the infoDev global network for economic growth and development (for of incubators (box 3.7). example, Aghion and Durlauf 2005; Baumol In conclusion, much policy and research 2002; Schumpeter 1934). Innovation is often attention has been focused on the financial defined to include not only the invention of inclusion of SMEs, but young firms also face new products, but also the adaptation of financial constraints. Relieving these con- products and processes from other countries straints can potentially foster productivity or firms; for example, see the Oslo Manual growth and job creation, although more evi- (OECD 2005). dence is needed to document this relationship. Angel investors provide a promising source Why innovative projects may be of finance for young firms. Currently, angel particularly difficult to finance investor groups are less common in develop- ing countries than in developed economies, Firms may face especially severe financing but financial and technical assistance can constraints on innovative investments because help support the creation of angel investor of information asymmetries. Firms seeking to networks. make new inventions through research and 136 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 development frequently have better informa- investment in innovation. Empirical evidence tion about the likelihood of success and the indicates there is a positive relationship nature of the completed project than potential between bank financing and innovation. Using investors. In fact, because inventions can be data from the World Bank enterprise surveys, easy to imitate, firms may be reluctant to dis- a study examined more than 19,000 small, close much information about their projects, medium, and large firms across 47 developing making it difficult to find financing. Firms are economies and found that the use of external thus often required to rely on internal funds finance is associated with greater innovation to finance research and development expendi- by private firms (Ayyagari, Demirgüç-Kunt, tures (Hall and Lerner 2010).32 and Maksimovic 2011b). In particular, bank Innovation that does not involve new inven- financing is positively associated with firms tions, but involves the technological upgrading that undertake core innovation, which is of processes may also be unattractive for lend- defined as upgrading an existing product line ers. Banks often prefer to use physical assets or introducing a new product line or new to secure loans and are reluctant to lend if a technology (figure 3.9). project involves knowledge investment, such Survey data on microenterprises and SMEs as contracting business development services, in Sri Lanka also show that firms are more rather than investment in plant and equip- likely to innovate if they have a bank loan (de ment. A study in Mexico found that financing Mel, McKenzie, and Woodruff 2009a). This constraints were a key reason why firms did relationship may not be causal, though, and not hire business consultants, although the use could be driven by other factors. For exam- of consulting services had positive effects on ple, more well managed firms may both be productivity and firm growth in the longer run more likely to innovate and more likely to (Bruhn, Karlan, and Schoar 2013). obtain a bank loan. More empirical within- country research is thus needed to shed light on the causal effects of access to bank finance Can bank finance foster innovation? on innovation. Given the challenges, it is not clear whether policies that promote access to bank financ- Nonbank finance and government ing, as discussed in earlier sections of this subsidies for innovation chapter, are sufficient to foster innovation. They may help to some extent because firms The availability of nonbank financing, espe- could use bank financing for noninnovative cially venture capital or angel investment, activities and thus free up internal funds for may be critical in fostering innovation. As discussed above, venture capital firms and FIGURE 3.9 The Estimated Effect of Financing Sources on Innovation angel investors address information asymme- try problems by intensively scrutinizing firms before providing capital, and then they moni- tor them afterwards. However, venture capi- Bank financing 0.02 tal firms and angel investors tend to focus on high-technology sectors and a small number of firms. A broad range of firms in other sec- External financing 0.01 tors could potentially also benefit from tech- nological upgrading (Bloom and others 2013; 0 0.005 0.010 0.015 0.020 0.025 Bruhn, Karlan, and Schoar 2013). To facili- Increase in core innovation index when proportion of new investment tate access to business development services financed by banks (or externally) increases by 10 percentage points that can help upgrade technologies in such firms, some governments have implemented Source: Ayyagari, Demirgüç-Kunt, and Maksimovic 2011b. subsidy or matching grant programs. Note: The core innovation index is formed by adding 1 if the firm has developed a new product line, upgraded an existing product line, or introduced a new technology. External financing comprises Matching grant programs are one of the bank financing and all other financing that is not from internal funds or retained earnings. most common policy tools used by develop- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 137 ing-country governments to foster technologi- example, in six African countries. However, cal upgrading and innovation. A matching these attempts have ultimately failed in part grant consists of a partial subsidy to a firm because the programs received few applica- for the use of business development services tions (Campos and others 2013). This lack or similar activities. They often cover 50 per- of demand may be caused by overly strict eli- cent of the cost. Despite their widespread gibility criteria, red tape, capture by special use, there is little evidence on the effective- interest groups, or incentives facing project ness of the grants in spurring firms to under- implementation staff and suggests that match- take activities they would not have otherwise ing grant programs are difficult to implement. undertaken or whether the grants merely sub- Several countries have experimented with sidize firms for actions they would have taken encouraging innovation and job creation anyway. A randomized impact evaluation of through business plan competitions. A recent a matching grant program in Mexico finds example that has garnered much attention is that subsidized business consulting services Nigeria’s YouWiN! competition (box 3.8). led to short-run improvements in the pro- Fiscal incentives can also be used to stimu- ductivity of firms and to a long-run increase late innovation and research and develop- in employment among firms, that is, to firm ment. More and more countries are imple- growth (Bruhn, Karlan, and Schoar 2013). To menting these policies to incentivize firms to expand the evidence base, researchers have invest in research and development. Mulkay tried to conduct impact evaluations of match- and Mairesse (2013) have examined the ing grant programs in other countries, for impact of the research and development tax BOX 3.8 Case Study: Nigeria’s YouWiN! Business Plan Competition The Youth Enterprise with Innovation in Nigeria As is typical with business plan competitions, Program (YouWiN!) is a business plan competition YouWiN! does not provide access to fi nance on a for young entrepreneurs in Nigeria sponsored by the broad scale. Instead, the program seeks to target country’s Ministry of Finance, Ministry of Commu- fi rms and business ideas with high potential. This nication Technology, and Ministry of Youth Devel- is refl ected in the fact that the level of educational opment, with support from the U.K. Department attainment is higher among applicants than among for International Development and the World Bank. the overall population. According to data from the The program was launched on October 11, 2011, by 2008 general household survey, among the overall President Goodluck Jonathan in a ceremony aired youth population, 5.5 percent have a university edu- live on national television. It has the stated objective cation, compared with slightly more than 50 percent of encouraging innovation and job creation through of the YouWiN! applicants. the establishment of new businesses and the expan- YouWiN! applications were scored by the Enter- sion of existing businesses. prise Development Center of the Pan-African Uni- The program combines training with cash grants versity, a private, nonprofit educational institution to build business capacity and reduce fi nancing con- located in Nigeria’s capital, Lagos. Based on these straints to promote business creation and growth. In scores and geographical location, 6,000 candi- response to advertisements throughout the country dates were selected to attend a four-day training via television, radio, newspapers, and road shows, session, which took place in December 2011, with the program received almost 24,000 applica- 4,873 individuals participating. All applicants who tions from youth aged 18 to 40 years. Nigeria has attended the training were then given until late Janu- approximately 50 million people in this age range. ary 2012 to submit a business plan. In total, 4,510 The 24,000 applications therefore represent only business plan applications were received. These 0.05 percent of the overall youth population. were scored by a joint Enterprise Development (box continued next page) 138 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BOX 3.8 Case Study: Nigeria’s YouWiN! Business Plan Competition (continued) Center–PwC team, narrowing down the field, fi rst, FIGURE B3.8.1 Business Sectors of YouWiN! Winners to 2,400 semifi nalists and, then, to 1,200 winners. Figure B3.8.1 illustrates that the main sectors of the Agriculture YouWiN! winners are agriculture, information tech- IT and computer services nology, and computer services, manufacturing, and Manufacturing other professional services. Other professional services YouWiN! winners received up to $32,000 each Other industries in grants for new businesses and $64,000 for exist- Tailoring ing businesses. For the median existing business that Construction won the competition, this is equivalent to more than Retail trade six years of annual turnover. The grants can repre- Personal services sent a large increase in access to fi nance. According to the IFC Enterprise Finance Gap Database, only Food preparation 8 percent of Nigerian microenterprises and SMEs 0 5 10 15 20 25 30 35 have a loan from a bank, with an average loan Firms in each business sector, % size of 18 percent of annual revenue. a A World New businesses Existing businesses Bank team is currently conducting an evaluation of Source: YouWiN! program organizers. YouWiN! to measure its impact on fi rm start-ups, Note: The data are based on the self-classification of applicants at the time of fi rm survival (for existing businesses), employment, the submission of their business plans. IT = information technology. sales, and profits. a. IFC Enterprise Finance Gap Database, SME Finance Forum, http://smefi nanceforum.org. credit on private research and development (Bruhn 2009; GPFI 2011c). While several investment among French firms. They find factors, such as type of business, managerial that, in the long run, the tax relief raised skills, or the regulatory environment, may be research and development investment among driving these differences, access to financial firms by 12 percent, without taking into services may matter. This section first exam- account spillover effects. ines why women may face more constraints In summary, innovative projects are partic- in financial markets than men. It then asks ularly difficult to finance with external funds whether boosting access to finance, defined as because of information asymmetries. Govern- access to credit and savings instruments, can ments can step in by sponsoring matching spur the growth of woman-owned firms. grant programs and business plan competi- tions. Evidence indicates that matching grants Is there a gender gap? can enhance firm productivity and employ- ment growth, but the grants are difficult to The gender gap in access to finance is still a implement well. Research on the impact of relatively unexplored topic, but a growing business plan competitions is ongoing. literature documents that, relative to men, women face less favorable conditions in seeking financing. Even after controlling for DOES GENDER MATTER IN THE a range of demographic and socioeconomic ACCESS OF FIRMS TO FINANCE? characteristics, a recent cross-country study by Empirical evidence suggests that woman- Demirgüç-Kunt, Klapper, and Singer (2013) owned firms tend to be smaller than firms finds that the ownership of bank accounts owned by men and grow at a slower rate and the usage of savings and lending instru- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 139 ments are substantially less prevalent among FIGURE 3.10 Number of Countries with Joint women. Other cross-country studies focusing Titling of Major Assets for Married Couples on the access of entrepreneurs to credit reach similar conclusions. The findings suggest that women entrepreneurs are less likely than their No joint men counterparts to obtain financing from titling, formal institutions and are also more likely 11 to pay higher interest rates or to receive less- favorable loan terms (Demirgüç-Kunt, Beck, Joint titling and Honohan 2008; GPFI 2011c; Muravyev, is default, 54 Schäfer, and Talavera 2009). Gender gaps in the credit market have also been documented Joint titling available, but in particular settings, such as in informal loan not default, 76 markets in rural Mexico, where women are more likely to be credit constrained (Love and Sanchez 2009). Unfavorable conditions in women’s access to finance may also affect the demand among women for external financing and the busi- Source: Calculations based on 2012 data of the Women, Business, and ness decisions of women. For instance, the Law (database), World Bank, Washington, DC, http://wbl.worldbank women may select into businesses that need .org/. Note: The sample includes 141 countries. Major assets include land or the less credit. However, gender gaps in access marital home. to finance are less evident in other regions. A study by Bruhn (2009) finds that women rower’s immediate family. Given women’s lim- and men entrepreneurs have similar access to ited mobility and the prevalent social norms, it credit in Latin America. may be difficult for women to find men guar- antors who are not relatives. Determinants of the gender gap Another common obstacle women face in seeking credit is that they often own fewer Gender differences in access to finance may assets than men, limiting their collateral. be explained by several factors, ranging from This can be partly caused by cultural prac- cultural, regulatory, and legislative barriers to tices that serve to give women less access to statistical or even taste discrimination. Identi- assets, as shown by Hallward-Driemeier and fying which factors are at play can allow poli- Hasan (2012) in the case of Africa. Lack of cies to be tailored to help level the field for ownership can also be driven by laws and women in financial markets. regulations. The Women, Business, and the In many environments, cultural beliefs, Law database documents that, in 11 out of laws, or regulatory mandates may prevent 141 countries, joint titling of major assets women from participating in financial arrange- (such as land or the marital home) does not ments. Demirgüç-Kunt, Klapper, and Singer exist among married couples (figure 3.10). (2013) find evidence supporting that gender In 54 of the 130 countries where joint titling norms, by law or custom, are associated with does exist, it is not the default titling proce- lower access and usage of financial services dure for marital property. Moreover, in 26 by women. In a recent World Bank report, of 141 countries, sons and daughters do not Safavian and Haq (2013) document that few have equal inheritance rights to the property women in Pakistan obtain individual loans of their parents, and in 25 countries women instead of group loans. One reason is that the and men surviving spouses do not have equal guarantor requirements for individual loans inheritance rights. may be prohibitive for women. Most MFIs Raising the incidence of property owner- typically require two men guarantors for a ship among women may thus require changes loan, only one of whom may be from the bor- in the law. While this may be a longer-term 140 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 endeavor, financial institutions can also take at a disadvantage. For example, 90 percent steps to increase the availability of collateral of the loan officers at a large public sector for women. A bank in Uganda, the Devel- bank in India are men (Cole, Kanz, and Klap- opment Finance Company of Uganda Bank, per 2012). Another study in India finds that launched the Women in Business Program in cultural proximity between loan officers and 2007 (GPFI 2011c). As part of this program, borrowers increases lending, lowers default the bank has specifically designed some of rates, and reduces the amount of collateral its loan and savings products to address the required. These findings suggest that cultural needs of women entrepreneurs. Because col- proximity mitigates informational problems, lateral requirements are a major obstacle which, in turn, relaxes financial constraints among Ugandan women who have diffi- and improves access to finance (Fisman, Para- culty accessing property, the bank created a vasini, and Vig 2012). It is not clear whether land loan for women. Through this product, these findings easily translate to gender, but, women are able to obtain a loan to purchase if they do, the policy implication would be property that they can later use as collateral that the presence of more women loan offi- for a business loan. cers could relax the financial constraints on Women’s differential access to finance can women. In fact, there is some evidence that be determined by gender differences in out- women make better loan officers. Beck, Behr, comes such as education, income, or busi- and Guettler (2012) examine the relationship ness experience. These differences could, between the gender of loan officers and loan for instance, influence women’s ability to performance. They show that loans moni- keep adequate financial records that may be tored by women loan officers exhibit lower required to obtain a loan. Moreover, women arrear probabilities than loans handled by may be more prone to default if they have men loan officers. This result is explained by less education or if they have less experience the greater capability of women loan officers in running businesses or understanding finan- to build trust relationships with borrowers. cial contracts, which makes financial institu- tions less inclined to serve women clients. This Does increased access to finance type of discriminatory treatment is known as promote growth among woman-owned statistical discrimination. Aterido, Beck, and firms? Iacovone (2011) find evidence of statistical gender discrimination in access to finance. Improving access to financial services by itself In nine countries in Sub-Saharan Africa, they may not stimulate the growth of woman- find that the gender gap in financial services is owned firms if the factors causing the differ- explained by differences in education, income, ential access are not addressed. For instance, formal employment, and status as the house- if women select to run businesses with lower hold head. Once they control for these char- capital returns relative to firms operated acteristics, the gender gap in access to finance by men, interventions that improve access disappears. to capital will not be more effective among Gender bias or taste discrimination may woman-owned firms. An impact evaluation of also limit women’s access to finance. A recent a program in Sri Lanka that distributed capital study by Beck, Behr, and Madestam (2012) grants found higher profits among enterprises argues that own-gender preferences affect owned by men, but not among woman-owned both credit supply and demand. More spe- enterprises. The difference in the returns to cifically, the authors find that borrowers capital did not appear to be driven by differ- matched with loan officers of the opposite ences in entrepreneurial ability or risk aver- gender pay higher interest rates, receive lower sion between men and women, but by the fact loan amounts, and are less likely to return that the sectors that women selected typically for a second loan. In settings in which loan had lower returns to capital in the first place. officers are predominantly men, own-gender But, even in sectors in which both man- and preferences could put women loan applicants woman-owned firms are common, woman- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 141 owned enterprises may have lower returns to ing among microfinance clients raised profits capital if women face household constraints. among woman-owned firms, but not among A study by de Mel, McKenzie, and Woodruff firms owned by men. One substantial benefit (2009b) finds that women have higher returns of the training was that it enabled women cli- to capital if they have more decision-making ents to obtain more favorable loan terms and power in the household or if their spouses are explore alternative funding options (Bruhn more cooperative with regard to the manage- and Zia 2013). ment of the enterprises. Overall, it appears that, if policies tackle Recent evidence suggests that financial the underlying causes of the gender gap, interventions might be more effective if they boosting access to finance can improve the were combined with business training. For performance of woman-owned enterprises. In instance, a study in Sri Lanka finds that capi- settings in which gender bias or taste discrimi- tal grants can temporarily raise the profitabil- nation is relevant, policies should consider ity of woman-run subsistence enterprises if antidiscriminatory programs or competition- the women are provided with business train- enhancing strategies (Beck, Behr, and Mad- ing, but this impact dissipates two years after estam 2012). If statistical discrimination is completion of the training (de Mel, McKenzie, the main driver, then policies should aim at and Woodruff 2012b). improving women’s education, employment However, in contexts in which women status, and income opportunities, if these are face household constraints, these interven- the relevant dimensions. For instance, policies tions may not be as effective. In rural Paki- promoting financial inclusion may have to be stan, Giné and Mansuri (2012) evaluate the combined with business training. Depending impact of a business training program and on the local context, household constraints the offering of larger loans to microfinance may prevent women from making changes clients. Their results suggest that the training in their businesses so that even the combina- enhanced the business knowledge of both men tion of credit and training may not be effec- and women microentrepreneurs. However, tive. In these types of settings, policies aimed the performance of the woman-run businesses at strengthening the position of women in did not improve, though the training lowered society should be promoted. This may require business failure rates and boosted the sales reforming the laws and regulations pertaining among man-run businesses. Offering larger to property rights and inheritance to encour- loans had little effect on any enterprises in the age women’s ownership of assets. sample. As in the findings of other studies, the lack of effect on women may in part be AGRICULTURAL FIRMS: driven by household constraints (for instance, CAN FINANCE INFLUENCE see Fletschner and Mesbah 2011). About PRODUCTIVITY? 40 percent of women report that their spouses are responsible for most business decisions, Agriculture is an important sector in most suggesting that woman-owned enterprises developing countries. According to the Food show no improvement because women have and Agriculture Organization of the United little decision-making control in their own Nations, the agricultural sector is the main businesses. For a sample of five Sub-Saharan source of income and employment among African countries, Aterido and Hallward- 70 percent of the world’s poor in rural areas. Driemeier (2011) show that, in enterprises in If the rural poor benefit more from growth which women are part owners, a man is the in agriculture than from growth in other sec- decision maker in 77 percent of the cases. They tors, then agriculture can be a relevant vehicle stress the importance of looking at decision- for reducing extreme poverty. A recent cross- making control and not partial ownership by country study by Christiaensen, Demery, and women in evaluating enterprise performance. Kuhl (2011) suggests that this may be the A study in Bosnia and Herzegovina reaches case. According to the results, poverty among a different conclusion. There, business train- the poorest of the poor is more sensitive to 142 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FIGURE 3.11 Adults with an Account Used for Business Purposes 25 24 23 20 Adults, % 15 10 8 7 6 6 5 5 4 4 5 5 4 2 2 0 High East Asia Europe and Latin America Middle East and South Sub-Saharan income and Pacific Central Asia and the North Africa Asia Africa Caribbean Rural Urban Source: Calculations based on the Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex. growth in this sector. While the results of ture, such as weather variability, natural haz- other studies are more modest, they also indi- ards, and commodity price volatility (World cate that agricultural growth translates into Bank 2007a). Improving access to finance in lower rural poverty rates (for instance, see the sector can raise the investment choices Foster and Rosenzweig 2003; Ravallion and of farmers and provide farmers with more- Datt 1996; Suryahadi, Suryadarma, and Sum- effective tools to manage risks. arto 2009; World Bank 2007a). This section discusses the main constraints In this respect, agricultural finance may have faced by institutions in supplying financial the potential to increase the productivity of the instruments such as savings, credit, and insur- sector. Currently, the use of basic accounts for ance products to the agricultural sector. It then business purposes in rural areas remains lim- outlines the most promising developments in ited, lagging far behind the situation in urban the field and, finally, provides concrete policy areas, particularly in developing countries (fig- recommendations to enhance financial inclu- ure 3.11). In the absence of formal financial sion within the agricultural sector. instruments, farmers and agricultural firms rely on alternative informal mechanisms (such Why have financial institutions been as moneylenders, transfers across households, reluctant to serve the sector? or informal savings arrangements) to supply their finance needs (Lim and Townsend 1998; One relevant barrier that has historically Munshi and Rosenzweig 2005; Rosenzweig prevented financial providers from serving and Wolpin 1993; Townsend 1994). How- the agriculture sector and its supply chain ever, these informal arrangements may prevent is geographical. Farmers and agricultural farmers from adopting better technologies, firms find it difficult to access banks because purchasing agricultural inputs, or improving most of these farmers and firms are located the efficiency of their businesses if appropriate in rural areas, where banks are discouraged risk mitigation products are lacking or if the from operating at a profitable scale because of available financial instruments do not match low population density and large geographi- the needs of farmers. cal dispersion. Hence, the provision of formal Financial constraints are costly, particularly savings, insurance, and credit instruments to to the smallest farmers and agribusinesses, farmers and agricultural SMEs is limited. limiting their ability to compete and protect Another major factor inhibiting financial against a variety of risks inherent to agricul- institutions from serving the sector is the GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 143 systemic risk that characterizes agricultural 10 percentage points. The new loans are activities. When natural hazards or adverse directed at districts with a political interest weather conditions take place, they typically for the ruling state party, are less likely to be affect a large number of farmers and firms repaid, and are not associated with greater simultaneously, making it more challeng- agricultural output. In line with these find- ing for financial providers to diversify their ings, de la Torre, Giné, and Vishwanath client portfolios because, when one client (2011) conclude that primary agricultural fails to pay, many others will be in the same credit cooperatives in India have been used situation. as political instruments, and the responses of The lack of financial infrastructure poses borrowers have been to prioritize their debt a third challenge. Tracking the identity of cli- payments from institutions other than the ents or monitoring production outcomes is cooperatives because the borrowers associ- extremely difficult in rural areas. If natural ate the cooperatives with frequent govern- hazards cannot be mapped to the production ment relief packages. Their results suggest of farmers and SMEs, or if the lack of infor- that the greater involvement of governments mation about clients means financial pro- in credit markets might lead to unintended viders face difficulty in tracking them, then consequences such as incentivizing defaults. farmers and firms might well prefer to default Kanz (2012) finds that India’s largest bailout or underperform, especially in settings with program, the Debt Waiver and Debt Relief low contract enforcement. Hence, potential Scheme for Small and Marginal Farmers, lenders or insurers may be discouraged from did not alleviate problems of debt overhang engaging with farmers and agricultural SMEs among beneficiaries. Instead, program recipi- in the first place or may be reluctant to supply ents augmented their reliance on informal certain products potentially demanded by cli- credit and reduced their productive invest- ents, such as longer maturity loans. Financial ment. This conclusion indicates that ben- institutions serving rural areas have actually eficiaries were concerned about the stigma of responded by excessive credit rationing or being identified as defaulters because of the overreliance on traditional forms of collat- program and the effect this may have on their eral, which many small and medium farmers future access to formal credit. and firms lack in the first place (Stein, Rand- Recently, other reasons inhibiting financial hawa, and Bilandzic 2011). institutions from serving this market have Additionally, the lack of interest of finan- begun receiving more attention. If individuals cial providers in serving farmers and agricul- in rural areas do not trust banks or lack finan- tural firms is aggravated by the paternalistic cial training, they will be less likely to use behavior or political motives that govern- financial products. Moreover, if the products ments may have. Policies ranging from sub- that financial institutions offer do not ade- sidies with no proper assessment of project quately match the needs of farmers, then the feasibility to political loans to the sector or products will be associated with low demand. unconditional bailouts to relieve households In a study in Kenya, Dupas and others (2012) from their debt obligations may distort firm find that these factors partially explain the and farmer incentives and discourage finan- low usage of financial services among rural cial providers from entering the market. customers. The study randomly waived the Evidence from recent studies cautions that cost of opening a basic savings account to government interventions in the credit market unbanked individuals and found that only 18 can be costly not only in terms of political percent of these individuals actively saved in capture, but also in amplifying market distor- their new accounts. The main reasons people tions, thereby aggravating financial exclusion. did not save in their accounts were lack of Cole (2009) finds evidence that, during elec- trust in the bank, unreliable service, and high tion years, the amount of agricultural credit withdrawal fees. The study also provided supplied by public banks increases by 5 to information on credit options and the lower- 144 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ing of loan requirements. After six months, Based on the system of warehouse receipt only 3 percent of the people under study financing, different approaches have been had applied for a loan. People did not bor- developed to extend financial services to agri- row from the bank because they feared losing businesses and farmers. Factoring and leasing their collateral. These results suggest that not are two examples. Through factoring, farm- only access matters, but also the quality of the ers can use their accounts receivables as col- services offered. lateral for loans (see above).34 Leasing is of particular relevance for the agricultural sector because it has the potential New developments in agricultural to expand the medium- and long-term finance finance available to agribusinesses. Through leas- In the last two decades, new approaches have ing, farmers can acquire machinery and farm been emerging in agricultural finance. Most equipment, while paying for it on a more flex- of these approaches are designed based on ible basis (IFC 2012c). microfinance principles and adapted to fit Agricultural production is being trans- the needs of farmers and agricultural firms. formed into integrated market chains, link- Following Kloeppinger-Todd and Sharma ing smaller farmers to large multinational (2010), the most promising developments can firms.35 Value chain finance has become one be grouped into four thematic areas: tailor- of the most popular mechanisms for financ- ing financial services to the business reality of ing commercial agriculture in various devel- farmers and agribusinesses, using technology oping countries. Value chain finance is par- to reach out to new clients and reduce transac- ticularly useful in helping link small farmers tion costs, developing innovative risk manage- and agribusinesses into effective market sys- ment strategies, and bundling financial instru- tems (Miller and Jones 2010).36 An innova- ments with other financial or nonfinancial tive business model in this area in Mexico services to overcome the multiple constraints is Agrofinanzas. Agrofinanzas specializes in faced by farmers and agricultural SMEs. lending to small farmers with little experience Because most commercial banks operate with banks and formal financing. Its business exclusively in urban markets, they are highly model is based on relationships with larger inexperienced with rural settings, where the firms that are connected to smaller farmers. reality of business and the demand for finan- Agrofinanzas identifies its borrowers through cial products are different.33 For instance, the information obtained by large firms on their reliance of commercial banks on traditional small suppliers. This information is criti- land collateral excludes from borrowing many cal to making credit risk manageable for the farmers and agricultural firms that lack secure institution. rights to land. New financing models such as Another example in Mexico is FIRA (Trust warehouse receipts can help relieve this con- Funds for Rural Development), which pro- straint. Through warehouse receipt financing, vides a broad range of financial products and farmers can obtain finance by using nonper- services to banks to promote the develop- ishable goods deposited in a warehouse as col- ment of the rural sector. Among other initia- lateral for their loans. While several elements tives, FIRA supplies structured financing, that are needed to develop a well-functioning is, specific, tailored products based on client warehouse receipt financing system, including needs and the local business environment to an enabling regulatory framework, licensed help clients manage the risks involved in their warehouses, and appropriate training, these everyday operations. instruments have been operating successfully The use of technology to facilitate financial in several countries (Höllinger, Rutten, and transactions has great potential in agricultural Kiriakov 2009). However, a rigorous assess- settings, where transaction and information ment on how important these systems have costs are high. Credit and movable collateral been for farmer productivity is still needed. registries, mobile banking, and correspondent GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 145 banking are examples of ways in which tech- been rising over the years, providing revenue nology can help ease market failures in an agri- to the municipalities when adverse weather cultural setting. In a recent project in Malawi events take place. Mostly, the insurance prod- (see chapter 2), Giné, Goldberg, and Yang ucts offered in the Mexican market focus (2012) find that the use of fingerprints to iden- on two schemes. The first is weather index tify clients makes the threat of future credit insurance, which operates based on weather denial more credible. The incentives for clients indexes measured through weather stations to pay back loans are thereby increased, while that determine crop damage. The second is simultaneously incentivizing lenders to engage yield index insurance, in which the payment in more transactions. Even though projects of is determined when the observed yield, esti- this type are at the pilot stage, they show great mated through sampling in the risk unit, is potential for reducing the information costs of lower than the covered yield.38 lenders or insurers. Other examples include However, index insurance still faces chal- Kenya’s M-PESA (chapter 2) and initiatives to lenges, including low take-up rates, the diffi- introduce registries for movable collateral (see culties of farmers in understanding and thus box 3.5). valuing the complicated insurance, and the New research suggests that a major con- failure to dissipate a considerable part of the straint on farmer investment is uninsured risk. risks among farmers (Carter 2008; World Karlan, Osei, and others (2012) find that, Bank 2007a). Other studies also suggest that when farmers in north Ghana were insured the lack of trust of users in the insurance against the primary catastrophic risk, they products can partially explain the low take- increased expenditure on their businesses. Cai up rates (Cai and others 2009). and others (2009) have evaluated a Chinese Commodity exchange markets, which insurance program aimed at increasing the have a prominent history in Asia and Latin supply of pork. The insurance was found to America and are now being piloted in sev- have substantial effects on the decisions of eral African countries, such as Ethiopia, can farmers to raise pigs. be good tools to help farmers hedge against Evidence suggests that instruments such as adverse price fluctuations. In a commodity index insurance succeed in minimizing moral exchange market, different financial instru- hazard and adverse selection and, under some ments can be traded, such as futures, for- circumstances, can incentivize farmers to wards, options, derivatives, and swaps. The make riskier, but more profitable investments objective of all these instruments is to lock in (Giné, Menand, and others 2010; Karlan, the price of a product in the future. Through Osei-Akoto, and others 2012; Mobarak and future contracts, for instance, the purchase Rosenzweig 2012). While index insurance or sale of a specific quantity of a commodity represents only a small fraction of the broad on a determined date in the future is agreed range of insurance products available, its in advance (IFC 2012c). Many commodities, presence is growing in various countries. An ranging from orange juice to cereals and cot- example is the Kilimo Salama microinsurance ton, are traded in futures markets. However, scheme in Kenya, which is a weather index many conditions are required to develop a insurance offered, together with loans, to help well-functioning exchange market, such as farmers buy farming products. The scheme the availability of a large number of market has insured 64,000 farmers across Kenya participants, adequate infrastructure, and an and is to be expanded to other countries in appropriate legal framework (Rashid, Winter- Africa, such as Rwanda.37 Another example Nelson, and Garcia 2010). is CADENA, a public insurance initiative in While financial products can contribute to Mexico. Through CADENA, the govern- risk mitigation and the promotion of more ment purchases insurance for municipalities profitable investments in agriculture, bun- on behalf of residents, thereby promoting dling them with other services can help over- take-up at the municipal level. Coverage has come various constraints affecting farmers 146 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 and agricultural SMEs. For instance, financial interest rates and the lack or nonenforcement services should be provided with the proper of appropriate rules and regulations (World financial training. For effective value chains Bank 2007a). to operate, financial instruments must be bundled with the timely availability of inputs, NOTES efficient marketing, and distribution channels for agricultural outputs. 1. For microenterprises and small firms, the dis- tinction between households and enterprise finance may be blurry because firm owners Policy recommendations may mix their personal finances with the finances of the firm. However, in contrast to Summing up, evidence suggests that produc- chapter 2, this chapter zooms in on the role tivity in the agricultural sector can benefit of finance in promoting business activities. from better access to financial instruments It also discusses programs and interventions tailored to the needs of farmers and agribusi- that are targeted specifically at firms, such as nesses. Policy makers can take a series of business training, risk-sharing facilities, and steps to make this happen. First, investing in factoring. rural financial infrastructure can overcome 2. This follows the definition used in the World the information asymmetries that discourage Bank enterprise surveys (http://www.enter financial providers from serving agricultural prisesurveys.org/), which is based only on the firms. The availability of public databases number of employees and which also defines small firms (5–19 employees) and medium on agricultural and weather statistics would firms (20–99 employees). Many countries allow lenders and insurers to distinguish good and institutions use their own definitions of clients from bad ones more precisely and SMEs. For instance, in a survey of banks in monitor their actions. Governments have a the Middle East and North Africa, the cutoff comparative advantage in providing informa- used by the banks to distinguish small and tion to help lenders or insurers identify their medium firms ranged from 5 to 50 employ- risks and price them accordingly (World Bank ees, and the cutoff between medium and large 2007a). firms ranged from 15 to 100 employees (IFC Second, strengthening property rights and 2010b). Some of the SME definitions are contract enforcement can open up access to based not only on the number of employees, important financial products to farmers and but also on other variables, such as sales and assets. For example, the European Union SMEs, for instance, by allowing the develop- defines SMEs as firms with 10–250 employ- ment of value chain finance. ees, less than €50 million in turnover, or less Third, governments should abstain from than €43 million in total balance sheet (IFC paternalistic policies that discourage financial 2010b), although these thresholds may be providers from entering the market and that high for most developing economies. The distort the incentives for farmers and firms. heterogeneity in definitions of SMEs across Public subsidies directed at agriculture should countries and institutions can pose an obsta- be carefully considered because they provide cle to collecting accurate and comprehensive inappropriate incentives for farmers to invest data on SMEs and to designing policies tar- in unprofitable farming activities. While cer- geted at SMEs. tain subsidized insurance products could be 3. The distinction between microenterprises and justified on the basis of achieving the higher SMEs based on number of employees is not absolute, and some firms with fewer than take-up of these products and allowing users five employees may also seek funding from to understand their value, subsidies that do sources other than MFIs, particularly if they not involve proper assessments of the quality aim to grow quickly or are young firms. or feasibility of projects should be avoided. 4. See the SME Finance Forum website, at http:// The gravest risks to sustainable financing in smefinanceforum.org/. The database is based agriculture often arise from misguided gov- on the Enterprise Surveys (database), Interna- ernment interventions such as subsidized tional Finance Corporation and World Bank, GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 147 Washington, DC, http://www.enterprise 12. Another potential explanation for the find- surveys.org. ings is that small firms were less productive so 5. This information is from the same IFC Enter- that they could not pay market interest rates. prise Finance Gap Database through the SME However, more than 95 percent of small firms Finance Forum, http://smefinanceforum.org. had at least some credit at market interest The data set considers as informal all micro- rates. Moreover, the paper does not find that enterprises and SMEs that are not registered more productive firms were less affected by with the authorities and all nonemployer the drop in subsidized credit. firms (independently of registration status). 13. For most countries, the survey does not 6. For additional information on the infor- include comprehensive information on the mal surveys, see “Enterprise Surveys Data,” share of loan applications that were rejected World Bank, Washington, DC, http://www or on the reasons for rejection. .enterprisesurveys.org/Data. 14. The incentives for banks to lend to SMEs can 7. The evidence on the impact of these train- also depend on the regulatory framework, for ing courses is summarized in McKenzie and example, on capital requirements. Regulators Woodruff (forthcoming). are currently introducing Basel III, which is a 8. The study measures aggregate effects, that new global regulatory standard on the capi- is, increases in average income (Bruhn and tal adequacy and liquidity of banks agreed by Love 2013). Critics have been concerned that the Basel Committee on Banking Supervision Banco Azteca may do more harm than good in response to the deficiencies in financial among some borrowers because of its high regulation revealed by the global financial interest rates and diligent repossession of col- crisis. Basel III introduces new regulatory lateral, including household appliances, in the requirements on bank capital, liquidity, and case of default. These issues are of particular leverage. The higher capital requirements are concern with respect to individuals with low expected to raise the average cost of bank lia- levels of financial literacy. Note, however, bilities, which could push up the interest rates that a recent study examining the impact of charged on loans, including to SMEs (GFPI loans with a 110 percent annual interest rate 2011f). Also, see chapter 1 for an in-depth given out by the largest microlender in Mex- discussion of how banking market structure ico, Compartamos Banco, finds little support influences access to finance. for the hypothesis that microcredit causes 15. These numbers also reflect a merger with harm (Angelucci, Karlan, and Zinman 2013). Fortis Bank in 2010. 9. Although average returns to capital are high 16. For more information, see IFC (2012b) and among microenterprises, there may be con- the underlying IFC case study on Turkey. siderable variation in these returns across 17. Other products offered by IFC include lines firms. of credit to banks in developing economies 10. The study relies on cross-sectional data for on-lending to SMEs. There is a lack of rig- whereby firms report employment levels for orous evidence on the impact of these credit multiple years so that employment growth lines. More research is needed to determine may be calculated. The data thus do not cap- how effective they are in increasing access to ture firms that have closed by the time the credit among SMEs that would not otherwise survey was conducted. A study on the United have received a loan. States using repeated survey data finds that 18. Another type of credit guarantee helps finan- small firms display higher net employment cial institutions raise long-term funds. For growth than large firms (Haltiwanger, Jar- example, the World Bank is providing a par- min, and Miranda 2010). However, once the tial credit guarantee for up to €200 million authors control for firm age, there is no rela- to the Croatian Bank for Reconstruction and tionship between firm size and employment Development to support fundraising in finan- growth. See the discussion in our chapter here cial markets for on-lending to private sector on young firms. exporters and foreign currency earners. 11. SMEs in developing economies face many 19. Given the scale of credit guarantee schemes other constraints to growth, including regula- around the world, there is relatively little evi- tory obstacles, missing physical infrastructure dence on whether and how they work. More (such as roads and reliable electricity supply), research is needed in this area to complement and lack of skills. the few existing studies. 148 FINANCIAL INCLUSION FOR FIRMS GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 20. For an in-depth review, see the Global Finan- could potentially lead to confounding results cial Development Report 2013 (World Bank (reverse causality) if, for example, individu- 2012a). als whose businesses failed and who had to 21. Enterprise Surveys (database), International default on loans have fewer entrepreneurial Finance Corporation and World Bank, Wash- attitudes because of this experience. ington, DC, http://www.enterprisesurveys 31. The number of angel groups operating in the .org. United States was about 350 in 2009, and 22. See also UNCITRAL (2010) for a guidebook it was about 400 in all European countries on efficient and effective secured transaction combined (OECD 2011a). laws. 32. These issues may be less severe in economies 23. In an effort to provide more information on in which financial intermediaries are able to firms that have not previously had a loan, monitor firms closely or in which contracts some credit bureaus also collect payment his- are in place, encouraging firms to reveal prof- tories on utility bills or other services. its truthfully so that investors can be fairly 24. Additionally, there is a forthcoming study compensated. Monitoring is less costly in by the International Committee on Credit economies with ample credit information. Reporting, to be released by the end of 2013, Complex contracts rely on a sound legal sys- on credit reporting and SMEs, which will tem for enforcement (Cole, Greenwood, and examine, for example, the role of trade credit Sanchez 2012). and how it is captured by credit reporting 33. Several World Bank initiatives, such as Agri- systems. Fin, provide technical assistance to financial 25. Credit Bureau Singapore is an example of institutions. AgriFin supports financial insti- an institution that collects trade credit data tutions in Africa and Asia in the development on SMEs. It combines this information with of models of agricultural finance that reach credit history data from banks and with the smallholder farmers. The Centenary Bank of personal credit histories of business owners Uganda partnered with AgriFin to expand and key stakeholders to quantify default risk lending to the sector by taking several mea- (see GPFI 2011f for more detail). sures such as opening new branches in rural 26. Doing Business (database), International areas or investing in staff training. Over the Finance Corporation and World Bank, Wash- following four years, its lending portfolio to ington, DC, http://www.doingbusiness.org the sector is expected to double to $34 million /data. (World Bank 2013b). 27. The website is at http://www.opic.gov/. 34. NAFIN (Nacional Financiera), a Mexican 28. A key reason for the interest in job creation development bank institution, has been pro- is that jobs are the main source of income for viding reverse factoring services to SMEs the majority of households and a key driver through cadenas productivas (productive of poverty reduction (World Bank 2012c). chains). The main feature of the program 29. There is an overlap between firm age and is that it links small, risky firms with large, size. According to data from the World Bank creditworthy firms that buy from them. enterprise surveys, about 90 percent of young Through cadenas productivas, the small firms firms are SMEs, that is, they have between can use the receivables from their larger cli- 5 and 100 employees, while 10 percent are ents to obtain loans. While not yet exported large firms. See Enterprise Surveys (data- to the agriculture sector, the program shows base), International Finance Corporation and great potential for such future expansion World Bank, Washington, DC, http://www (World Bank 2006a). .enterprisesurveys.org. 35. As defined in Miller and Jones (2010), agricul- 30. The Entrepreneurial Finance Lab website tural value chain finance refers to any or all of (http://www.eflglobal.com/) does not specify the financial services, products, and support the methodology used to obtain these results. services flowing to or through a value chain It is not clear whether borrowers took a psy- to address the needs and constraints of those chometric test before obtaining a loan, and involved in the chain, whether a need to access the statistics represent default rates for these finance, secure sales, procure products, reduce borrowers, or whether borrowers took the risk, or improve efficiency within the chain. test after they had a loan that they had either 36. The Agricultural Finance Program of IFC repaid or defaulted on. The second scenario provides support to financial and nonfinan- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 FINANCIAL INCLUSION FOR FIRMS 149 cial institutions in practices that help mitigate April 28, 2013, http://allafrica.com and manage the risks related to lending in the /stories/201304291357.html. sector along the entire value chain. In addi- 38. See Jesús Escamilla-Juárez and Luisarturo tion, IFC assists farmers and agribusinesses Castellanos-Hernández, “The Usage of Grids in building capacity in many areas, such in Risk Management of Agriculture in Mex- as financial training and links to financial ico,” FARMD, Forum for Agricultural Risk institutions. Management in Development, 2012, https:// 37. See Peer Stein and Denis Salord, “Rwanda: www.agriskmanagementforum.org/content Turning the Tide on Rural Poverty Requires /usage-grids-risk-management-agriculture Innovation,” allAfrica.com, New Times, -mexico. Statistical Appendixes This section consists of three appendixes. to the 2014 Global Financial Development Report. Appendix A presents basic country-by- country data on financial system characteris- Appendix C contains additional country-by- tics around the world. It also presents aver- country information on Islamic banking and ages of the same indicators for peer groups financial inclusion in Organization of Islamic of countries, together with summary maps. It Cooperation (OIC) member countries. It is is an update on information from the 2013 also specific to the 2014 Global Financial Global Financial Development Report. The Development Report. eight indicators in this part remain the same as those shown in the previous report, with These appendixes present only a small the exception of “account at a formal fi nan- part of the Global Financial Development cial institution (%, age 15+),” which replaces Database (GFDD), available at http://www the previously shown “accounts per thousand .worldbank.org/financialdevelopment. The adults, commercial banks.” A key reason 2014 Global Financial Development Report for this change is the higher coverage of the is also accompanied by The Little Data Book newly shown indicator. on Financial Development 2014, which is a pocket edition of the GFDD. It presents Appendix B provides additional country-by- country-by-country and also regional fig- country information on key aspects of finan- ures of a larger set of variables than what are cial inclusion around the world. It is specific shown here. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 151 152 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A BASIC DATA ON FINANCIAL SYSTEM CHARACTERISTICS, 2009–11 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility Afghanistan 7.8 9.0 8.8 Albania 35.7 28.3 6.3 3.2 Algeria 14.4 33.3 6.3 20.3 Andorra 19.0 Angola 18.3 39.2 10.1 12.4 Antigua and Barbuda 77.7 7.3 Argentina 12.7 33.1 3.0 5.2 16.7 29.4 5.1 35.1 Armenia 25.6 17.5 9.6 17.7 1.5 0.3 Aruba 58.8 7.8 20.9 Australia 122.4 99.1 3.1 11.7 166.3 57.0 84.2 22.0 Austria 120.9 97.1 27.4 71.6 36.8 57.8 35.5 Azerbaijan 17.0 14.9 8.3 10.2 Bahamas, The 84.8 2.1 29.2 Bahrain 79.4 64.5 6.1 17.9 87.2 2.5 11.6 Bangladesh 41.2 39.6 5.2 8.1 12.0 146.7 Barbados 81.2 6.1 13.8 119.8 0.4 Belarus 33.3 58.6 0.5 15.8 Belgium 94.3 96.3 6.0 104.0 48.4 25.1 Belize 63.2 6.3 18.4 Benin 22.5 10.5 17.0 Bermuda 15.9 Bhutan 36.4 11.1 37.1 Bolivia 32.9 28.0 9.1 9.4 15.8 0.6 Bosnia and Herzegovina 51.7 56.2 4.6 14.3 12.6 Botswana 25.3 30.3 6.0 14.6 32.2 3.1 7.7 Brazil 50.3 55.9 33.1 21.2 81.7 45.6 68.5 33.8 Brunei Darussalam 39.5 5.0 9.5 Bulgaria 52.8 6.5 16.7 15.2 4.8 26.5 Burkina Faso 17.3 13.4 8.5 Burundi 14.7 7.2 19.3 Cambodia 25.4 3.7 9.7 Cameroon 11.1 14.8 30.5 Canada 95.8 2.4 20.1 142.3 70.4 78.0 25.5 Cape Verde 59.0 7.5 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 153 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 (continued) Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility Cayman Islands 9.0 Central African Republic 7.6 3.3 13.7 Chad 4.9 9.0 17.9 Chile 64.6 42.2 4.0 15.1 145.5 53.4 18.4 19.3 China 118.1 63.8 3.1 19.4 95.2 74.5 188.9 30.9 Colombia 31.2 30.4 6.5 6.5 57.7 22.9 12.6 20.2 Comoros 15.1 21.7 5.2 Congo, Dem. Rep. 3.7 39.8 4.2 Congo, Rep. 5.0 9.0 Costa Rica 46.0 50.4 12.2 28.3 4.4 2.4 Côte d’Ivoire 17.6 19.4 28.2 2.0 Croatia 68.6 88.4 8.3 58.1 40.5 4.6 27.7 Cuba 9.6 Cyprus 272.7 85.2 3.8 25.9 17.9 11.9 Czech Republic 80.7 4.7 14.7 35.0 36.0 32.5 Denmark 99.7 12.3 246.5 78.5 27.7 Djibouti 25.8 12.3 9.4 9.4 Dominica 52.4 6.2 7.8 Dominican Republic 20.9 38.2 8.4 18.4 Ecuador 27.2 36.7 2.1 7.1 7.0 15.4 Egypt, Arab Rep. 33.2 9.7 4.9 39.9 38.1 56.6 45.3 33.2 El Salvador 4.6 13.8 27.1 21.2 1.0 Equatorial Guinea 7.1 16.6 Estonia 99.5 96.8 5.4 8.0 11.2 14.1 26.1 Ethiopia 10.4 Fiji 48.1 2.9 13.8 1.7 Finland 93.3 99.7 12.7 72.0 102.3 28.6 France 112.2 97.0 14.2 123.3 79.9 29.7 Gabon 8.6 18.9 13.5 Gambia, The 13.2 13.4 5.5 Georgia 30.8 33.0 15.5 8.1 6.3 0.3 Germany 108.0 98.1 12.9 68.5 53.7 115.5 27.8 Ghana 13.9 29.4 9.0 9.4 3.2 Greece 109.2 77.9 0.3 44.3 39.2 62.9 36.3 Grenada 79.5 7.6 13.6 Guatemala 23.7 22.3 8.1 18.8 (appendix continued next page) 154 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 (continued) Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility Guinea 5.0 3.7 3.2 Guinea-Bissau 6.7 Guyana 33.6 12.3 18.9 14.0 Haiti 13.2 22.0 14.8 19.5 Honduras 48.7 20.5 9.4 30.0 Hong Kong SAR, China 166.3 88.7 5.0 11.9 465.5 61.8 150.6 32.6 Hungary 72.7 3.3 11.7 25.6 4.1 97.9 35.2 Iceland 110.0 –2.4 84.0 17.5 India 45.5 35.2 40.2 76.7 70.3 85.2 30.7 Indonesia 24.4 19.6 5.6 2.8 37.9 55.6 56.3 27.6 Iran, Islamic Rep. 24.6 73.7 0.1 15.3 57.4 30.7 Iraq 5.9 10.6 21.8 Ireland 225.0 93.9 1.6 142.9 19.0 26.9 33.7 Israel 92.7 90.5 2.9 25.5 83.0 45.6 59.9 24.8 Italy 115.7 71.0 11.6 55.5 38.1 172.8 31.6 Jamaica 26.9 71.0 13.1 3.2 49.4 2.8 12.2 Japan 105.1 96.4 1.1 13.0 107.3 65.9 111.6 28.2 Jordan 70.5 25.5 5.0 45.1 119.8 29.9 28.3 16.7 Kazakhstan 42.2 42.1 –0.8 34.9 5.2 40.4 Kenya 30.9 42.3 9.4 14.6 35.8 7.0 11.0 Korea, Rep. 100.8 93.0 1.8 10.2 154.6 67.0 200.1 26.3 Kosovo 32.3 44.3 Kuwait 66.1 86.8 3.0 17.9 80.6 43.3 14.4 Kyrgyz Republic 3.8 26.8 23.9 1.7 34.7 Lao PDR 17.9 26.8 20.3 4.6 Latvia 89.7 7.3 4.1 5.7 2.4 29.6 Lebanon 67.9 37.0 2.0 52.7 30.5 9.8 17.1 Lesotho 12.5 18.5 7.8 13.2 Liberia 16.4 18.8 10.5 Libya 9.3 3.5 48.1 Lithuania 73.8 4.0 4.2 12.1 6.1 25.2 Luxembourg 184.3 94.6 27.6 172.8 3.7 0.2 Macao SAR, China 52.0 5.2 38.5 Macedonia, FYR 43.0 73.7 2.8 14.7 7.9 6.6 22.3 Madagascar 11.0 5.5 38.0 11.9 Malawi 14.6 16.5 20.8 26.4 26.3 2.3 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 155 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 (continued) Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility Malaysia 106.3 66.2 2.5 30.7 188.6 62.3 30.4 13.5 Maldives 84.2 6.3 Mali 17.8 8.2 13.2 Malta 126.2 95.3 14.8 40.4 6.4 1.2 Mauritania 25.5 17.5 9.8 26.8 Mauritius 84.0 80.1 1.0 21.1 59.0 43.2 7.0 15.8 Mexico 18.0 27.4 4.4 21.9 51.7 35.0 27.2 25.2 Micronesia, Fed. Sts. 14.0 25.9 Moldova 33.3 18.1 7.1 10.3 Mongolia 39.8 77.7 7.6 22.4 12.3 4.6 24.9 Montenegro 71.2 50.4 5.9 85.7 4.1 28.8 Morocco 71.8 39.1 31.4 69.3 27.7 24.2 14.4 Mozambique 22.6 39.9 6.3 2.2 Myanmar 5.0 0.5 Namibia 47.5 4.7 6.8 9.2 2.0 32.0 Nepal 48.9 25.3 4.8 6.2 32.0 2.8 Netherlands 204.2 98.7 4.5 145.0 105.6 29.0 New Zealand 145.1 99.4 2.0 25.6 47.9 44.8 29.8 13.3 Nicaragua 32.1 14.2 9.0 7.7 Niger 12.0 1.5 18.0 Nigeria 29.9 29.7 8.8 0.1 19.5 11.4 23.7 Norway 2.0 21.5 83.1 29.1 106.6 37.0 Oman 40.6 73.6 3.4 13.3 31.4 22.6 23.2 Pakistan 21.0 10.3 6.0 13.4 17.5 52.3 22.3 Panama 79.5 24.9 4.7 41.3 30.5 1.2 10.1 Papua New Guinea 25.1 8.9 8.6 113.4 0.4 Paraguay 32.8 21.7 25.6 11.8 1.9 3.0 Peru 23.5 20.5 17.3 13.1 56.2 36.1 5.0 32.4 Philippines 28.7 26.6 4.5 22.8 58.5 55.9 22.9 23.8 Poland 70.2 9.6 32.2 45.5 52.7 30.3 Portugal 186.9 81.2 16.9 94.9 46.6 24.1 Qatar 41.9 65.9 3.6 26.1 79.1 22.6 27.1 Romania 38.4 44.6 6.0 11.0 16.0 8.3 36.6 Russian Federation 42.1 48.2 5.2 7.0 53.6 36.7 107.1 45.9 Rwanda 32.8 9.6 8.0 Samoa 43.7 7.7 20.4 (appendix continued next page) 156 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 (continued) Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility San Marino 11.3 São Tomé and Príncipe 31.6 17.2 Saudi Arabia 44.8 46.4 14.1 69.8 40.5 88.2 27.4 Senegal 25.2 5.8 40.4 Serbia 46.9 62.2 6.7 16.2 25.2 3.7 28.1 Seychelles 23.1 8.2 15.1 Sierra Leone 8.5 15.3 12.2 4.2 Singapore 99.7 98.2 5.2 27.8 153.8 71.1 85.9 23.8 Slovak Republic 47.9 79.6 19.4 10.4 4.5 23.2 Slovenia 92.0 97.1 4.5 12.1 27.1 20.0 5.9 20.6 Solomon Islands 22.3 11.1 Somalia 31.0 South Africa 72.4 53.6 3.3 8.9 205.4 66.8 55.9 25.1 Spain 209.6 93.3 21.7 138.9 62.5 128.8 31.3 Sri Lanka 25.3 68.5 3.8 12.9 25.4 58.4 21.1 18.8 St. Kitts and Nevis 66.2 4.4 20.5 85.8 1.0 St. Lucia 111.5 7.2 14.5 St. Vincent and the Grenadines 51.4 6.2 Sudan 10.5 6.9 19.7 Suriname 22.6 5.4 14.5 Swaziland 23.2 28.6 6.0 16.5 Sweden 99.0 21.6 156.2 98.7 29.2 Switzerland 165.6 2.7 7.8 223.2 36.0 78.4 22.9 Syrian Arab Republic 19.4 23.3 3.7 7.7 Tajikistan 2.5 15.7 9.0 Tanzania 15.0 17.3 7.7 12.9 5.6 2.6 6.5 Thailand 96.7 72.7 4.8 6.3 77.8 53.0 99.1 25.8 Timor-Leste 12.2 10.2 Togo 21.6 10.2 4.4 Tonga 43.9 7.4 4.3 Trinidad and Tobago 32.8 75.9 7.6 20.8 58.7 1.5 Tunisia 61.2 32.2 23.3 20.2 15.2 11.4 Turkey 38.5 57.6 5.8 31.6 52.4 160.1 31.0 Turkmenistan 0.4 4.3 Uganda 12.3 20.5 10.8 19.0 20.7 0.3 Ukraine 64.6 41.3 6.8 2.6 18.3 8.3 44.8 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 157 TABLE A.1 Countries and Their Financial System Characteristics, Averages, 2009–11 (continued) Financial institutions Financial markets Stock market capitalization Market Account Bank + outstanding capitalization Private credit at a formal lending- domestic excluding top Stock by deposit financial deposit private debt 10 companies market Stock money banks institution spread Bank securities to to total market turnover price Economy to GDP (%) (%, age 15+) (%) Z-score GDP (%) capitalization (%) ratio (%) volatility United Arab Emirates 70.4 59.7 21.2 24.8 48.3 21.4 United Kingdom 202.4 97.2 6.6 132.9 65.9 118.6 24.7 United States 55.7 88.0 26.1 209.9 72.6 240.6 28.2 Uruguay 22.1 23.5 7.4 2.7 0.4 1.6 Uzbekistan 22.5 6.3 Vanuatu 62.4 4.0 15.9 Venezuela, RB 18.8 44.1 3.2 10.1 1.4 1.0 14.3 Vietnam 104.4 21.4 2.4 18.6 16.5 87.0 30.0 West Bank and Gaza 19.4 17.8 21.0 Yemen, Rep. 6.0 3.7 5.8 26.8 Zambia 2.2 21.4 13.4 11.3 17.9 3.6 Zimbabwe 39.7 2.3 ˇ ihák and others 2013. Source: Data from and calculations based on the Global Financial Development Database. For more information, see C Note: Empty cells indicate lack of data. NOTES Visualization: To illustrate where a coun- try’s observation is in relation to the global Table layout: The layout of the table follows distribution of the variable, the table includes the 4x2 matrix of financial system character- four bars on the left of each observation. istics introduced in the 2013 Global Finan- The four-bar scale is based on the location cial Development Report, with four variables of the country in the statistical distribu- approximating depth, access, efficiency, and tion of the variable in the Global Financial stability of financial institutions and financial Development Database: values below the markets, respectively. 25th percentile show only one full bar, val- ues equal to or greater than the 25th and less Additional data: The above table presents a than the 50th percentile show two full bars, small fraction of observations in the Global values equal to or greater than the 50th and Financial Development Database, accom- less than the 75th percentile show three full panying this report. For additional variables, bars, and values greater than the 75th per- historical data, and detailed metadata, see centile show four full bars. The bars are cal- the full data set at http://www.worldbank culated using “winsorized” and “rescaled” .org/financialdevelopment. variables, as described in the 2013 Global Financial Development Report. To prepare for this, the 95th and 5th percentile for each Period covered: The table shows averages for variable for the entire pooled country-year 2009–11. data set are calculated, and the top and bot- tom 5 percent of observations are truncated. Averaging: Each observation is an arithmetic Specifically, all observations from the 5th average of the corresponding variable over percentile to the minimum are replaced by 2009–11. When a variable is not reported the value corresponding to the 5th percen- or not available for a part of this period, the tile, and all observations from the 95th per- average is calculated for the period for which centile to the maximum are replaced by the observations are available. value corresponding to the 95th percentile. 158 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 To convert all the variables to a 0–100 scale, of returns). Return of Assets (ROA), equity, each score is rescaled by the maximum and and assets are country-level aggregate figures the minimum for each indicator. The res- (calculated from underlying bank-by-bank caled indicator can be interpreted as the per- unconsolidated data from Bankscope). cent distance between the worst (0) and the best (100) financial development outcome, Stock market capitalization + outstanding defined by the 5th and 95th percentile of the domestic private debt securities to GDP (%) original distribution (for further informa- measures the market capitalization plus the tion see the 2013 Global Financial Develop- amount of outstanding domestic private debt ment Report). The four bars on the left of the securities as percentage of GDP. Market capi- country name show the unweighted arithme- talization (also known as market value) is the tic average of the “winsorized” and rescaled share price times the number of shares out- variables (dimensions) for each country. This standing. Listed domestic companies are the average is reported only for those countries domestically incorporated companies listed where data for 2009–11 are available for at on the country’s stock exchanges at the end least four variables (dimensions). of the year. Listed companies do not include investment companies, mutual funds, or Private credit by deposit money banks to GDP other collective investment vehicles. Data are (%) measures the domestic private credit to from Standard & Poor’s Global Stock Mar- the real sector by deposit money banks as a kets Factbook and supplemental Standard & percentage of local currency GDP. Data on Poor’s data, and are compiled and reported by domestic private credit to the real sector by the World Development Indicators. Amount deposit money banks are from the Interna- of outstanding domestic private debt securi- tional Financial Statistics (IFS), line 22D, pub- ties is from table 16A (domestic debt amount) lished by the International Monetary Fund of the Securities Statistics by the Bank for (IMF). Local currency GDP is also from IFS. International Settlements. The amount includes all issuers except governments. Account at a formal financial institution (%, age 15+) measures the percentage of adults Market capitalization excluding top 10 com- with an account (self or together with some- panies to total market capitalization (%) one else) at a bank, credit union, another measures the ratio of market capitalization financial institution (e.g., cooperative, micro- outside of the top 10 largest companies to finance institution), or the post office (if total market capitalization. The World Fed- applicable), including adults who report hav- eration of Exchanges (WFE) provides data ing a debit card. The data are from the Global on the exchange level. This variable is aggre- Financial Inclusion (Global Findex) Database gated up to the country level by taking a sim- (Demirgüç-Kunt and Klapper 2012). ple average over exchanges. Bank lending-deposit spread (percentage Stock market turnover ratio (%) is the total points) is lending rate minus deposit rate. value of shares traded during the period Lending rate is the rate charged by banks on divided by the average market capitaliza- loans to the private sector and deposit interest tion for the period. Average market capi- rate is the rate paid by commercial or similar talization is calculated as the average of banks for demand, time, or savings deposits. the end-of-period values for the current The lending and deposit rates are from IFS period and the previous period. Data are lines 60P and 60L, respectively. from Standard & Poor’s Global Stock Markets Factbook and supplemental Standard Bank Z-score is calculated as [ROA + (equity / & Poor’s data, and are compiled and reported assets)] / (standard deviation of ROA). To by the World Development Indicators. approximate the probability that a country’s banking system defaults, the indicator com- Stock price volatility is the 360-day standard pares the system’s buffers (returns and capital- deviation of the return on the national stock ization) with the system’s riskiness (volatility market index. The data are from Bloomberg. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 159 MAP A.1 DEPTH—FINANCIAL INSTITUTIONS To approximate financial institutions’ depth, this published by the International Monetary Fund map uses domestic private credit to the real sector (IMF). Local currency GDP is also from IFS. The by deposit money banks as a percentage of local cur- four shades of blue in the map are based on the aver- rency GDP. Data on domestic private credit to the age value of the variable in 2009–11: the darker the real sector by deposit money banks are from the blue, the higher the quartile of the statistical distribu- International Financial Statistics (IFS), line 22D, tion of the variable. TABLE A.1.1 Depth—Financial Institutions Number of Standard Weighted Private credit by deposit money banks to GDP (%) countries Average Median deviation Minimum Maximum averagea World 163 53.8 36.9 49.1 0.0 284.6 88.6 By developed/developing economies Developed economies 43 107.4 96.3 59.0 6.5 284.6 100.9 Developing economies 120 34.6 26.8 25.3 0.0 121.5 64.1 By income level High income 43 107.4 96.3 59.0 6.5 284.6 100.9 Upper-middle income 46 48.1 43.2 28.8 7.9 121.5 72.3 Lower-middle income 49 31.2 27.8 19.9 0.0 109.1 36.9 Low income 25 17.1 14.3 10.6 3.9 51.1 26.9 By region High income: OECD 25 126.8 111.4 49.5 47.1 237.6 102.0 High income: non-OECD 18 79.5 64.5 60.8 6.5 284.6 77.4 East Asia & Pacific 16 51.2 41.4 34.8 11.5 121.5 105.8 Europe & Central Asia 16 40.4 37.7 13.9 16.5 83.6 41.3 Latin America & the Caribbean 28 39.8 31.4 25.0 4.4 112.6 35.5 Middle East & North Africa 12 36.3 28.1 26.3 3.3 74.5 28.6 South Asia 8 38.8 39.5 22.2 6.8 90.4 42.4 Sub-Saharan Africa 40 20.9 16.6 17.4 0.0 86.7 38.2 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by current GDP. 160 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP A.2 ACCESS—FINANCIAL INSTITUTIONS To approximate access to fi nancial institutions, this Inclusion (Global Findex) Database. The four shades map uses the percentage of adults (age 15+) who of blue in the map are based on the value of the vari- reported having an account at a formal financial insti- able in 2011: the darker the blue, the higher the quar- tution. The data are taken from the Global Financial tile of the statistical distribution of the variable. TABLE A.1.2 Access—Financial Institutions Account at a formal financial institution Number of Standard Weighted (%, age 15+) countries Average Median deviation Minimum Maximum averagea World 147 45.7 38.2 31.7 0.4 99.7 50.6 By developed/developing economies Developed economies 40 87.1 93.2 13.2 46.4 99.7 89.2 Developing economies 107 30.3 25.5 20.9 0.4 89.7 41.9 By income level High income 40 87.1 93.2 13.2 46.4 99.7 89.2 Upper-middle income 40 46.3 44.4 20.1 0.4 89.7 57.1 Lower-middle income 38 24.0 21.4 15.3 3.7 77.7 28.5 Low income 29 16.5 13.4 12.9 1.5 42.3 23.3 By region High income: OECD 28 91.2 96.1 9.5 70.2 99.7 90.5 High income: non-OECD 12 77.4 80.6 15.8 46.4 98.2 66.7 East Asia & Pacific 9 42.0 26.8 27.7 3.7 77.7 55.1 Europe & Central Asia 23 40.7 44.3 24.2 0.4 89.7 44.8 Latin America & the Caribbean 20 32.0 27.7 14.7 13.8 71.0 39.3 Middle East & North Africa 12 26.6 24.4 18.8 3.7 73.7 33.9 South Asia 6 31.3 30.3 22.1 9.0 68.5 33.1 Sub-Saharan Africa 37 21.0 17.5 16.3 1.5 80.1 23.8 Source: Global Financial Development Database, 2011 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by total adult population in 2011. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 161 MAP A.3 EFFICIENCY—FINANCIAL INSTITUTIONS To approximate efficiency of fi nancial institutions, The lending and deposit rates are from IFS, lines 60P this map uses the spread (difference) between lending and 60L, respectively. The four shades of blue in the rate and deposit interest rate. Lending rate is the rate map are based on the average value of the variable in charged by banks on loans to the private sector, and 2009–11: the darker the blue, the higher the quartile deposit interest rate is the rate paid by commercial or of the statistical distribution of the variable. similar banks for demand, time, or savings deposits. TABLE A.1.3 Efficiency—Financial Institutions Number of Standard Weighted Bank lending-deposit spread (percentage points) countries Average Median deviation Minimum Maximum averagea World 126 7.9 6.2 6.6 0.1 49.3 3.7 By developed/developing economies Developed economies 26 4.2 4.3 2.0 1.0 8.6 2.0 Developing economies 100 8.8 6.9 7.0 0.1 49.3 6.2 By income level High income 26 4.2 4.3 2.0 1.0 8.6 2.0 Upper-middle income 43 6.5 5.5 5.3 0.1 35.4 6.2 Lower-middle income 39 8.9 8.0 4.8 1.9 26.8 5.7 Low income 18 14.3 10.7 10.9 3.2 49.3 5.5 By region High income: OECD 13 3.1 2.7 1.4 1.0 6.7 1.8 High income: non-OECD 13 5.3 5.2 1.9 1.7 8.6 5.0 East Asia & Pacific 17 7.2 5.8 4.7 1.9 21.5 3.3 Europe & Central Asia 17 8.2 6.7 6.2 0.1 33.8 5.4 Latin America & the Caribbean 26 9.7 7.6 6.9 1.4 35.4 26.4 Middle East & North Africa 9 4.7 4.5 2.5 0.1 9.7 4.5 South Asia 6 6.2 5.9 2.6 3.0 12.0 5.3 Sub-Saharan Africa 25 11.5 9.1 9.4 0.5 49.3 5.0 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by total banking assets. 162 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP A.4 STABILITY—FINANCIAL INSTITUTIONS To approximate stability of financial institutions, indicator compares the banking system’s buffers this map uses the Z-score for commercial banks. The (returns and capital) with its riskiness (volatility of indicator is estimated as follows: [ROA + (equity / returns). The four shades of blue in the map are based assets)] / (standard deviation of ROA). Return on on the average value of the variable in 2009–11: the equity (ROA), equity, and assets are country-level darker the blue, the higher the quartile of the statisti- aggregate figures (calculated from underlying bank- cal distribution of the variable. by-bank unconsolidated data from Bankscope). The TABLE A.1.4 Stability—Financial Institutions Number of Standard Weighted Bank Z-score countries Average Median deviation Minimum Maximum averagea World 175 15.5 13.6 10.6 -4.5 65.3 15.8 By developed/developing economies Developed economies 54 16.3 14.4 10.1 -3.3 58.4 14.9 Developing economies 121 15.1 13.1 10.9 -4.5 65.3 19.2 By income level High income 54 16.3 14.4 10.1 -3.3 58.4 14.9 Upper-middle income 47 15.2 12.7 12.3 -4.5 65.3 18.0 Lower-middle income 45 17.6 16.8 10.7 -4.1 49.5 30.5 Low income 29 11.2 9.9 6.9 0.1 27.3 2.4 By region High income: OECD 32 14.2 13.0 8.2 -3.3 35.8 14.8 High income: non-OECD 22 19.6 16.7 11.7 1.0 58.4 18.4 East Asia & Pacific 15 14.6 16.4 9.3 0.1 31.8 17.9 Europe & Central Asia 22 9.6 8.5 6.2 -4.5 26.4 7.0 Latin America & the Caribbean 27 15.2 13.6 9.7 2.0 42.3 19.0 Middle East & North Africa 12 28.7 25.3 15.1 6.4 65.3 36.5 South Asia 7 18.1 13.1 14.1 3.6 49.5 37.4 Sub-Saharan Africa 38 13.7 12.7 8.5 -4.1 42.7 9.8 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by total banking assets. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 163 MAP A.5 DEPTH—FINANCIAL MARKETS To approximate depth of financial markets, this map Stock Markets Factbook and supplemental S&P uses market capitalization plus the amount of out- data, and are compiled and reported by the World standing domestic private debt securities as percent- Development Indicators. Amount of outstanding age of GDP. Market capitalization (also known as domestic private debt securities is from table 16A market value) is the share price times the number of (domestic debt amount) of the Securities Statistics by shares outstanding. Listed domestic companies are the Bank for International Settlements. The amount the domestically incorporated companies listed on includes all issuers except governments. The four the country’s stock exchanges at the end of the year. shades of blue in the map are based on the average Listed companies do not include investment compa- value of the variable in 2009–11: the darker the blue, nies, mutual funds, or other collective investment the higher the quartile of the statistical distribution vehicles. Data are from Standard & Poor’s Global of the variable. TABLE A.1.5 Depth—Financial Markets Stock market capitalization plus outstanding Number of Standard Weighted domestic private debt securities to GDP (%) countries Average Median deviation Minimum Maximum averagea World 107 66.5 41.5 69.2 0.4 538.5 123.0 By developed/developing economies Developed economies 45 101.7 82.1 80.5 9.1 538.5 144.9 Developing economies 62 39.7 24.7 43.1 0.4 229.8 71.8 By income level High income 45 101.7 82.1 80.5 9.1 538.5 144.9 Upper-middle income 33 50.0 32.0 51.6 0.4 229.8 77.2 Lower-middle income 22 29.8 18.7 28.4 1.4 141.1 53.2 Low Income 7 20.6 22.9 12.8 1.5 38.4 18.7 By region High income: OECD 32 103.1 93.3 62.2 9.1 259.4 145.7 High income: non-OECD 13 98.3 63.6 114.7 20.4 538.5 124.0 East Asia & Pacific 9 68.2 58.6 57.4 8.4 202.2 89.5 Europe & Central Asia 14 22.6 14.8 23.0 1.4 93.2 40.8 Latin America & the Caribbean 16 34.5 20.7 34.8 0.4 150.0 62.0 Middle East & North Africa 6 53.1 36.5 38.4 15.3 140.8 40.0 South Asia 5 32.7 24.7 24.6 7.6 87.9 66.8 Sub-Saharan Africa 12 41.8 25.3 55.2 5.6 229.8 108.3 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by current GDP. 164 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP A.6 ACCESS—FINANCIAL MARKETS To approximate access to financial markets, this map aggregated up to the country level by taking a simple uses the ratio of market capitalization excluding the average over exchanges. The four shades of blue in top 10 largest companies to total market capitaliza- the map are based on the average value of the vari- tion. The World Federation of Exchanges (WFE) able in 2009–11: the darker the blue, the higher the provides data on the exchange level. This variable is quartile of the statistical distribution of the variable. TABLE A.1.6 Access—Financial Markets Market capitalization excluding top Number of Standard Weighted 10 companies (%) countries Average Median deviation Minimum Maximum averagea World 46 45.7 47.4 19.4 2.8 76.7 64.9 By developed/developing economies Developed economies 25 43.1 43.6 22.4 2.8 76.3 66.2 Developing economies 21 48.7 51.8 15.0 20.7 76.7 61.0 By income level High income 25 43.1 43.6 22.4 2.8 76.3 66.2 Upper-middle income 15 46.6 46.9 15.1 20.7 76.7 60.2 Lower-middle income 6 54.1 56.4 13.5 25.7 72.4 65.1 Low income 0 By region High income: OECD 20 44.0 45.2 21.6 2.8 76.3 66.5 High income: non-OECD 5 39.5 41.5 25.7 5.4 74.3 59.1 East Asia & Pacific 5 60.3 58.8 8.4 51.6 76.7 71.3 Europe & Central Asia 2 44.5 44.6 9.1 32.4 55.1 40.1 Latin America & the Caribbean 6 37.1 35.0 10.5 20.7 55.0 42.1 Middle East & North Africa 4 42.9 40.8 15.1 25.7 60.7 46.9 South Asia 2 64.3 66.0 7.2 53.9 72.4 70.4 Sub-Saharan Africa 2 55.0 49.6 15.5 39.0 74.8 65.5 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by stock market capitalization. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX A 165 MAP A.7 EFFICIENCY—FINANCIAL MARKETS To approximate efficiency of fi nancial markets, this Factbook and supplemental S&P data, and is com- map uses the total value of shares traded during the piled and reported by the World Development Indi- period divided by the average market capitalization cators. The four shades of blue in the map are based for the period. Average market capitalization is cal- on the average value of the variable in 2009–11: the culated as the average of the end-of-period values darker the blue, the higher the quartile of the statisti- for the current period and the previous period. Data cal distribution of the variable. are from Standard & Poor’s Global Stock Markets TABLE A.1.7 Efficiency—Financial Markets Number of Standard Weighted Stock market turnover ratio (%) countries Average Median deviation Minimum Maximum averagea World 106 43.7 17.4 54.7 0.1 347.0 150.7 By developed/developing economies Developed economies 45 65.5 58.0 59.2 0.1 347.0 161.6 Developing economies 61 26.7 5.9 44.1 0.2 226.5 114.5 By income level High income 45 65.5 58.0 59.2 0.1 347.0 161.6 Upper-middle income 33 28.8 7.0 47.5 0.4 226.5 125.2 Lower-middle income 21 21.9 6.2 32.0 0.2 144.5 67.3 Low income 7 30.7 4.1 58.9 0.3 213.9 53.0 By region High income: OECD 32 76.9 73.9 60.1 0.1 347.0 164.7 High income: non-OECD 13 37.2 17.4 46.7 0.3 161.9 106.8 East Asia & Pacific 9 54.6 30.0 63.1 0.2 226.5 163.3 Europe & Central Asia 14 25.7 5.2 48.6 0.2 172.3 107.2 Latin America & the Caribbean 15 11.1 3.0 18.3 0.4 76.3 47.1 Middle East & North Africa 6 24.9 17.3 16.0 4.9 59.2 32.5 South Asia 5 61.6 37.0 60.5 1.7 213.9 83.7 Sub-Saharan Africa 12 8.9 3.3 15.6 0.3 63.8 49.9 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by stock market capitalization. 166 APPENDIX A GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 MAP A.8 STABILITY—FINANCIAL MARKETS To approximate stability of fi nancial markets, this map are based on the average value of the variable in map uses the 360-day standard deviation of the 2009–11: the darker the blue, the higher the quartile return on the national stock market index. Data of the statistical distribution of the variable. are from Bloomberg. The four shades of blue in the TABLE A.1.8 Stability—Financial Markets Number of Standard Weighted Stock price volatility countries Average Median deviation Minimum Maximum averagea World 83 25.3 24.0 10.8 2.4 68.0 29.4 By developed/developing economies Developed economies 38 26.8 26.3 9.3 8.8 51.1 28.9 Developing economies 45 24.0 22.6 11.8 2.4 68.0 31.5 By income level High income 38 26.8 26.3 9.3 8.8 51.1 28.9 Upper-middle income 31 23.9 23.6 12.1 5.8 68.0 31.6 Lower-middle income 12 26.3 24.3 9.9 9.9 52.9 31.2 Low Income 2 8.3 10.9 4.5 2.4 12.5 10.8 By region High income: OECD 29 27.9 27.4 8.7 8.8 51.1 28.9 High income: non-OECD 9 23.2 22.0 10.5 9.5 47.1 30.3 East Asia & Pacific 7 25.2 22.7 8.1 9.2 41.0 30.9 Europe & Central Asia 12 31.0 28.4 13.1 10.7 68.0 38.5 Latin America & the Caribbean 10 21.8 17.0 11.0 5.8 48.4 30.3 Middle East & North Africa 6 19.0 16.2 9.3 8.4 40.2 27.4 South Asia 3 23.9 20.4 9.0 15.6 43.8 32.0 Sub-Saharan Africa 7 17.7 16.5 11.0 2.4 43.1 24.9 Source: Global Financial Development Database, 2009–11 data. Note: OECD = Organisation for Economic Co-operation and Development. a. Weighted average by total value of stocks traded. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX B 167 APPENDIX B KEY ASPECTS OF FINANCIAL INCLUSION TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Afghanistan 9.0 7.4 0.2 4.7 73.1 3.4 1.4 2.5 1.9 Albania 28.3 7.5 3.2 21.1 92.4 42.2 12.4 33.3 22.2 Algeria 33.3 1.5 1.8 13.5 83.8 31.1 8.9 28.6 5.3 Angola 39.2 7.9 17.0 29.8 86.4 9.5 13.1 13.4 10.5 Antigua and Barbuda 100.0 49.2 49.4 46.3 23.4 Argentina 33.1 6.6 5.7 29.8 96.2 49.3 30.3 33.3 13.5 Armenia 17.5 18.9 2.2 5.2 89.5 44.3 31.9 18.8 Aruba 19.5 Australia 99.1 17.0 79.2 79.1 29.6 Austria 97.1 8.3 55.3 86.8 15.2 Azerbaijan 14.9 17.7 0.7 10.0 75.9 19.9 19.0 9.9 Bahamas, The 97.6 34.2 14.6 28.5 38.0 Bahrain 64.5 21.9 6.0 62.2 Bangladesh 39.6 23.3 0.5 2.3 95.3 24.7 43.1 7.8 Barbados 97.4 58.2 45.5 38.7 19.9 Belarus 58.6 16.1 10.4 50.3 92.3 49.5 35.8 2.1 Belgium 96.3 10.5 71.1 85.8 44.0 Belize 100.0 43.9 36.7 57.0 23.2 Benin 10.5 4.2 0.6 0.7 99.2 42.8 4.2 32.9 Bhutan 92.6 58.6 64.2 59.5 16.4 Bolivia 28.0 16.6 0.7 12.8 95.6 49.1 27.8 40.5 9.7 Bosnia and Herzegovina 56.2 13.0 6.2 34.4 99.8 65.0 59.7 31.3 Botswana 30.3 5.6 6.8 15.6 99.0 50.0 32.8 32.1 8.6 Brazil 55.9 6.3 16.6 41.2 99.4 65.3 48.4 60.0 46.2 Brunei Darussalam 23.1 Bulgaria 52.8 7.8 4.6 45.8 96.8 40.2 34.7 58.6 Burkina Faso 13.4 3.1 0.6 2.0 96.8 28.4 25.6 33.1 Burundi 7.2 1.7 0.1 0.8 90.5 35.3 12.3 25.5 2.4 Cambodia 3.7 19.5 0.5 2.9 20.7 11.3 12.6 4.3 Cameroon 14.8 4.5 0.4 2.1 92.5 30.3 31.4 41.6 1.7 Canada 95.8 20.3 69.2 88.0 24.3 Cape Verde 96.5 41.5 35.3 49.8 30.7 (appendix continued next page) 168 APPENDIX B GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 (continued) Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Central African Republic 3.3 0.9 0.1 1.0 98.5 26.0 25.3 25.3 0.9 Chad 9.0 6.2 1.6 5.3 95.9 20.6 4.2 16.1 0.7 Chile 42.2 7.8 11.1 25.8 97.9 79.6 44.8 55.1 17.5 China 63.8 7.3 6.9 41.0 Colombia 30.4 11.9 6.8 22.7 95.8 57.2 35.0 49.2 15.0 Comoros 21.7 7.2 0.4 5.7 Congo, Dem. Rep. 3.7 1.5 0.3 1.7 71.3 10.7 6.7 8.8 Congo, Rep. 9.0 2.8 2.1 3.6 86.7 12.8 7.7 9.7 2.7 Costa Rica 50.4 10.0 14.5 43.8 97.5 56.8 22.2 30.1 23.1 Côte d’Ivoire 67.4 11.5 13.9 8.3 Croatia 88.4 14.4 17.3 74.8 99.8 67.3 60.0 63.2 34.8 Cyprus 85.2 27.0 30.2 46.4 103.9 Czech Republic 80.7 9.5 44.7 61.0 98.1 46.6 33.4 23.1 Denmark 99.7 18.8 85.6 90.1 39.0 Djibouti 12.3 4.5 1.5 7.6 Dominica 100.0 32.8 46.2 37.9 17.7 Dominican Republic 38.2 13.9 4.4 21.3 98.4 56.9 39.1 72.4 10.7 Ecuador 36.7 10.6 4.2 17.1 100.0 48.9 17.0 42.3 Egypt, Arab Rep. 9.7 3.7 0.4 5.1 74.3 17.4 5.6 7.5 El Salvador 13.8 3.9 3.0 10.9 94.7 53.1 31.7 44.5 Equatorial Guinea 4.9 Eritrea 98.2 10.9 11.9 5.7 Estonia 96.8 7.7 74.1 92.3 97.4 50.8 41.5 18.6 Ethiopia 91.8 46.0 10.9 40.7 2.0 Fiji 96.1 37.8 37.1 50.7 11.0 Finland 99.7 23.9 88.2 89.3 15.0 France 97.0 18.6 65.1 69.2 41.6 Gabon 18.9 2.3 3.3 8.6 83.6 9.0 6.3 8.5 5.8 Gambia, The 72.8 16.6 7.6 14.3 8.9 Georgia 33.0 11.0 2.0 20.2 90.8 41.8 38.2 19.6 Germany 98.1 12.5 64.2 88.0 45.0 42.2 Ghana 29.4 5.8 2.9 11.4 83.5 22.2 16.0 21.4 5.5 Greece 77.9 7.9 7.7 34.0 25.9 26.3 38.7 Grenada 98.7 49.0 37.3 50.3 34.5 Guatemala 22.3 13.7 2.6 13.0 61.0 49.1 26.6 26.2 37.1 Guinea 3.7 2.4 0.5 2.3 53.9 6.0 0.9 2.6 1.5 Guinea-Bissau 59.0 2.8 0.7 1.1 Guyana 100.0 50.5 34.5 59.4 7.6 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX B 169 TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 (continued) Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Haiti 22.0 8.3 2.9 2.7 2.7 Honduras 20.5 7.1 1.4 11.1 81.3 31.2 17.0 25.6 21.6 Hong Kong SAR, China 88.7 7.9 51.2 75.8 23.8 Hungary 72.7 9.4 28.7 62.4 97.7 43.0 48.7 15.7 Iceland 52.4 India 35.2 7.7 2.0 8.4 46.6 36.4 10.6 Indonesia 19.6 8.5 3.1 10.5 51.5 18.2 11.7 13.8 8.5 Iran, Islamic Rep. 73.7 30.7 32.9 58.3 29.5 Iraq 10.6 8.0 1.0 3.3 43.2 3.8 2.7 4.6 5.1 Ireland 93.9 15.7 61.5 70.5 37.4 46.1 27.7 Israel 90.5 16.7 54.4 7.5 20.4 Italy 71.0 4.6 27.8 35.2 66.3 Jamaica 71.0 7.9 7.2 41.1 99.8 27.2 44.2 53.1 6.2 Japan 96.4 6.1 44.8 13.0 34.0 Jordan 25.5 4.5 3.4 14.7 94.2 25.5 8.6 18.3 21.1 Kazakhstan 42.1 13.1 4.5 31.3 92.1 33.2 31.0 3.4 Kenya 42.3 9.7 5.4 29.9 89.1 25.4 22.9 26.0 5.2 Kiribati 4.0 Korea, Rep. 93.0 16.6 64.8 57.9 39.9 41.2 18.8 Kosovo 44.3 6.1 5.9 29.0 96.6 15.0 25.3 Kuwait 86.8 20.8 21.9 83.9 19.4 Kyrgyz Republic 3.8 11.3 0.6 1.7 68.9 20.4 17.9 7.3 Lao PDR 26.8 18.1 0.3 6.5 91.8 18.5 0.0 10.7 Latvia 89.7 6.8 52.7 77.8 99.5 48.5 37.3 30.0 Lebanon 37.0 11.3 2.0 21.4 86.7 69.4 23.8 51.3 31.5 Lesotho 18.5 3.0 2.7 14.5 89.7 32.2 32.7 31.9 3.2 Liberia 18.8 6.5 3.6 3.3 67.8 14.0 10.1 12.8 3.8 Lithuania 73.8 5.6 31.5 61.3 98.3 53.0 47.4 Luxembourg 94.6 17.4 67.7 73.2 88.6 Macao SAR, China 37.2 Macedonia, FYR 73.7 10.6 14.1 36.3 96.8 61.1 47.0 24.3 Madagascar 5.5 2.3 0.1 0.9 94.1 20.6 12.2 20.2 1.4 Malawi 16.5 9.2 0.8 9.4 96.9 40.1 20.6 31.0 1.1 Malaysia 66.2 11.2 12.6 23.1 97.7 60.4 48.6 49.3 10.5 Maldives 17.2 Mali 8.2 3.7 0.1 1.8 85.6 16.6 29.3 21.4 Malta 95.3 10.0 34.5 71.2 41.6 Marshall Islands 12.8 (appendix continued next page) 170 APPENDIX B GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 (continued) Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Mauritania 17.5 7.9 2.6 6.3 76.3 16.0 3.2 13.5 Mauritius 80.1 14.3 7.4 50.9 97.2 47.4 37.5 39.5 21.3 Mexico 27.4 7.6 8.3 22.3 61.8 32.0 16.2 26.9 14.9 Micronesia, Fed. Sts. 98.5 43.0 7.2 19.4 14.2 Moldova 18.1 6.4 2.2 16.0 88.2 39.6 30.8 11.3 Mongolia 77.7 24.8 21.6 60.6 61.4 52.9 26.5 66.4 Montenegro 50.4 21.8 3.5 22.0 78.5 49.6 75.8 39.6 Morocco 39.1 4.3 7.4 22.4 86.8 33.4 12.3 30.2 22.3 Mozambique 39.9 5.9 17.3 37.3 75.7 14.2 10.5 8.5 3.6 Myanmar 1.7 Namibia 97.5 24.0 8.1 19.6 7.1 Nepal 25.3 10.8 0.5 3.7 73.7 39.1 17.5 32.1 6.7 Netherlands 98.7 12.6 80.2 97.6 21.5 New Zealand 99.4 26.6 83.2 93.8 34.0 Nicaragua 14.2 7.6 1.5 8.3 75.7 43.4 21.9 18.4 7.4 Niger 1.5 1.3 0.2 0.8 94.0 29.7 9.3 33.4 Nigeria 29.7 2.1 2.4 18.6 3.8 2.7 4.3 6.4 Norway 10.9 Oman 73.6 9.2 17.5 53.0 23.6 Pakistan 10.3 1.6 0.2 2.9 64.7 8.6 9.7 4.6 8.7 Panama 24.9 9.8 3.0 11.3 69.1 20.7 1.1 9.0 23.9 Paraguay 21.7 12.9 4.2 11.3 89.7 60.2 30.1 48.0 9.5 Peru 20.5 12.7 1.9 14.1 87.4 66.8 45.9 49.9 58.7 Philippines 26.6 10.5 2.1 13.2 97.8 33.2 21.9 19.1 8.1 Poland 70.2 9.6 31.4 37.3 95.8 50.1 40.7 32.3 Portugal 81.2 8.3 48.3 68.2 24.4 20.3 64.2 Qatar 65.9 12.6 21.9 49.5 17.8 Romania 44.6 8.4 10.5 27.7 50.4 42.3 37.3 Russian Federation 48.2 7.7 7.7 37.0 98.0 31.3 30.6 37.1 Rwanda 32.8 8.4 0.3 5.3 71.6 46.3 24.2 44.5 5.5 Samoa 97.0 51.3 48.3 68.7 18.4 São Tomé and Príncipe 23.4 Saudi Arabia 46.4 2.1 22.6 42.3 8.7 Senegal 5.8 3.5 0.5 1.8 83.4 15.3 19.8 9.6 Serbia 62.2 12.3 9.6 43.1 100.0 67.6 42.8 9.6 Seychelles 37.2 Sierra Leone 15.3 6.1 1.1 4.0 67.8 17.4 6.9 24.6 3.0 Singapore 98.2 10.0 41.5 28.6 10.2 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX B 171 TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 (continued) Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Slovak Republic 79.6 11.4 43.4 68.3 18.0 42.4 33.5 25.8 Slovenia 97.1 12.8 40.6 91.9 99.9 71.2 52.2 38.3 Solomon Islands 7.1 Somalia 31.0 1.6 21.5 15.6 South Africa 53.6 8.9 13.1 45.3 97.9 30.1 34.8 21.1 10.7 Spain 93.3 11.4 43.4 62.2 32.6 35.8 89.7 Sri Lanka 68.5 17.7 0.5 10.0 89.4 40.4 43.6 40.6 16.7 St. Kitts and Nevis 100.0 49.3 46.4 52.0 37.7 St. Lucia 100.0 24.5 52.2 49.1 22.5 St. Vincent and the Grenadines 98.5 56.5 55.8 52.7 21.2 Sudan 6.9 1.8 2.1 3.3 2.4 Suriname 100.0 44.3 37.0 57.6 11.2 Swaziland 28.6 11.5 4.7 21.0 97.8 21.9 7.7 16.0 7.2 Sweden 99.0 23.4 84.9 95.5 Switzerland 51.0 Syrian Arab Republic 23.3 13.1 3.1 6.2 92.7 37.4 20.7 16.0 Taiwan, China 87.3 9.6 29.2 37.0 Tajikistan 2.5 4.8 0.7 1.8 86.9 33.6 21.4 6.7 Tanzania 17.3 6.6 3.5 12.0 86.2 16.3 6.8 17.3 1.9 Thailand 72.7 19.4 8.6 43.1 99.6 72.5 74.4 71.9 11.3 Timor-Leste 87.8 6.9 1.6 2.6 Togo 10.2 3.8 0.0 1.2 94.2 21.6 16.9 16.9 Tonga 100.0 54.3 33.9 3.0 21.5 Trinidad and Tobago 75.9 8.4 9.3 64.1 99.9 53.7 36.7 63.8 Tunisia 32.2 3.2 2.7 21.0 17.2 Turkey 57.6 4.6 11.1 56.6 90.6 56.8 51.9 18.3 Turkmenistan 0.4 0.8 0.0 0.3 Uganda 20.5 8.9 3.1 10.3 85.8 17.2 7.7 14.0 2.4 Ukraine 41.3 8.1 6.4 33.6 90.2 31.8 32.1 1.6 United Arab Emirates 59.7 10.8 14.8 55.4 14.5 United Kingdom 97.2 11.8 65.3 87.6 United States 88.0 20.1 64.3 71.8 35.4 Uruguay 23.5 14.8 3.2 16.4 90.8 48.6 13.7 26.4 13.7 Uzbekistan 22.5 1.5 4.3 20.4 93.8 10.5 8.2 47.7 Vanuatu 96.0 45.8 41.4 33.2 20.9 Venezuela, RB 44.1 1.7 15.0 35.1 96.5 35.4 35.3 27.1 17.1 Vietnam 21.4 16.2 2.5 14.6 89.4 49.9 21.5 47.0 3.6 West Bank and Gaza 19.4 4.1 1.7 10.7 87.8 18.0 4.2 14.2 (appendix continued next page) 172 APPENDIX B GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 TABLE B.1 Countries and Their Level of Financial Inclusion, 2011 (continued) Individuals Firms (formal sector) Providers Account Loan from Electronic Firms with Firms using Firms using at a formal a fi nancial payments a checking Firms with a banks to banks to Bank fi nancial institution in used to make or savings bank loan/ fi nance fi nance branches institution the past year payments Debit card account line of credit investments working per 100,000 Economy (%, age 15+) (%, age 15+) (%, age 15+) (%, age 15+) (%) (%) (%) capital (%) adults Yemen, Rep. 3.7 0.9 0.6 2.2 31.3 8.1 4.2 6.0 1.8 Zambia 21.4 6.1 3.3 15.7 95.0 16.0 10.2 15.0 4.4 Zimbabwe 39.7 4.9 6.9 28.3 93.5 12.5 13.1 12.8 Source: Data on individuals are from the Global Financial Inclusion (Global Findex) Database, data on firms are from Enterprise Surveys, and data providers are from Financial Access Survey (FAS). Note: Global Findex data pertain to 2011. Data from Enterprise Survey range from 2005 to 2011. Financial Access Survey covers 2001 through 2011. For both the Enterprise Survey and Financial Access Survey, the table shows data from 2011 or the most recent year. Empty cells indicate lack of data. NOTES Electronic payments used to make payments (%, age 15+): Percentage of adults who Additional data. The above table presents a report having made electronic payments small fraction of observations in the Global or that are made automatically, including Findex, the Enterprise Surveys, and Financial wire transfers or payments made online to Access Survey. These data can be accessed at make payments on bills or purchases using Global Findex: http://www.worldbank money from their account. The data are .org/globalfindex from Global Findex (Demirgüç-Kunt and Klapper 2012). Enterprise Survey: http://www.enterprise surveys.org/ Debit card (%, age 15+): Percentage of adults Financial Access http://fas.imf.org/ who report having a debit card where a debit Survey: card is defined as a card that allows a holder to make payments, get money, or make pur- Period covered. The table shows 2011 or chases and the money is taken out of the the most recent data for individuals, formal holder’s bank account right away. The data firms, and providers. are from Global Findex (Demirgüç-Kunt and Klapper 2012). Account at a formal financial institution (%, age 15+): Percentage of adults with an Firms with a checking or savings account (%): account (self or together with someone else) Percentage of fi rms in the survey that report at a bank, credit union, another financial having a checking or savings account. The data institution (e.g., cooperative, microfinance are based on surveys of more than 130,000 institution), or the post office (if applicable) firms spanning 2005 and 2011 and conducted including adults who report having a debit by the World Banks’ enterprise unit. card to total adults. The data are from Global Findex (Demirgüç-Kunt and Klapper 2012). Firms with a bank loan/line of credit (%): Percentage of firms in the survey that report Loan from a financial institution in the past having a loan or a line of credit from a fi nan- year (%, age 15+): Percentage of adults who cial institution. The data are based on sur- report borrowing any money from a bank, veys of more than 130,000 fi rms spanning credit union, microfinance institution, or 2005 and 2011 and conducted by the World another fi nancial institution such as a coop- Banks’ enterprise unit. erative in the past 12 months. The data are from Global Findex (Demirgüç-Kunt and Firms using banks to finance investments Klapper 2012). (%): Percentage of firms in the survey that GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX B 173 report using banks to finance their invest- capital. The data are based on surveys of ment. The data are based on surveys of more more than 130,000 firms spanning 2005 and than 130,000 firms spanning 2005 and 2011 2011 and conducted by the World Banks’ and conducted by the World Banks’ enter- enterprise unit. prise unit. Bank branches per 100,000 adults: Number Firms using banks to finance working capi- of commercial bank branches per 100,000 tal (%): Percentage of firms in the survey that adults. The data are from IMF’s Financial report using banks to fi nance their working Access Survey (FAS). 174 APPENDIX C GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX C ISLAMIC BANKING AND FINANCIAL INCLUSION TABLE C.1 Organization of Islamic Cooperation (OIC) Member Countries, Account Penetration Rates, and Islamic Financial Institutions, 2011 Religiosity and fi nancial inclusion Islamic fi nancial institutions (IFIs) Adults with no Account at a account due Adults with no formal fi nancial to religious account due to Islamic assets Number of Number Religiosity institution reasons religious reasons Number per adult IFIs per of IFIs per Economy (%) (%, age 15+) (%, age 15+) (thousands, age 15+) of IFIs (US$) 10 million adults 10,000 km 2 Afghanistan 97 9.0 33.6 5,830 2 1.1 0.03 Albania 39 28.3 8.3 150 1 4.0 0.36 Algeria 95 33.3 7.6 1,330 2 0.8 0.01 Azerbaijan 50 14.9 5.8 355 1 1.4 0.12 Bahrain 94 64.5 0.0 0 32 29,194 301.6 421.05 Bangladesh 99 39.6 4.5 2,840 12 14 1.2 0.92 Benin 10.5 1.7 77 0 0 0.0 0.00 Burkina Faso 13.4 1.2 98 1 1.1 0.04 Cameroon 96 14.8 1.1 114 2 1.7 0.04 Chad 95 9.0 10.0 573 0 0 0.0 0.00 Comoros 97 21.7 5.8 20 0 0 0.0 0.00 Djibouti 98 12.3 22.8 117 0 0 0.0 0.00 Egypt, Arab Rep. 97 9.7 2.9 1,480 11 146 1.9 0.11 Gabon 18.9 1.5 12 0 0 0.0 0.00 Guinea 3.7 5.0 279 0 0 0.0 0.00 Indonesia 99 19.6 1.5 2,110 23 30 1.3 0.13 Iraq 84 10.6 25.6 4,310 14 98 7.4 0.32 Jordan 25.5 11.3 329 6 1,583 15.4 0.68 Kazakhstan 43 42.1 1.7 126 0 0 0.0 0.00 Kuwait 91 86.8 2.6 7 18 28,102 87.2 10.10 Kyrgyz Republic 72 3.8 7.3 272 0 0 0.0 0.00 Lebanon 87 37.0 7.6 155 4 12.4 3.91 Malaysia 96 66.2 0.1 8 34 4,949 16.8 1.03 Mali 95 8.2 2.8 218 0 0 0.0 0.00 Mauritania 98 17.5 17.7 312 1 76 4.7 0.01 Morocco 97 39.1 26.8 3,810 0 0 0.0 0.00 Mozambique 39.9 2.3 189 0 0 0.0 0.00 Niger 99 1.5 23.6 1,910 0 0 0.0 0.00 Nigeria 96 29.7 3.9 2,520 0 0 0.0 0.00 Oman 73.6 14.2 78 3 14.4 0.10 Pakistan 92 10.3 7.2 7,400 29 40 2.5 0.38 Qatar 95 65.9 11.6 64 14 13,851 86.5 12.08 Saudi Arabia 93 46.4 24.1 2,540 18 1,685 9.2 0.08 Senegal 96 5.8 6.0 411 0 0 0.0 0.00 Sierra Leone 15.3 9.9 287 0 0 0.0 0.00 GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 APPENDIX C 175 TABLE C.1 OIC Member Countries, Account Penetration Rates, and Islamic Financial Institutions, 2011 (continued) Religiosity and fi nancial inclusion Islamic fi nancial institutions (IFIs) Adults with no Account at a account due Adults with no formal fi nancial to religious account due to Islamic assets Number of Number Religiosity institution reasons religious reasons Number per adult IFIs per of IFIs per Economy (%) (%, age 15+) (%, age 15+) (thousands, age 15+) of IFIs (US$) 10 million adults 10,000 km 2 Somalia 31.0 8.9 325 0 0 0.0 0.00 Sudan 93 6.9 4.5 871 29 103 14.0 0.12 Syrian Arab Republic 89 23.3 15.3 1,560 4 18 3.0 0.22 Tajikistan 85 2.5 7.6 329 0 0 0.0 0.00 Togo 10.2 1.2 40 0 0 0.0 0.00 Tunisia 93 32.2 26.8 1,490 3 72 3.7 0.19 Turkey 82 57.6 7.9 1,820 5 538 0.9 0.06 Turkmenistan 80 0.4 9.9 360 0 0 0.0 0.00 Uganda 93 20.5 3.4 485 0 0 0.0 0.00 United Arab Emirates 91 59.7 3.2 84 22 9,298 33.5 2.63 Uzbekistan 51 22.5 5.9 952 0 0 0.0 0.00 West Bank and Gaza 93 19.4 26.7 502 9 0 38.5 14.95 Yemen, Rep. 99 3.7 8.9 1,190 8 179 5.8 0.15 Sources: Calculations based on BankScope, Islamic Development Bank, Gallup Poll, and the Global Financial Inclusion (Global Findex) Database. Note: This project is in progress; data in this table are preliminary and subject to change. OIC = Organization of Islamic Cooperation. Empty cells indicate lack of data. Religiosity (%): Percentage of adults in a Number of IFIs (Islamic financial institu- given country who responded affirmatively to tions): Number of banks in a country that the question, “Is religion an important part offer Shari‘a-compliant fi nancial services to of your daily life?” in a 2010 Gallup poll. their clients. The data are compiled by the Global Financial Development Report team Account at a formal financial institution (%, members. age 15+): Percentage of adults with an account (self or together with someone else) at a bank, Islamic assets per adult (US$): Size of the credit union, another financial institution Islamic assets in the banking sector of an (such as a cooperative or microfinance institu- economy per its adult population. The size of tion), or the post office (if applicable) includ- the Islamic assets is taken from BankScope. ing adults who reported having a debit card to Adult population is taken from World Devel- total adults. The data are from Global Findex opment Indicators. (Demirgüç-Kunt and Klapper 2012). Number of Islamic financial institutions Adults with no account due to religious rea- per 10 million adults: Number of banks sons (%, age 15+): Percentage of those adults in a country that offer Shari‘a-compliant who point to a religious reason for not having fi nancial services to their clients per 10 mil- an account at a formal fi nancial institution. lion adults. The data are compiled by the The data are from Global Findex (Demirgüç- Global Financial Development Report team Kunt and Klapper 2012). members. Adults with no account due to religious rea- Number of Islamic financial institutions per sons (thousands, age 15+): Number of adults 10,000 km2: Number of banks in a country that point to a religious reason for not having that offer Shari‘a-compliant financial services an account at a formal fi nancial institution. to their clients per 10,000 km2. The data are The data are from Global Findex (Demirgüç- compiled by the Global Financial Develop- Kunt and Klapper 2012). ment Report team members. Bibliography Abdel Aziz, Teymour, and Gunhild Berg. 2012. ies Prepared on Behalf of the Global Partnership “Financing Small- and Medium-Sized Enter- for Financial Inclusion, Bangkok, Thailand. prises in Rwanda.” Unpublished working paper, Agarwal, Sumit, John Driscoll, Xavier Gabaix, and World Bank, Washington, DC. David Laibson. 2009. “The Age of Reason: Fi- Abreu, Margarida, and Victor Mendes. 2010. “Fi- nancial Decisions over the Life Cycle and Im- nancial Literacy and Portfolio Diversification.” plications for Regulation.” Brookings Papers on Quantitative Finance 10 (5): 515–28. Economic Activity 2: 51–117. Acemoglu, Daron, and Simon Johnson. 2005. ———. 2011. “Learning in the Credit Card Mar- “Unbundling Institutions.” Journal of Political ket.” NBER Working Paper 13822, National Economy 113 (5): 949–95. Bureau of Economic Research, Cambridge, MA. Adasme, Osvaldo, Giovanni Majnoni, and Myr- Aghion, Philippe, and Patrick Bolton. 1997. “A iam Uribe. 2006. “Access and Risk: Friends or Theory of Trickle-Down Growth and Develop- Foes? Lessons from Chile.” Policy Research ment.” Review of Economics Studies 64 (2): Working Paper 4003, World Bank, Washington, 151–72. DC. Aghion, Philippe, and Steven Durlauf. 2005. Hand- Adelaja, Tai. 2012. “Cash or Credit?” Russia Profile book of Economic Growth. Amsterdam: North- .org, Moscow. http://russiaprofile.org/business Holland Elsevier. /58837.html. Aggarwal, Shilpa, and Leora Klapper. 2012. “De- Adrianova, Svetlana, Panicos Demetriades, and signing Government Policies to Expand Fi- Anja Shortland. 2010. “Is Government Own- nancial Inclusion: Evidence from around the ership of Banks Really Harmful to Growth?” World.” Unpublished working paper, World Discussion Paper 987, German Institute for Eco- Bank, Washington, DC. nomic Research, Berlin, Germany. Ahlin, Christopher and Robert M. Townsend. AFI (Alliance for Financial Inclusion). 2010. “The 2007a. “Using Repayment Data to Test across AFI Survey on Financial Inclusion Policy in De- Models of Joint Liability Lending.” Economic veloping Countries: Preliminary Findings.” AFI Journal 117 (February): F11–51. Policy Note, Bangkok, Thailand. ———. 2007b. “Selection into and across Credit ———. 2011. “Principles for Innovative Financial Contracts: Theory and Field Research.” Journal Inclusion: Bringing the Principles to Life.” Stud- of Econometrics 136 (2): 665–98. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 177 178 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Aker, Jenny, Rachid Boumnijel, Amanda McClel- “The Foundations of Financial Inclusion: Un- land, and Niall Tierney. 2011. “Zap It to Me: derstanding Ownership and Use of Formal Ac- The Short-Term Impacts of a Mobile Cash counts.” Policy Research Working Paper 6290, Transfer Program.” Working Paper 268, Center World Bank, Washington, DC. for Global Development, Washington, DC. Allen, Franklin, and Douglas Gale. 1990. “Incom- Akin, G. Gulsun, Ahmet Faruk Aysan, Serap Ozce- plete Markets and Incentives to Set Up an Op- lik, and Levent Yildiran. 2012. “Credit Card tions Exchange.” The Geneva Papers on Risk Satisfaction and Financial Literacy: Evidence and Insurance Theory 15 (1): 17–46. from an Emerging Market Economy.” Emerg- ———. 1991. “Arbitrage, Short Sales, and Fi- ing Markets Finance & Trade 48 (5): 103–15. nancial Innovation.” Econometrica 59 (4): Akpalu, Wisdom, Samuel Erasmus Alnaa, and 1041–68. Peter Aglobitse. 2012. “Access to Microfi- ———. 2004. “Financial Intermediaries and Mar- nance and Intra-Household Business Decision kets.” Econometrica 72 (4): 1023–61. Making: Implication for Efficiency of Female- Almenberg, Johan, and Jenny Save-Soderbergh. Owned Enterprises in Ghana.” Journal of 2011. “Financial Literacy and Retirement Plan- Socio-Economics 41 (5): 513–18. ning in Sweden.” Journal of Pension Economics Alem, Mauro, and Robert M. Townsend. Forth- and Finance 10 (4): 585–98. coming. “An Evaluation of Financial Institu- Alonso, Javier, Santiago Fernández de Lis, Carmen tions: Impact on Consumption and Investment Hoyo, Carlos López-Moctezuma, and David Using Panel Data and the Theory of Risk- Tuesta. 2013. “Mobile Banking in Mexico as Bearing.” Journal of Econometrics. a Mechanism for Financial Inclusion: Recent Alessie, Rob, Maarten Van Rooij, and Annamaria Developments and a Closer Look into the Po- Lusardi. 2011. “Financial Literacy and Retire- tential Market.” Working Paper 13/20, BBVA ment Preparation in the Netherlands.” Jour- Research, Mexico City, Mexico. nal of Pension Economics and Finance 10 (4): Alvarez de la Campa, Alejandro, ed. 2011. Secured 527–45. Transactions and Collateral Registries. IFC Ad- Ali, Shahla, and Antonio Da Roza. 2012. “Alter- visory Services, International Finance Corpora- native Dispute Resolution Design in Financial tion, Washington, DC. Markets—Some More Equal Than Others: Angelucci, Manuela, Dean Karlan, and Jonathan Hong Kong’s Proposed Financial Dispute Reso- Zinman. 2013. “Win Some, Lose Some? Evi- lution Center in the Context of the Experience dence from a Randomized Microcredit Pro- in the United Kingdom, United States, Austra- gram Placement Experiment by Compartamos lia, and Singapore.” Pacific Rim Law and Pol- Banco.” Working paper, Yale University, New icy Journal 21 (3): 485–531. Haven, CT. Allen, Franklin, Elena Carletti, Robert Cull, Jun Anzoategui, Diego, Aslı Demirgüç-Kunt, and Qian, Lemma Senbet, and Patricio Valenzu- María Soledad Martínez Pería. 2011. “Remit- ela. 2012a. “Improving Access to Banking: tances and Financial Inclusion: Evidence from Evidence from Kenya.” SSRN Working Paper, El Salvador.” Policy Research Working Paper Financial Institutions Center, Wharton School, 5839, World Bank, Washington, DC. University of Pennsylvania, Philadelphia, PA. Aportela, Fernando. 1999. Effects of Financial ———. 2012b. “Resolving the African Financial Access on Savings by Low-Income People. Re- Development Gap: Cross-Country Compari- search Department, Banco de México, Mexico sons and a Within-Country Study of Kenya.” City, Mexico. NBER Working Paper 18013, National Bureau Ardic, Oya Pinar, Gregory Chen, and Alexia Lator- of Economic Research, Cambridge, MA. tue. 2012. Financial Access 2011: An Overview Allen, Franklin, Rajesh Chakrabarti, Sankar De, of the Supply-Side Data Landscape. Washing- Jun Qian, and Meijun Qian. 2006. “Financing ton, DC: Consultative Group to Assist the Poor Firms in India.” Policy Research Working Paper and International Finance Corporation, World 3975, World Bank, Washington, DC. Bank. Allen, Franklin, Aslı Demirgüç-Kunt, Leora Klap- Ardic, Oya Pinar, Joyce Ibrahim, and Natailya Myl- per, and María Soledad Martínez Pería. 2012. enko. 2011. “Consumer Protection Laws and GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 179 Regulations in Deposit and Loan Services: A tant Are Financing Constraints? The Role of Cross-Country Analysis with a New Dataset.” Finance in the Business Environment.” World Policy Research Working Paper 5536, World Bank Economic Review 22 (3): 483–516. Bank, Washington, DC. ———. 2010. “Formal Versus Informal Finance: Armendáriz, Beatriz, and Jonathan Morduch. Evidence from China.” Review of Financial 2010. The Economics of Microfinance. Cam- Studies 23 (8): 3048–97. bridge, MA: MIT Press. ———. 2011a. “Small vs. Young Firms across Ashraf, Nava, Diego Aycinena, Claudia Marti- the World: Contribution to Employment, nez, and Dean Yang. 2011. Remittances and Job Creation, and Growth.” Policy Research the Problem of Control: A Field Experiment Working Paper 5631, World Bank, Washing- Among Migrants from El Salvador. Depar- ton, DC. tamento de Economía, Universidad de Chile, ———. 2011b. “Firm Innovation in Emerging Santiago, Chile. Markets: Role of Finance, Governance, and Ashraf, Nava, Dean Karlan, and Wesley Yin. 2006. Competition.” Journal of Financial Quantita- “Tying Odysseus to the Mast: Evidence from tive Analysis 46 (6): 1545–80. a Commitment Savings Product in the Philip- ———. 2012. “Financing of Firms in Develop- pines.” Quarterly Journal of Economics 121 ing Countries: Lessons from Research.” Policy (2): 635–72. Research Working Paper 6036, World Bank, ———. 2010. “Female Empowerment: Further Washington, DC. (Forthcoming in Handbook Evidence from a Commitment Savings Product of Finance: Empirical Corporate Finance, in the Philippines.” World Development 28 (3): edited by G. Constantinedes, M. Harris, and 333–44. R. Stulz. Amsterdam: North-Holland.) Aterido, Reyes, Thorsten Beck, and Leonardo Bae, Kee-Hong, and Vidhal Goyal. 2009. “Credi- Iacovone. 2011. “Gender and Finance in Sub- tor Rights, Enforcement, and Bank Loans.” Saharan Africa: Are Women Disadvantaged?” Journal of Finance 64 (2): 823–60. Policy Research Working Paper 5571, World Banerjee, Abhijit, Timothy Besley, and Timothy Bank, Washington, DC. (Forthcoming in World Guinnane. 1994. “Thy Neighbor’s Keeper: The Development.) Design of a Credit Cooperative with Theory Aterido, Reyes, and Mary Hallward-Driemeier. and a Test.” Quarterly Journal of Economics 2011. “Whose Business Is It Anyway? Closing 109 (2): 491–515. the Gender Gap in Entrepreneurship in Sub- Banerjee, Abhijit, and Esther Duflo. 2007. “The Saharan Africa.” Small Business Economics 37 Economic Lives of the Poor.” Journal of Eco- (4): 443–64. nomic Perspectives 21 (1): 141–68. Atkinson, Adele. 2008. Evidence of Impact: An ———. 2008. “What Is Middle Class about the Overview of Financial Education Evaluations. Middle Classes around the World?” Journal of University of Bristol, Bristol, U.K. Economic Perspectives 22 (2): 3–28. Attanasio, Orazio, Britta Augsburg, Ralph De ———. 2011. Poor Economics: A Radical Re- Haas, Emla Fitzsimons, and Heike Harmgart. thinking of the Way to Fight Global Poverty. 2011. “Group or Individual Lending? Evidence New York, NY: Public Affairs Press. from a Randomised Field Experiment in Mon- ———. 2012. “Do Firms Want to Borrow More? golia.” Working Paper 136, European Bank Testing Credit Constraints Using a Directed for Reconstruction and Development, London, Lending Program.” Working paper, Massachu- U.K. setts Institute of Technology, Cambridge, MA. Augsburg, Britta, Ralph De Haas, Heike Harm- Banerjee, Abhijit, Esther Duflo, Rachel Glenner- gart, and Costas Meghir. 2012. “Microfinance ster, and Cynthia Kinnan. 2013. “The Miracle at the Margin: Experimental Evidence from of Microfinance? Evidence from a Randomized Bosnia and Herzegovina.” Working Paper 146, Evaluation.” NBER Working Paper 18950, European Bank for Reconstruction and Devel- National Bureau of Economic Research, Cam- opment, London, U.K. bridge, MA. Ayyagari, Meghana, Aslı Demirgüç-Kunt, and Banerjee, Abhijit, and Andrew Newman. 1993. Vojislav Maksimovic. 2008. “How Impor- “Occupational Choice and the Process of De- 180 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 velopment.” Journal of Political Economy 101 Bias in Microfinance?” Discussion Paper 2012– (2): 274–98. 062, Tilburg University Center for Economic Bar-Gill, Oren, and Elizabeth Warren. 2008. Research, Tilburg, the Netherlands. “Making Credit Safer.” University of Pennsyl- Beck, Thorsten, Berrak Buyukkarabacak, Felix vania Law Review 157 (1): 1–101. Rioja, and Neven Valev. 2008. “Who Gets Barber, Brad, and Terrance Odean. 2001. “Boys the Credit? And Does It Matter? Household Will Be Boys: Gender, Overconfidence, and vs. Firm Lending across Countries.” Policy Common Stock Investment.” Quarterly Journal Research Working Paper 4661, World Bank, of Economics 116 (1): 261–92. Washington, DC. Barr, Michael, Sendhil Mullainathan, and Eldar Beck, Thorsten, Robert Cull, Michael Fuchs, Jared Shafir. 2008. “Behaviorally Informed Financial Getenga, Peter Gatere, John Randa, and Mir- Services Regulation.” New America Founda- cea Trandafir. 2011. “Banking Sector Stabil- tion White Paper, New America Foundation, ity, Efficiency, and Outreach in Kenya.” Policy Washington, DC. Research Working Paper 5442, World Bank, Barron, John, and Michael Staten. 2003. “The Washington, DC. Value of Comprehensive Credit Reports: Les- Beck, Thorsten, and Augusto de la Torre. 2007. sons from the U.S. Experience.” In Credit Re- “The Basic Analytics of Access to Financial porting Systems and the International Econ- Services.” Financial Markets, Institutions, and omy, edited by Margaret J. Miller, 273–311. Instruments 16 (2): 79–117. Cambridge, MA: MIT Press. Beck, Thorsten, Aslı Demirgüç-Kunt, Luc Laeven, Basel Committee on Banking Supervision. 2010. and Vojislav Maksimovic. 2006. “The Determi- “Microfinance Activities and the Core Principles nants of Financing Obstacles.” Journal of In- for Effective Banking Supervision.” Bank for In- ternational Money and Finance 25 (6): 932–52. ternational Settlements, Basel, Switzerland. Beck, Thorsten, Aslı Demirgüç-Kunt, and Ross Bauchet, Jonathan, Cristobal Marshall, Laura Levine. 2005. “SMEs, Growth, and Poverty: Starita, Jeanette Thomas, and Anna Yalouris. Cross-Country Evidence.” Journal of Economic 2011. “Latest Findings from Randomized Eval- Growth 10 (3): 199–229. uations of Microfinance.” Access to Finance ———. 2007. “Finance, Inequality, and the Poor.” FORUM, Reports by the Consultative Group Journal of Economic Growth 12 (1): 27–49. to Assist the Poor and Its Partners, World Bank, ———. 2010. “Financial Institutions and Markets Washington, DC. across Countries and over Time.” World Bank Bauchet, Jonathan, and Jonathan Morduch. 2011. Economic Review 24 (1): 77–92. “Is Micro Too Small? Microcredit vs. SME Beck, Thorsten, Aslı Demirgüç-Kunt, and Vojislav Finance.” Wagner Research Paper 2011–2015, Maksimovic. 2005. “Financial and Legal Con- New York University, New York, NY. straints to Firm Growth: Does Firm Size Mat- Baumol, William. 2002. The Free-Market Innova- ter?” Journal of Finance 60 (1): 137–77. tion Machine: Analyzing the Growth Miracle of Beck, Thorsten, Aslı Demirgüç-Kunt, and María Capitalism. Princeton, NJ: Princeton University Soledad Martínez Pería. 2007. “Reaching Out: Press. Access to and Use of Banking Services across Bayer, Patrick, Douglas Bernheim, and John Karl Countries.” Journal of Financial Economics 85 Scholz. 2009. “The Effects of Financial Educa- (1): 234–66. tion in the Workplace: Evidence from a Sur- ———. 2008a. “Bank Financing for SMEs around vey of Employers.” Economic Inquiry 47 (4): the World: Drivers, Obstacles, Business Models, 605–24. and Lending Practices.” Policy Research Work- Beck, Thorsten, Patrick Behr, and Andre Guettler. ing Report 4785, World Bank, Washington, 2012. “Gender and Banking: Are Women Bet- DC. ter Loan Officers?” Review of Finance 17 (4): ———. 2008b. “Banking Services for Everyone? 1279–321. Barriers to Bank Access and Use around the Beck, Thorsten, Patrick Behr, and Andreas Mades- World.” World Bank Economic Review 22 (3): tam. 2012. “Sex and Credit: Is There a Gender 397–430. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 181 ———. 2011. “Bank Financing for SMEs: Evi- Working Paper 6563, World Bank, Washing- dence across Countries and Bank-Ownership ton, DC. Types.” Journal of Financial Services Research Berg, Gunhild, Michael Fuchs, Leonardo Iaco- 39 (1–2): 35–54. vone, Thomas Jaeggi, Andrew Lovegrove, and Beck, Thorsten, Aslı Demirgüç-Kunt, and Dorothe Carolina Sanchez Villegas. 2012. “Financing Singer. Forthcoming. “Is Small Beautiful? Fi- Small- and Medium-Sized Enterprises in Nige- nancial Structure, Size, and Access to Finance.” ria.” Unpublished working paper, World Bank, World Development 52, 19–33. Washington, DC. Beck, Thorsten, Vasso Ioannidou, and Larissa Berg, Gunhild, and Bilal Zia. 2013. “Harnessing Schäfer. 2012. “Foreigners vs. Natives: Bank Emotional Connections to Improve Financial Lending Technologies and Loan Pricing.” Decisions: Evaluating the Impact of Finan- European Banking Center Discussion Paper cial Education in Mainstream Media.” Policy 2012-014, Tilburg University, Tilburg, the Research Working Paper 6407, World Bank, Netherlands. Washington, DC. Beck, Thorsten, Leora Klapper, and Juan Car- Berger, Allen, Adrian Cowan, and W. Scott Frame. los Mendoza. 2010. “The Typology of Partial 2011. “The Surprising Use of Credit Scoring in Credit Guarantee Funds around the World.” Small Business Lending by Community Banks Journal of Financial Stability 6 (1): 10–25. and the Attendant Effects on Credit Availabil- Beck, Thorsten, and Ross Levine. 2004. “Stock ity, Risk, and Profitability.” Journal of Financial Markets, Banks, and Growth: Panel Evidence.” Services Research 39 (1): 1–17. Journal of Banking and Finance 28 (3): 423–42. Berger, Allen, W. Scott Frame, and Nathan Miller. Beck, Thorsten, Ross Levine, and Alexey Levkov. 2005. “Credit Scoring and the Availability, 2010. “Big Bad Banks? The Winners and Losers Price, and Risk of Small Business Credit.” from Bank Deregulation in the United States.” Journal of Money, Credit, and Banking 37 (2): Journal of Finance 65 (5): 1637–67. 191–222. Beck, Thorsten, Chen Lin, and Yue Ma. 2010. Berger, Allen, Leora Klapper, and Gregory Udell. “Why Do Firms Evade Taxes? The Role of 2001. “The Ability of Banks to Lend to Infor- Information Sharing and Financial Sector Out- mationally Opaque Small Businesses.” Journal reach.” Discussion Paper 2010–93, Tilburg of Banking and Finance 25 (12): 2127–67. University Center for Economic Research, Til- Berger, Allen, and Gregory Udell. 1995. “Universal burg, the Netherlands (forthcoming in Journal Banking and the Future of Small Business Lend- of Finance). ing.” Finance and Economics Discussion Series, Beck, Thorsten, Samuel Munzele Maimbo, Issa Board of Governors of the Federal Reserve Sys- Faye, Thouraya Triki. 2011. Financing Africa: tem, Washington, DC. Through the Crisis and Beyond. Washington, ———. 1998. “The Economics of Small Business DC: World Bank. Finance: The Roles of Private Equity and Debt Behrman, Jere R., Olivia S. Mitchell, Cindy Soo, Markets in the Financial Growth Cycle.” Jour- and David Bravo. 2010. “Financial Literacy, nal of Banking and Finance 22 (6–8): 613–73. Schooling and Wealth Accumulation.” NBER ———. 2006. “A More Complete Conceptual Working Paper 16452, National Bureau of Eco- Framework for SME Financing.” Journal of nomic Research, Cambridge, MA. Banking and Finance 30 (11): 2945–66. Benavente, José Miguel, Alexander Galetovic, and Berlin, Mitchell, and Loretta Mester. 1998. “On Ricardo Sanhueza. 2006. “FOGAPE: An Eco- the Profitability and Cost of Relationship Lend- nomic Analysis.” Working Paper 222, Depart- ing.” Journal of Banking and Finance 22 (6–8): ment of Economics, University of Chile, San- 873–97. tiago, Chile. ———. 1999. “Deposits and Relationship Lend- Berg, Gunhild, and Michael Fuchs. 2013. “Bank ing.” Review of Financial Studies 12 (3): 579– Financing of SMEs in Five Sub-Saharan African 607. Countries: The Role of Competition, Innova- Bernheim, Douglas, and Daniel Garrett. 2003. tion, and the Government.” Policy Research “The Effects of Financial Education in the 182 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Workplace: Evidence from a Survey of House- Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, holds.” Journal of Public Economics 87 (7–8): David McKenzie, and John Roberts. 2013. 1487–519. “Does Management Matter? Evidence from Bernheim, Douglas, Daniel Garrett, and Dean India.” Quarterly Journal of Economics 128 Maki. 2001. “Education and Saving: The Long- (1): 1–51. Term Effects of High School Financial Curricu- Bold, Chris. 2011. “Branchless Banking in Paki- lum Mandates.” Journal of Public Economics stan: A Laboratory for Innovation.” CGAP 80 (3): 435–65. Policy Brief, Consultative Group to Assist the Berry, Jim, Dean Karlan, and Menno Pradhan. Poor, World Bank, Washington, DC. 2013. “Social or Financial: What to Focus on Bold, Chris, David Porteous, and Sarah Rotman. in Youth Financial Literacy Training?” Draft 2012. “Social Cash Transfers and Financial In- working paper, Innovations for Poverty Action, clusion: Evidence from Four Countries.” Focus New Haven, CT. Note 77, Consultative Group to Assist the Poor, Berthaud, Alexandre, and Gisela Davico. 2013. World Bank, Washington, DC. Global Panorama on Postal Financial Inclu- Breza, Emily. 2012. “Peer Effects and Loan Re- sion: Business Models and Key Issues. Universal payment: Evidence from the Krishna Default Postal Union, Berne, Switzerland. Crisis.” Job Market Paper, Department of Eco- Bertola, Giuseppe, Richard Disney, and Charles nomics, Massachusetts Institute of Technology, Grant. 2006. The Economics of Consumer Cambridge, MA. Credit. Cambridge, MA: MIT Press. Brix, Laura, and Kate McKee. 2010. “Consumer Bertrand, Marianne, and Adair Morse. 2011. “In- Protection Regulation in Low-Access Environ- formation Disclosure, Cognitive Biases, and ments: Opportunities to Promote Responsible Payday Borrowing.” Journal of Finance 66 (6): Finance.” Focus Note 60, Consultative Group to 1865–93. Assist the Poor, World Bank, Washington, DC. Bertrand, Marianne, Sendhil Mullainathan, and Brown, Gregory, Larry W. Chavis, and Leora Klap- Eldar Shafir. 2004. “A Behavioral Economics per. 2010. “A New Lease on Life: Institutions, View of Poverty.” American Economic Review External Financing, and Business Growth.” 94 (2): 419–23. AFA 2008 New Orleans Meetings Paper, Amer- Beshears, John, James Choi, David Laibson, and ican Finance Association, Duke University, Brigitte Madrian. 2009. “How Does Simpli- Durham, NC. fied Disclosure Affect Individuals’ Mutual Brown, Martin, Tullio Jappelli, and Marco Pa- Fund Choices.” NBER Working Paper 14859, gano. 2009. “Information Sharing and Credit: National Bureau of Economic Research, Cam- Firm-Level Evidence from Transition Coun- bridge, MA. tries.” Journal of Financial Intermediation 18 Besley, Timothy, and Stephen Coate. 1995. “Group (2): 151–72. Lending, Repayment Incentives, and Social Col- Brown, Martin, and Christian Zehnder. 2008. lateral.” Journal of Development Economics 46 “Credit Reporting, Relationship Lending and (1): 1–18. Loan Repayment.” Journal of Money, Credit, Bianchi, Milo, and Matteo Bobba. Forthcoming. and Banking 39 (8): 1883–918. “Liquidity, Risk, and Occupational Choices.” Bruhn, Miriam. 2009. “Female-Owned Firms in Review of Economic Studies. Latin America: Characteristics, Performance, Black, Bernhard, and Ronald Gilson. 1998. “Ven- and Obstacles to Growth.” Policy Research ture Capital and the Structure of Capital Mar- Working Paper 5122, World Bank, Washing- kets: Banks Versus Stock Markets.” Journal of ton, DC. Financial Economics 47 (3): 243–77. ———. 2013. “A Tale of Two Species: Revisiting Blackburn, Keith, Niloy Bose, and Salvatore Ca- the Effect of Registration Reform on Informal passo. 2012. “Tax Evasion, the Underground Business Owners in Mexico.” Journal of Devel- Economy, and Financial Development.” Jour- opment Economics (103): 275–83. nal of Economic Behavior and Organization 83 Bruhn, Miriam, Subika Farazi, and Martin Kanz. (2): 243–53. 2012. “Bank Competition, Concentration, and GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 183 Credit Reporting.” Policy Research Working Social Banking Experiment.” American Eco- Paper 6442, World Bank, Washington, DC. nomic Review 95 (3): 780–95. Bruhn, Miriam, Dean Karlan, and Antoinette Burgess, Robin, Rohini Pande, and Grace Wong. Schoar. 2010. “What Capital Is Missing in De- 2005. “Banking for the Poor: Evidence from veloping Countries?” American Economic Re- India.” Journal of the European Economic As- view: Papers & Proceedings 100 (2): 629–33. sociation 3 (2–3): 268–78. ———. 2013. “The Impact of Consulting Services Cadena, Ximena, and Antoinette Schoar. 2011. on Small and Medium Enterprises: Evidence “Remembering to Pay? Reminders vs. Financial from a Randomized Trial in Mexico.” Policy Incentives for Loan Payments.” NBER Work- Research Working Paper 6508, World Bank, ing Paper 17020, National Bureau of Economic Washington, DC. Research, Cambridge, MA. Bruhn, Miriam, Luciana de Souza Leão, Arianna Cai, Hongbin, Yuyu Chen, Hanming Fang, and Legovini, Rogelio Marchetti, and Bilal Zia. Li-An Zhou. 2009. “Microinsurance, Trust 2013. “Financial Education and Behavior For- and Economic Development: Evidence from a mation: Large-Scale Experimental Evidence Randomized Natural Field Experiment.” PIER from Brazil.” Working paper, World Bank, Working Paper 09–034, Penn Institute for Eco- Washington, DC. nomic Research, University of Pennsylvania, Bruhn, Miriam, and Inessa Love. 2013. “The Eco- Philadelphia, PA. nomic Impact of Expanding Access to Finance ———. 2010. “The Effect of Microinsurance on in Mexico.” In Banking the World: Empirical Economic Activities: Evidence from a Ran- Foundations of Financial Inclusion, edited by domized Natural Field Experiment.” Working Robert Cull, Aslı Demirgüç-Kunt, and Jona- paper, Department of Economics, University of thon Morduch, 137–56. Cambridge, MA: MIT Pennsylvania, Philadelphia, PA. Press. Calomiris, Charles. 2011. “The Mortgage Cri- ———. Forthcoming. “The Real Impact of Im- sis: Some Inside Views.” Wall Street Journal, proved Access to Finance: Evidence from Mex- October 27. ico.” Journal of Finance. Calvet, Laurent, John Campbell, and Paolo Sodini. Bruhn, Miriam, and Bilal Zia. 2013. “Stimulating 2007. “Down or Out: Assessing the Welfare Managerial Capital in Emerging Markets: The Costs of Household Investment Mistakes.” Impact of Business Training for Young Entre- Journal of Political Economy 115 (5): 707–47. preneurs.” Journal of Development Effective- Campbell, John, Howell Jackson, Brigitte Madrian, ness 5(2): 232–66. and Peter Tufano. 2011. “The Regulation of Brune, Lasse, Xavier Giné, Jessica Goldberg, and Consumer Financial Products: An Introductory Dean Yang. 2011. “Commitments to Save: Essay with a Case Study on Payday Lending.” A Field Experiment in Rural Malawi.” Policy In Moving Forward: The Future of Consumer Research Working Paper 5748, World Bank, Credit and Mortgage Finance, edited by Nico- Washington, DC. las Retsinas and Eric Belsky, 206–44 . Harvard ———. 2012. “Reducing Barriers to Savings in University Joint Center for Housing Studies. Rural Malawi: A Field Experiment Approach.” Washington, DC: Brookings Institution Press. Unpublished working paper. Campos, Francisco, Aidan Coville, Ana Fernandes, Bucher-Koenen, Tabea, and Annamaria Lusardi. Markus Goldstein, and David McKenzie. 2013. 2011. “Financial Literacy and Retirement Plan- “Learning from the Experiments That Never ning in Germany.” Journal of Pension Econom- Happened: Lessons from Trying to Conduct ics and Finance 10 (4): 565–84. Randomized Evaluations of Matching Grant Buera, Francisco, Joseph Kaboski, and Yongseok Programs in Africa.” Policy Research Working Shin. 2012. “The Macroeconomics of Microfi- Paper 6296, World Bank, Washington, DC. nance.” NBER Working Paper 19705, National Campos, Francisco, Markus Goldstein, and David Bureau of Economic Research, Cambridge, MA. McKenzie. 2013 “Business Registration Impact Burgess, Robin, and Rohini Pande. 2005. “Do Evaluation in Malawi.” Unpublished paper, Rural Banks Matter? Evidence from the Indian World Bank, Washington, DC. 184 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Carlin, Bruce I. 2009. “Strategic Price Complexity Cheng, Xiaoqiang, and Hans Degryse. 2010. in Retail Financial Markets.” Journal of Finan- “Information Sharing and Credit Rationing: cial Economics 91 (3): 278–87. Evidence from the Introduction of a Public Carpena, Fenella, Shawn Cole, Jeremy Shapiro, Credit Registry.” Discussion Paper 2010-34S, and Bilal Zia. 2011. “Unpacking the Casual Center for Economic Research, Tilburg Univer- Chain of Financial Literacy.” Policy Research sity, Tilburg, the Netherlands. Working Paper 5978, World Bank, Washing- Chiappori, Pierre-Andre, Krislert Samphanth- ton, DC. arak, Samuel Schulhofer-Wohl, and Robert M. ———. 2013. “The ABCs of Financial Literacy: Townsend. Forthcoming. “Heterogeneity and Experimental Evidence on Attitudes, Behavior, Risk-Sharing in Village Economies.” Quantita- and Cognitive Biases.” Unpublished working tive Economics. paper, World Bank, Washington, DC. Chien, Jennifer. 2012. “Designing Disclosure Re- Carter, Michael. 2008. “Inducing Innovation: gimes for Responsible Financial Inclusion.” Risk Instruments for Solving the Conundrum Focus Note 78, Consultative Group to Assist of Rural Finance.” Keynote paper prepared for the Poor, World Bank, Washington, DC. the Agence Française de Développement and Choi, James, David Laibson, and Brigitte Madrian. European Development Network’s 6th Annual 2010. “Why Does the Law of One Price Fail? Conference, Paris, November 12. An Experiment on Index Mutual Funds.” Re- Central Bank of Brazil; Secretariat for Economic view of Financial Studies 15 (5): 1499–524. Monitoring, Ministry of Finance; and Secre- Choi, Woon Gyu, and Yungsan Kim. 2005. “Trade tariat of Economic Law, Ministry of Justice. Credit and the Effect of Macro-Financial 2010. Report on the Brazilian Payment Cards Shocks: Evidence from the U.S. Panel Data.” Industry. Brasília: Department of Banking Op- Journal of Financial and Quantitative Analysis erations and Payments System, Banco Central 40 (4): 897–925. do Brasil. Christen, Robert, and Douglas Pearce. 2005a. CGAP (Consultative Group to Assist the Poor). “Managing Risks and Designing Products 2009. Financial Access 2009: Measuring Access for Agricultural Microfinance: Features of an to Financial Services around the World. Wash- Emerging Model.” Occasional Paper 11, Con- ington, DC: CGAP, World Bank. sultative Group to Assist the Poor, World Bank, ———. 2010a. “Andhra Pradesh 2010: Global Washington, DC. Implications of the Crisis in Indian Microfi- ———. 2005b. “Managing Risks and Designing nance.” Focus Note 67, CGAP, World Bank, Products for Agricultural Microfinance: Fea- Washington, DC. tures of an Emerging Model.” Occasional Paper ———. 2010b. Financial Access 2010: The State 11, Consultative Group to Assist the Poor, of Financial Inclusion through Crisis. Washing- World Bank, Washington, DC. ton, DC: CGAP, World Bank. Christiaensen, Luc, Lionel Demery, and Jesper Chandrasekhar, Arun, Cynthia Kinnan, and Hora- Kuhl. 2011. “The (Evolving) Role of Agricul- cio Larreguy. 2011. “Informal Insurance, Social ture in Poverty Reduction: An Empirical Per- Networks, and Savings Access: Evidence from a spective.” Journal of Development Economics Lab Experiment in the Field.” Working Paper, 96 (2): 239–54. Consortium on Financial Systems and Poverty, Churchill, Craig, and Michal Matul. 2012. Protect- Chicago, IL. ing the Poor: A Microinsurance Compendium, Chavis, Larry, Leora Klapper, and Inessa Love. Volume 2. Geneva, Switzerland: International 2011. “The Impact of the Business Environ- Labour Organization. ment on Young Firm Financing.” World Bank Čihák, Martin. 2012. “Finance: Good, Bad, or Economic Review 25 (3): 486–507. Something in Between? Tell Us What You Chen, Greg, Stephen Rasmussen, and Xavier Think.” All About Finance (blog), blogs.world Reille. 2010. “Growth and Vulnerabilities in bank.org. World Bank, Washington, DC. http:// Microfinance.” Focus Note 61, Consultative blogs.worldbank.org/allaboutfinance/finance Group to Assist the Poor, World Bank, Wash- -good-bad-or-something-in-between-tell-us ington, DC. -what-you-think. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 185 Čihák, Martin, Aslı Demirgüç-Kunt, Erik Feyen, Cole, Harold, Jeremy Greenwood, and Juan San- and Ross Levine. 2013. “Financial Develop- chez. 2012. “Why Doesn’t Technology Flow ment in 205 Economies, 1960 to 2010.” Jour- from Rich to Poor Countries?” Unpublished nal of Financial Perspectives 1 (2): 17–36. working paper, St. Louis Federal Reserve Bank, (Earlier version issued as Policy Research St. Louis. Working Paper 6175, World Bank, Washing- Cole, Shawn. 2009. “Fixing Market Failures or ton, DC). Fixing Elections? Agricultural Credit in India.” Čihák, Martin, Aslı Demirgüç-Kunt, María Sole- American Economic Journal: Applied Econom- dad Martínez Pería, and Amin Mohseni. 2012. ics 1 (1): 219–50. “Banking Regulation and Supervision around Cole, Shawn, and Nilesh Fernando. 2008. “As- the World: A Crisis Update.” Policy Research sessing the Importance of Financial Literacy.” Working Paper 6308, World Bank, Washing- Asian Development Bank: Finance for the Poor ton, DC. 9 (3): 1–7. Čihák, Martin, and Heiko Hesse. 2010. “Islamic Cole, Shawn, Xavier Giné, Jeremy Tobacman, Banks and Financial Stability: An Empirical Petia Topalova, Robert M. Townsend, and Analysis.” Journal of Financial Services Re- James Vickery. 2013. “Barriers to Household search 38 (2): 95–113. Risk Management: Evidence from India.” Čihák, Martin, and Richard Podpiera. 2008. “In- American Economic Journal: Applied Econom- tegrated Financial Supervision: Which Model?” ics 5 (1): 104–35. North American Journal of Economics and Fi- Cole, Shawn, Xavier Giné, and James Vickery. nance 19 (2): 135–52. 2012. “Can Financial Innovation Reduce the Cirasino, Massimo, and José Antonio García. Costs of Weather Shocks? Evidence from Ex 2007. Retail Payment Systems to Support Fi- Ante Agricultural Investment in India.” Unpub- nancial Access: Infrastructure and Policy. Du- lished working paper, Harvard Business School, rango, Mexico: Centro de Estudios Monetarios Cambridge, MA; World Bank, Washington, DC; Latinoamericanos (CELMA) and World Bank. Federal Reserve Bank of New York, New York. Cirmizi, Elena, Leora Klapper, and Mahesh Ut- Cole, Shawn, Martin Kanz, and Leora Klapper. tamchandani. 2012. “The Challenges of Bank- 2012. “Incentivizing Calculated Risk-Taking: ruptcy Reform.” World Bank Research Ob- Evidence from an Experiment with Commercial server 27 (2): 185–203. Bank Loan Officers.” Policy Research Working Claessens, Stijn, and Erik Feyen. 2006. “Access to Paper 6146, World Bank, Washington, DC. Financial Services: A Review of the Issues and Cole, Shawn, Anna Paulson, and Gauri Kartini Public Policy Objectives.” World Bank Re- Shastry. 2012. “Smart Money: The Effect of Ed- search Observer 21 (2): 207–40. ucation on Financial Behavior.” Finance Work- ———. 2007. “Financial Development and the ing Paper 09–071, Harvard Business School, Millenium Development Goals.” World Bank Boston, MA. Working Paper 89, World Bank, Washington, Cole, Shawn, Thomas Sampson, and Bilal Zia. DC. 2009. “Smart Money: The Effect of Education, Claessens, Stijn, Patrick Honohan, and L. Rojas- Cognitive Ability, and Financial Literacy on Fi- Suarez. 2009. “Policy Principles for Expand- nancial Market Participation.” Working Paper ing Financial Access: Report of the CGD Task 09–071, Harvard Business School, Boston, MA. Force on Access to Financial Services.” Center ———. 2011. “Prices or Knowledge? What Drives for Global Development, Washington, DC. Demand for Financial Services in Emerging Clarke, George, Lixin Colin Xu, and Heng-fu Zou. Markets?” Journal of Finance 66 (6): 1933–67. 2006. “Finance and Inequality: What Do the Cole, Shawn, and Gauri Kartini Shastry. 2010. “Is Data Tell Us?” Southern Economic Journal 72 High School the Right Time to Teach Savings (3): 578–96. Behavior? The Effect of Financial Education Coelho, Christiano, João De Mello, and Bruno and Mathematics Courses on Savings.” Un- Funchal. 2012. “The Brazilian Payroll Lending published paper, Wellesley University, Wellesley, Experiment.” Review of Economics and Statis- MA. http://academics.wellesley.edu/Economics tics 94 (4): 925–34. /gshastry/cole-shastry-math.pdf. 186 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Collins, Daryl, Nicola Jentzsch, and Rafael Mazer. ———. 2011. “Microfinance Tradeoffs: Regu- 2011. “Incorporating Consumer Research into lation, Competition, and Financing.” In The Consumer Protection Policy Making.” Focus Handbook of Microfinance, edited by Beztriz Note 74, Consultative Group to Assist the Poor, Armendáriz and Marc Labie, 141–57. Hacken- World Bank, Washington, DC. sack, NJ: World Scientific Publishing. Collins, Daryl, Jonathan Morduch, Stuart Ruther- ———. 2012. Banking the World: Empirical ford, and Orlanda Ruthven. 2009. Portfolios of Foundations of Financial Inclusion. Cambridge, the Poor. Princeton, NJ: Princeton University MA: MIT Press. Press. Cull, Robert, and Kinnon Scott. 2010. “Measuring Costa, Ana, and João De Mello. 2006. “Judicial Household Usage of Financial Services: Does It Risk and Credit Market Performance: Micro Matter How or Whom You Ask?” World Bank Evidence from Brazilian Payroll Loans.” NBER Economic Review 24 (2): 199–233. Working Paper 12252, National Bureau of Eco- Cumming, Douglas, Daniel Schmidt, and Uwe nomic Research, Cambridge, MA. Walz. 2010. “Legality and Venture Govern ance Cotler, Pablo, and Christopher Woodruff. 2007. around the World.” Journal of Business Ventur- “The Impact of Short-Term Credit on Micro- ing 25 (1): 54–72. enterprises: Evidence from the Bimbo Program Dahan, Frederique, and John Simpson. 2008. Se- in Mexico.” Working paper, School of Interna- cured Transactions Reform and Access to Credit. tional Relations and Pacific Studies, University Cheltenham, U.K: Edward Elgar Publishing. of California at San Diego, San Diego, CA. de Andrade, Gustavo Henrique, Miriam Bruhn, Cowan, Kevin, Alejandro Drexler, and Álvaro and David McKenzie. 2013. “A Helping Hand Yañez. 2009. “The Effect of Credit Insurance or the Long Arm of the Law? Experimental Evi- on Liquidity Constraints and Default Rates: dence on What Governments Can Do to For- Evidence from a Governmental Intervention.” malize Firms.” Policy Research Working Paper Working Paper 524, Central Bank of Chile, 6435, World Bank, Washington, DC. Santiago, Chile. De Giorgi, Giacomo, and Aminur Rahman. 2013. Crabtree, Steve. 2010. Religiosity Highest in “SME’s Registration: Evidence from an RCT World’s Poorest Nations. Gallup Global Re- in Bangladesh.” Economics Letters 120 (3): ports, August 31. http://www.gallup.com/ 573–78. poll/142727/religiosity-highest-world-poorest Deininger, Klaus, and Aparajita Goyal. 2012. -nations.aspx. “Going Digital: Credit Effects of Land Registry Crépon, Bruno, Florencia Devoto, Esther Duflo, Computerization in India.” Journal of Develop- and William Parienté. 2011. “Impact of Mi- ment Economics 99 (2): 236–43. crocredit in Rural Areas of Morocco: Evidence de Janvry, Alain, Craig McIntosh, and Elisabeth from a Randomized Evaluation.” Unpublished Sadoulet. 2010. “The Supply- and Demand- working paper, Massachusetts Institute of Tech- Side Impacts of Credit Market Information.” nology, Cambridge, MA. Journal of Development Economics 93 (2): Crossan, Diana, David Feslier, and Roger Hur- 173–88. nard. 2011. “Financial Literacy and Retirement De Koker, Louis, and Jentzsch, Nicola. 2011. Planning in New Zealand.” Journal of Pension “Financial Inclusion and Financial Integ- Economics and Finance 10 (4): 619–35. rity: Aligned Incentives?” Shadow 2011: The Cull, Robert, Aslı Demirgüç-Kunt, and Timothy Shadow Economy, Tax Evasion, and Money Lyman. 2012. “Financial Inclusion and Stabil- Laundering: Proceedings of the 2011 Shadow ity: What Does Research Show?” Brief (May), Conference, University of Münster, Münster, Consultative Group to Assist the Poor, World Germany. Bank, Washington, DC. de la Torre, Augusto, Juan Carlos Gozzi, and Ser- Cull, Robert, Aslı Demirgüç-Kunt, and Jonathan gio Schmukler. 2007. “Innovative Experiences Morduch. 2009. “Microfinance Meets the Mar- in Access to Finance: Market-Friendly Roles for ket.” Journal of Economic Perspectives 23 (1): the Visible Hand?” Policy Research Working 167–92. Paper 4326, World Bank, Washington, DC. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 187 de la Torre, Augusto, María Soledad Martínez ———. Forthcoming. “The Demand for, and Con- Pería, and Sergio Schmukler. 2010a. “Bank sequences of, Formalization among Informal Involvement with SMEs: Beyond Relationship Firms in Sri Lanka.” American Economic Jour- Lending.” Journal of Banking and Finance 34 nal: Applied Economics. (9): 2280–93. De Meza, David, Bernd Irlenbusch, and Diane ———. 2010b. “Drivers and Obstacles to Bank- Reyniers. 2008. “Financial Capability: A Be- ing SMEs: The Role of Competition and the havioural Economics Perspective.” Consumer Institutional Framework.” In Reforming Rules Research 69, Financial Services Authority, Lon- and Regulations: Laws, Institutions, and Im- don, U.K. plementation, edited by Vivek Ghosal, 215– De Meza, David, and David Webb. 1987. “Too 53. CESifo Seminar Series. Cambridge, MA: Much Investment: A Problem of Asymmetric MIT Press. Information.” Quarterly Journal of Economics de la Torre, Jordi, Xavier Giné, and Tara Vish- 102 (2): 281–92. wanath. 2011. “After the Microfinance Crisis: Demirgüç-Kunt, Aslı, Thorsten Beck, and Patrick Assessing the Role of Government-Led Micro- Honohan. 2008. “Finance for All? Policies and credit Alternatives.” Policy Research Working Pitfalls in Expanding Access.” World Bank Pol- Paper 5808, World Bank, Washington, DC. icy Research Report, World Bank, Washington, de Mel, Suresh, David McKenzie, and Christopher DC. Woodruff. 2008. “Returns to Capital in Mi- Demirgüç-Kunt, Aslı, Erik Feyen, and Ross Levine. croenterprises: Evidence from a Field Experi- 2011. “The Evolving Importance of Banks and ment.” Quarterly Journal of Economics 123 Securities Markets.” Policy Research Working (4): 1329–72. Paper 5805, World Bank, Washington, DC. ———. 2009a. “Innovative Firms, or Innova- Demirgüç-Kunt, Aslı, and Leora Klapper. 2012. tive Owners? Determinants of Innovation in “Measuring Financial Inclusion: The Global Micro, Small, and Medium Enterprises.” Policy Findex.” Policy Research Working Paper 6025, Research Working Paper 4934, World Bank, World Bank, Washington, DC. Washington, DC. Demirgüç-Kunt, Aslı, Leora Klapper, and Doug- ———. 2009b. “Are Women More Credit Con- las Randall. Forthcoming. “Islamic Finance strained? Experimental Evidence on Gender and Financial Inclusion: Measuring Use of and and Microenterprise Returns.” American Eco- Demand for Formal Financial Services among nomic Journal: Applied Economics 1 (3): 1–32. Muslim Adults.” Policy Research Working ———. 2010. “Who Are the Microenterprise Paper, World Bank, Washington, DC. Owners? Evidence from Sri Lanka on Tokman Demirgüç-Kunt, Aslı, Leora Klapper, and Dorothe versus de Soto.” In International Differences in Singer. 2013. “Financial Inclusion and Legal Entrepreneurship, edited by Josh Lerner and Discrimination against Women: Evidence from Antoinette Schoar, 63–87. Chicago: University Developing Countries.” Policy Research Work- of Chicago Press. ing Paper 6416, World Bank, Washington, DC. ———. 2011. “Getting Credit to High Return Demirgüç-Kunt, Aslı, and Ross Levine. 2008. “Fi- Microentrepreneurs: The Results of an Infor- nance, Financial Sector Policies, and Long Run mation Intervention.” World Bank Economic Growth.” M. Spence Growth Commission Back- Review 25 (3): 456–85. ground Paper 11, World Bank, Washington, DC. ———. 2012a. “One-Time Transfers of Cash or ———. 2009. “Finance and Inequality: Theory Capital Have Long-Lasting Effects on Micro- and Evidence.” Annual Review of Financial enterprises in Sri Lanka.” Science 335 (6071): Economics (1): 287–318. 962–66. Demirgüç-Kunt, Aslı, and Vojislav Maksimovic. ———. 2012b. “Business Training and Female 1998. “Law, Finance, and Firm Growth.” Jour- Enterprise Start-Up, Growth, and Dynamics: nal of Finance 53 (6): 2107–37. Experimental Evidence from Sri Lanka.” Policy ———. 1999. “Institutions, Financial Markets, Research Working Paper 6145, World Bank, and Firm Debt Maturity.” Journal of Financial Washington, DC. Economics 54 (3): 295–336. 188 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ———. 2001. “Firms as Financial Intermediar- Working Paper 0066, Center for Economic and ies: Evidence from Trade Credit Data.” Policy Financial Research, Moscow. Research Working Paper 2696, World Bank, Doi, Yoko, David McKenzie, and Bilal Zia. Washington, DC. 2012. “Who You Train Matters: Identifying ———. 2002. “Funding Growth in Bank-Based Complementary Effects of Financial Educa- and Market-Based Financial Systems: Evidence tion on Migrant Households.” Policy Re- from Firm-Level Data.” Journal of Financial search Working Paper 6157, World Bank, Economics 65 (3): 337–63. Washington, DC. Demombynes, Gabriel, and Aaron Thegeya. 2012. Drexler, Alejandro, Greg Fischer, and Antoinette “Kenya’s Mobile Revolution and the Promise Schoar. 2011. “Keeping It Simple: Financial of Mobile Savings.” Policy Research Working Literacy and Rules of Thumb.” Working paper, Paper 5988, World Bank, Washington, DC. Massachusetts Institute of Technology, Cam- Department of the Treasury, United States. 2011. bridge, MA. “Banking on Opportunity: A Scan of the Evolv- Drugov, Mikhail, and Rocco Macchiavello. 2008. ing Field of Bank on Initiatives.” Office of Fi- “Learning and Microlending.” CEPR Discus- nancial Education and Financial Access, U.S. sion Paper DP7011, Centre for Economic Pol- Department of the Treasury, Washington, DC. icy Research, London, U.K. http://www.treasury.gov/resource-center/finan Duflo, Esther, and Emmanuel Saez. 2003. “The cial-education/documents/banking%20on%20 Role of Information and Social Interactions in opportunity%20nov%2011.pdf. Retirement Plan Decisions: Evidence from a Diamond, Douglas. 1984. “Financial Interme- Randomized Experiment.” Quarterly Journal diation and Delegated Monitoring.” Review of of Economics 118 (3): 815–42. Economic Studies 51 (3): 393–414. Duflos, Eric, and Jasmina Glisovic-Mézières. 2008. Didier, Tatiana, and Sergio Schmukler. 2013. “The “National Microfinance Strategies.” Brief Financing and Growth of Firms in China and 45100, Consultative Group to Assist the Poor, India.” Policy Research Working Paper 6401, World Bank, Washington, DC. World Bank, Washington, DC. Dupas, Pascaline, Sarah Green, Anthony Keats, Di Maro, Vincenzo, Aidan Coville, Siegfried Zot- and Jonathan Robinson. 2012. “Challenges tel, and Felipe Alexander Dunsch. 2013. “The in Banking the Rural Poor: Evidence from Impact of Financial Literacy through Feature Kenya’s Western Province.” NBER Working Films: Evidence from a Randomized Experi- Paper 17851, National Bureau of Economic ment in Nigeria.” Unpublished paper, Financial Research, Cambridge, MA. Literacy and Education, Russia Trust Fund, Dupas, Pascaline, and Jonathan Robinson. 2011. World Bank, Washington, DC. “Why Don’t the Poor Save More? Evidence Ding, Lei, Roberto G. Quercia, and Janneke Rat- from Health Savings Experiments.” NBER cliffe. 2008. “Post-Purchase Counseling and De- Working Paper 17255, National Bureau of Eco- fault Resolutions among Low- and Moderate- nomic Research, Cambridge, MA. Income Borrowers.” Journal of Race, Ethnicity, ———. 2013. “Savings Constraints and Micro- and Religion 30 (3): 315–44. enterprise Development: Evidence from a Field Di Tella, Rafael, Sebastian Galiani, and Ernesto Experiment in Kenya.” American Economic Schargrodsky. 2007. “The Formation of Beliefs: Journal: Applied Economics 5 (1): 163–92. Evidence from the Allocation of Land Titles to Duygan-Bump, Burcu, and Charles Grant. 2009. Squatters.” Quarterly Journal of Economics “Household Debt Repayment Behaviour: What 122 (1): 209–41. Role Do Institutions Play?” Economic Policy Djankov, Simeon, Caralee McLiesh, and Andrei 24 (57): 107–40. Shleifer. 2007. “Private Credit in 129 Coun- Dwyer, Peggy, James Gilkeson, and John List. tries.” Journal of Financial Economics 84 (2): 2002. “Gender Differences in Revealed Risk 299–329. Taking: Evidence from Mutual Fund Inves- Djankov, Simeon, Yingyi Qian, Gerard Roland, tors.” Economics Letters 76 (2): 151–58. and Ekaterina Zhuravskaya. 2006. “Entrepre- Edelberg, Wendy. 2004. “Testing for Adverse Se- neurship in Brazil, China, and Russia.” CEFIR lection and Moral Hazard in Consumer Loan GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 189 Markets.” Working Paper 09/2004, Federal Re- Feyen, Erik. 2010. “Finances of Egyptian Listed serve Board, Washington, DC. Firms and the Performance of the Egyptian Ehrbeck, Tilman, Mark Pickens, and Michael Stock Exchange.” Policy Research Working Tarazi. 2012. “Financially Inclusive Ecosys- Paper 5213, World Bank, Washington, DC. tems: The Roles of Government Today.” Focus Field, Erica, Rohini Pande, John Papp, and Natalia Note 76, Consultative Group to Assist the Poor, Rigol. Forthcoming. “Does the Classic Micro- World Bank, Washington, DC. finance Model Discourage Entrepreneurship El-Zoghbi, Mayada, and Meritxell Martinez. among the Poor? Evidence from India.” Ameri- 2012. “Can Postal Networks Advance Finan- can Economic Review. cial Inclusion in the Arab World?” Brief (April), Financial Access Initiative, Ideas42, and Interna- Consultative Group to Assist the Poor, World tional Finance Corporation. 2011. “Emergency Bank, Washington, DC. (Hand) Loan.” http://www.microfinancegate El-Zoghbi, Mayada, and Michael Tarazi. 2013. way.org/gm/document-1.9.56808/Emergency “Trends in Shari‘a-Compliant Financial Inclu- %20Hand%20Loan.pdf. sion.” CGAP Focus Note 84, Conslutative Financial Action Task Force. 2013. “Revised Guid- Group to Assist the Poor, Washington, DC. ance on Anti–Money Laundering/Combating Evans, David, and Boyan Jovanovic. 1989. “An Terrorist Financing and Financial Inclusion.” Estimated Model of Entrepreneurial Choice Financial Action Task Force, Paris, France. under Liquidity Constraints.” Journal of Politi- Financial Services Authority. 2012. “Regulatory cal Economy 97 (4): 808–27. Reform: PRA and FCA Regimes Relating to As- Evans, David, and Richard Schmalensee. 1999. pects of Authorization and Supervision.” http:// Paying with Plastic: The Digital Revolution in www.fsa.gov.uk/static/pubs/cp/cp12-24.pdf. Buying and Borrowing. Cambridge, MA: MIT Financial Stability Board. 2011. “Consumer Fi- Press. nance Protection with Particular Focus on Evdokimov, Maxim. 2013. “From a Cash Econ- Credit.” http://www.financialstabilityboard.org omy to Digital Payments: Russia at the Cross- /publications/r_111026a.pdf. roads.” East-West Digital News. http://www Fischer, Greg. Forthcoming. “Contract Struc- .ewdn.com/2013/01/25/from-a-cash-economy ture, Risk Sharing, and Investment Choice.” -to-digital-payments-russia-at-the-crossroads/. Econometrica. Fabbri, Daniela, and Mario Padula. 2004. “Does Fishback, Bo, Christine Gulbranson, Robert Litan, Poor Legal Enforcement Make Households Lesa Mitchell, and Marisa Porzig. 2007. “Find- Credit-Constrained?” Journal of Banking and ing Business ‘Idols’: A New Model to Accelerate Finance 28 (10): 2369–97. Start-Ups.” Unpublished working paper, Kauff- Farazi, Subika. 2013. “Informal Firms and Finan- man Foundation, Kansas City, MO. cial Inclusion: Status and Determinants.” Un- Fisman, Raymond, Daniel Paravasini, and Vikrant published working paper, World Bank, Wash- Vig. 2012. “Cultural Proximity and Loan Out- ington, DC. comes.” Unpublished working paper, London Faz, Xavier, and Ted Moser. 2013. “Advancing Fi- School of Economics, London, U.K. nancial Inclusion through Use of Market Arche- Fleisig, Heywood, Mehnaz Safavian, and Nuria de types.” Focus Note 86, Consultative Group to la Peña. 2006. Reforming Collateral Laws to Assist the Poor, World Bank, Washington, DC. Expand Access to Finance. World Bank, Wash- Federal Deposit Insurance Corporation. 2009. ington, DC. “FDIC National Survey of Unbanked and Un- Fletcher, Matthew, Rachel Freeman, Murat Sul- derbanked Households.” http://www.fdic.gov tanov, and Umedjan Umarov. 2005. Leasing in /householdsurvey/. Development: Lessons from Emerging Econo- Fernandes, Daniel, John G. Lynch, and Richard mies. International Finance Corporation, Wash- Netemeyer. Forthcoming. A Meta-Analytic and ington, DC. Psychometric Investigation of the Effect of Fi- Fletschner, Diana. 2008. “Women’s Access to nancial Literacy on Downstream Financial Credit: Does it Matter for Household Effi- Behaviors. National Endowment for Financial ciency?” American Journal of Agricultural Eco- Education, Denver, CO. nomics 90 (3): 669–83. 190 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ———. 2009. “Rural Women’s Access to Credit: Land Titling.” Journal of Public Economics 94 Market Imperfections and Intrahousehold Dy- (9–10): 700–29. namics.” World Development 37 (3): 618–31. ———. 2011a. “The Dynamics of Land Titling Fletschner, Diana, and Dina Mesbah. 2011. “Gen- Regularization and Market Development.” der Disparity in Access to Information: Do Working Paper 2011/88, United Nations Spouses Share What They Know?” World De- University–World Institute for Development velopment 39 (8): 1422–33. Economics Research, Helsinki, Finland. Flores-Roux, Ernesto, and Judith Mariscal. 2010. ———. 2011b. “Land Property Rights and Re- “The Enigma of Mobile Money Systems.” source Allocation.” Journal of Law and Eco- Communications and Strategies 79: 41–62. nomics 54 (4): S329–45. Fonseca, Raquel, Kathleen J. Mullen, Gema Za- Galindo, Arturo, and Margaret Miller. 2001. “Can marro, and Julie Zissimopoulos. 2012. “What Credit Registries Reduce Credit Constraints? Em- Explains the Gender Gap in Financial Literacy? pirical Evidence on the Role of Credit Registries The Role of Household Decision Making.” Jour- in Firm Investment Decisions.” Paper presented nal of Consumer Affairs Spring 46 (1): 90–106. at Inter-American Development Bank’s 42nd Fornero, Elsa, and Chiara Monticone. 2011. “Fi- Annual Meeting, Santiago, Chile, March 19–21. nancial Literacy and Pension Plan Participation Galindo, Arturo, and Fabio Schiantarelli. 2002. in Italy.” Journal of Pension Economics and Fi- “Credit Constraints in Latin America: An nance 10 (4): 547–64. Overview of Micro Evidence.” Boston College Foster, Andrew, and Mark Rosenzweig. 2003. Working Papers in Economics 537, Boston Col- “Agricultural Development, Industrialization, lege, Boston, MA. and Rural Inequality.” Unpublished paper, Har- Galor, Oded, and Joseph Zeira. 1993. “Income vard University, Cambridge, MA. Distribution and Macroeconomics.” Review of Fracassi, Cesar, Mark Garmaise, Shimon Kogan, Economics Studies 60 (1): 35–52. and Gabriel Natividad. 2013. “How Much Gaurav, Sarthak, and Ashish Singh. 2012. “An Does Credit Matter for Entrepreneurial Success Inquiry into the Financial Literacy and Cogni- in the United States?” Unpublished working tive Ability of Farmers: Evidence from Rural paper, University of California at Los Angeles, India.” Oxford Development Studies 40 (3): Los Angeles, CA. 358–80. Freixas, Xavier. 1991. “El mercado hipotecario es- Gerardi, Kristopher, Lorenz Goette, and Stephen pañol: Situación actual y proyecto de reforma.” Meier. 2010. “Financial Literacy and the Sub- Fundación de Estudios de Economía Aplicada, prime Mortgage Delinquency: Evidence from Madrid, Spain. a Survey Matched to Administrative Data.” Fuchs, Michael, Leonardo Iacovone, Thomas Working Paper 2010–10, Federal Reserve Bank Jaeggi, Mark Napier, Roland Pearson, Giulia of Atlanta, Atlanta, GA. Pellegrini, and Carolina Villegas Sanchez. 2011. Ghatak, Maitreesh. 1999. “Group Lending, Local “Financing Small and Medium Enterprises in Information, and Peer Selection.” Journal of the Republic of South Africa.” World Bank, Development Economics 60 (1): 27–50. Washington, DC. ———. 2000. “Screening by the Company You Gabaix, Xavier, and David Laibson. 2005. Keep: Joint Liability Lending and the Peer Se- “Shrouded Attributes, Consumer Myopia, and lection Effect.” Economic Journal 110 (465): Information Suppression in Competitive Mar- 601–31. kets.” Quarterly Journal of Economics 12 (2): Ghatak, Maitreesh, and Timothy Guinnane. 1999. 505–40. “The Economics of Lending with Joint Liabil- Gale, William G., and Ruth Levine. 2010. “Finan- ity: Theory and Practice.” Journal of Develop- cial Literacy: What Works? How Could It Be ment Economics 60 (1): 195–228. More Effective?” Rand Corporation Working Ghosh, Jayati. 2012. “Microfinance and the Chal- Paper, Rand Corporation, Arlington, VA. lenge of Financial Inclusion for Development.” Galiani, Sebastian, and Ernesto Schargrodsky. Paper presented at the Annual Conference of 2010. “Property Rights for the Poor: Effects of the Central Bank of Argentina, October. http:// GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 191 www.bcra.gov.ar/pdfs/eventos/JMB_2012 Case Study of the Indian Rainfall Index Insur- _Ghosh.pdf. ance Market.” Policy Research Working Paper Ghosh, Parikshit, and Debraj Ray. 2001. “Infor- 5459, World Bank, Washington, DC. mation and Enforcement in Informal Credit Giné, Xavier, and Dean Yang. 2009. “Insurance, Markets.” Working Paper (November), New Credit, and Technology Adoption: Field Experi- York University, New York. http://web.cenet mental Evidence from Malawi.” Journal of De- .org.cn/upfile/57373.pdf. velopment Economics 89 (1): 1–11. Gibson, John, David McKenzie, and Bilal Zia. Glisovic, Jasmina, and Meritxell Martinez. 2012. 2012. “The Impact of Financial Literacy Train- “Financing Small Enterprises: What Role for ing for Migrants.” Policy Research Work- Microfinance Institutions?” Focus Note 81, ing Paper 6073, Impact Evaluation Series 57, Consultative Group to Assist the Poor, World World Bank, Washington, DC. Bank, Washington, DC. Giné, Xavier. 2009. “The Promise of Index Insur- Goldstein, Markus, and Christopher Udry. 2008. ance.” Finance and PSD Impact, No. 3, Lessons “The Profits of Power: Land Rights and Agri- from DECRG-FP Impact Evaluations, World cultural Investment in Ghana.” Journal of Po- Bank, Washington, DC. litical Economy 116 (6): 981–1022. Giné, Xavier, Jessica Goldberg, and Dean Yang. Gonzalez-Vega, Claudio, Mark Schreiner, Richard 2012. “Credit Market Consequences of Im- Meyer, Jorge Rodriguez, and Sergio Navajas. proved Personal Identification: Field Experi- 1997. “The Challenge of Growth for Micro- mental Evidence from Malawi.” American Eco- finance Organizations: The Case of Banco nomic Review 102 (6): 2923–54. Solidario in Bolivia.” In Microfinance for the Giné, Xavier, Pamela Jakiela, Dean Karlan, and Poor?, edited by Hartmut Schneider, 129–70. Jonathan Morduch. 2010. “Microfinance Paris: Organisation for Economic Co-operation Games.” American Economic Journal: Applied and Development. Economics 2 (3): 60–95. GPFI (Global Partnership for Financial Inclusion). Giné, Xavier, and Martin Kanz. 2013. “The Eco- 2010. “G20 Financial Inclusion Action Plan.” nomic Effects of a Borrower Bailout: Evidence http://www.gpfi.org/our-work/work-plans from an Emerging Market.” Working paper, /g20-financial-inclusion-action-plan. World Bank, Washington, DC. ———. 2011a. “G20 Principles for Innovative Fi- Giné, Xavier, and Dean Karlan. 2009. “Group Ver- nancial Inclusion.” Alliance for Financial Inclu- sus Individual Liability: Long-Term Evidence sion, Bangkok, Thailand. http://www.gpfi.org from Philippine Microcredit Lending Groups.” /knowledge-bank/publications?page=1. Economic Department Working Paper 61, Yale ———. 2011b. “Financial Inclusion Data: Assess- University, New Haven, CT. ing the Landscape and Country-Level Target Giné, Xavier, and Stefan Klonner. 2005. “Credit Approaches.” International Finance Corpora- Constraints as a Barrier to Technology Adop- tion, Washington, DC. http://www.gpfi.org/sites tion by the Poor: Lessons from South Indian /default/files/documents/WORKING DATA.pdf. Small-Scale Fishery.” Policy Research Working ———. 2011c. “Strengthening Access to Finance Paper 3665, World Bank, Washington, DC. for Women-Owned SMEs in Developing Coun- Giné, Xavier, and Ghazala Mansuri. 2012. “Money tries.” International Finance Corporation, or Ideas? A Field Experiment on Constraints to Washington, DC. http://www.gpfi.org/sites Entrepreneurship in Rural Pakistan.” Unpub- /default/files/documents/Strengthening.pdf. lished working paper, World Bank, Washing- ———. 2011d. “Scaling-Up Access to Finance ton, DC. for Agricultural SMEs.” http://www.gpfi.org Giné, Xavier, Cristina Martínez de Cuellar, and Ra- /knowledge-bank/publications. fael Mazer. 2013. “The Quality of Financial Ser- ———. 2011e. “Scaling-Up SME Access to Finan- vices: Evidence from an Audit Study in Mexico.” cial Services in the Developing World.” http:// Working paper, World Bank, Washington, DC. www.gpfi.org/knowledge-bank/publications. Giné, Xavier, Lev Menand, Robert M. Townsend, ———. 2011f. “SME Finance Policy Guide.” http:// and James Vickery. 2010. “Microinsurance: A www.gpfi.org/knowledge-bank/publications. 192 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ———. 2011g. “Global Standard-Setting Bodies sumption Choice.” Review of Economic Dy- and Financial Inclusion for the Poor.” http:// namics 7 (2): 243–64. www.gpfi.org/knowledge-bank/publications. Gutierrez, Eva, Heinz Rudolph, Theodore Homa, ———. 2011h. “GPFI Report on the G-20 and Enrique Blanco Beneit. 2011. “Develop- Leaders.” http://www.gpfi.org/knowledge-bank ment Banks: Role and Mechanisms to Increase /publications. Their Efficiency.” Policy Research Working ———. 2011i. “Toward Universal Access: Ad- Paper 5729, World Bank, Washington, DC. dressing the Global Challenge of Financial In- Gutierrez, Eva, and Sandeep Singh. 2013. “What clusion.” http://www.gpfi.org/knowledge-bank Regulatory Frameworks Are More Conducive /publications. to Mobile Banking? Empirical Evidence from ———. 2011j. “Principles and Report on Innova- FINDEX.” Policy Research Working Paper tive Financial Inclusion.” http://www.gpfi.org 5729, World Bank, Washington, DC. /knowledge-bank/publications. Hall, Bronwyn H., and Josh Lerner. 2010. “The ———. 2012. “Financial Inclusion—A Pathway Financing of R&D and Innovation.” In Hand- to Financial Stability? Understanding the Link- book of the Economics of Innovation, edited ages.” http://www.gpfi.org/knowledge-bank by Bronwyn H. Hall and Nathan Rosenberg, /publications. 609–39. Amsterdam: Elsevier-North Holland. Green, Anke. 2003. “Credit Guarantee Schemes Hallward-Driemeier, Mary, and Tazeen Hasan. for Small Enterprises: An Effective Instru- 2012. Empowering Women: Legal Rights and ment to Promote Private Sector-Led Growth?” Economic Opportunities in Africa. Africa De- UNIDO SME Technical Working Paper Series velopment Forum series. Washington, DC: 10, United Nations Industrial Development Or- World Bank. ganization, Vienna, Austria. Haltiwanger, John, Ron Jarmin, and Javier Miranda. Green, Edward. 1987. “Lending and the Smooth- 2010. “Who Creates Jobs? Small vs. Large vs. ing of Uninsurable Income.” In Contractual Ar- Young.” NBER Working Paper 16300, National rangements for Intertemporal Trade, edited by Bureau of Economic Research, Cambridge, MA. Edward Prescott and Neil Wallace, 3–25. Min- Haselmann, Rainer, and Paul Wachtel. 2010. “In- neapolis, MN: University of Minnesota Press. stitutions and Bank Behavior: Legal Environ- Gross, David, and Nicholas Souleles. 2002. “Do ment, Legal Perception, and the Composition of Liquidity Constraints and Interest Rates Matter Bank Lending.” Journal of Money, Credit, and for Consumer Behavior? Evidence from Credit Banking 42 (5): 965–84. Card Data.” Quarterly Journal of Economics Hashemi, Syed M., and Aude de Montesquiou. 117 (1): 149–85. 2011. “Reaching the Poorest: Lessons from the Group of 20 Financial Inclusion Experts. 2010. Graduation Model.” Focus Note 69, Consul- “Innovative Financial Inclusion: Principles tative Group to Assist the Poor, World Bank, and Report on Innovative Financial Inclusion Washington, DC. from Access through Innovation Sub-Group of Hastings, Justine S., Brigitte C. Madrian, and Wil- the G20 Financial Inclusion Experts Group.” liam L. Skimmyhorn. 2012. “Financial Literacy, ATISG Report, Canberra, Australia. Financial Education and Economic Outcomes.” Gu, Yiquan, and Tobias Wenzel. 2011. “Strategic NBER Working Paper 18412, National Bureau Obfuscation and Consumer Protection Policy of Economic Research, Cambridge, MA. in Financial Markets: Theory and Experimen- Hathaway, Ian, and Sameer Khatiwada. 2008. tal Evidence.” DICE Discussion Paper 76, Düs- “Do Financial Education Programs Work?” seldorf Institute for Competition Economics, Working Paper 08–03, Federal Reserve Bank of Heinrich-Heine-Universität Düsseldorf, Düs- Cleveland, Cleveland, OH. seldorf, Germany. http://www.webmeets.com He, Xin, Jeffrey Inman, and Vikas Mittal. 2008. /files/papers/earie/2012/308/complexity_4 “Gender Jeopardy in Financial Risk Taking.” .pdf. Journal of Marketing Research 45 (4): 414–24. Gul, Faruk, and Wolfgang Pesendorfer. 2004. Heidhues, Paul, and Botond Koszegi. 2010. “Ex- “Self-Control, Revealed Preference and Con- ploiting Naivete about Self Control in the GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 193 Credit Market.” American Economic Review ———. 2010b. Scaling-Up SME Access to Finan- 100 (5): 2279–303. cial Services in the Developing World. IFC, Henderson, Brian J., and Neil D. Pearson. 2011. Washington, DC. “The Dark Side of Financial Innovation: A Case ———. 2012a. Credit Reporting Knowledge Study of the Pricing of a Retail Financial Prod- Guide. IFC, Washington, DC. uct.” Journal of Financial Economics 100 (2): ———. 2012b. Why Banks in Emerging Markets 227–47. Are Increasingly Providing Non-Financial Ser- Höllinger, Frank, Lamon Rutten, and Krassimir vices to Small and Medium Enterprises. IFC, Kiriakov. 2009. The Use of Warehouse Receipt Washington, DC. Finance in Agriculture in Transition Countries. ———. 2012c. Innovative Agricultural SME Fi- FAO Investment Centre, Food and Agriculture nance Models. Global Partnership for Financial Organization of the United Nations, Rome, Italy. Inclusion and IFC, Washington, DC. Holzmann, Robert, Florentina Mulaj, and Valeria Inderst, Roman, and Marco Ottaviani. 2012. “Fi- Perotti. 2013. “Financial Capability in Low- nancial Advice.” Journal of Economic Litera- and Middle-Income Countries: Measurement ture 50 (2): 494–512. and Evaluation.” Working paper, Financial Lit- International Association of Insurance Supervi- eracy and Education, Russia Trust Fund, World sors. 2012. “Application Paper on Regulation Bank, Washington, DC. and Supervision Supporting Inclusive Insur- Honohan, Patrick. 2003. “Avoiding the Pitfalls ance Markets.” Microinsurance Network Joint in Taxing Financial Intermediation.” Policy Working Group, International Association of Research Working Paper 3056, World Bank, Insurance Supervisors, Basel, Switzerland. Washington, DC. Jack, William, and Tavneet Suri. 2011. “Mobile ———. 2004. “Financial Sector Policy and the Money: The Economics of M-Pesa.” NBER Poor: Selected Issues and Evidence.” World Working Paper 16721, National Bureau of Eco- Bank Working Paper 43, World Bank, Wash- nomic Research, Cambridge, MA. ington, DC. Jacoby, Hanan, Guo Li, and Scott Rozelle. 2002. ———. 2008. “Cross-Country Variation in House- “Hazards of Expropriation: Tenure Insecurity hold Access to Financial Services.” Journal of and Investment in Rural China.” American Banking and Finance 32 (11): 2493–500. Economic Review 92 (5): 1420–47. ———. 2009. “Household Financial Assets in the Jain, Anil, and Robert M. Townsend. 2013. “A Process of Development.” Chapter 13 in Per- Model of Two-Sided Markets in a Walrasian sonal Wealth from a Global Perspective, edited Framework: Payments, Platforms, and Pric- by James B. Davies. New York, NY: Oxford ing.” Unpublished paper, Massachusetts Insti- University Press. tute of Technology, Cambridge, MA. ———. 2010. “Partial Credit Guarantees: Princi- Jakiela, Pamela, and Owen Ozier. 2011. “Does Af- ples and Practice.” Journal of Financial Stability rica Need a Rotten Kin Theorem? Experimen- 6 (1): 1–9. tal Evidence from Village Economies.” Policy Hsieh, Chang-Tai, and Peter Klenow. 2012. “The Research Working Paper 6085, World Bank, Life Cycle of Plants in India and Mexico.” Washington, DC. Unpublished working paper, University of Jappelli, Tullio, and Marco Pagano. 2002. “Infor- Chicago, Chicago, IL. mation Sharing, Lending, and Defaults: Cross- Humphrey, David, and Robert Hunt. 2012. “Get- Country Evidence.” Journal of Banking and ting Rid of Paper: Savings from Check.” Federal Finance 26 (10): 2017–45. Reserve Bank of Philadelphia, Philadelphia, PA. ———. 2006. “The Role and Effects of Credit IASB (International Accounting Standards Board). Information Sharing.” In The Economics of 2012. A Guide to the IFRS for SMEs. IASB, Consumer Credit, edited by Giuseppe Bertola, London, U.K. Richard Disney, and Charles Grant, 347–72. IFC (International Finance Corporation). 2010a. Cambridge, MA: MIT Press. The SME Banking Knowledge Guide. IFC, Jappelli, Tullio, Marco Pagano, and Magda Bi- Washington, DC. anco. 2005. “Courts and Banks: Effects of Judi- 194 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 cial Enforcement on Credit Markets.” Journal Cambridge, MA. http://www.povertyactionlab of Money, Credit, and Banking 37 (2): 223–44. .org/evaluation/savings-account-labeling-ghana. Jayaratne, Jith, and Philip Strahan. 1996. “The Karlan, Dean, Margaret McConnell, Sendhil Mul- Finance-Growth Nexus: Evidence from Bank lainathan, and Jonathan Zinman. 2010. “Get- Branch Deregulation.” Quarterly Journal of ting to the Top of Mind. How Reminders In- Economics 111 (3): 639–70. crease Saving.” NBER Working Paper 16205, ———. 1998. “Entry Restrictions, Industry Evo- National Bureau of Economic Research, Cam- lution, and Dynamic Efficiency: Evidence from bridge, MA. Commercial Banking.” Journal of Law and Karlan, Dean, Melanie Morten, and Jonathan Zin- Economics 41 (1): 239–74. man. 2012. “A Personal Touch: Text Messaging Jeng, Leslie, and Philippe Wells. 2000. “The De- and Loan Repayment.” NBER Working Paper terminants of Venture Funding: Evidence across 17952, National Bureau of Economic Research, Countries.” Journal of Corporate Finance 6 (1): Cambridge, MA. 241–89. Karlan, Dean, Robert Darko Osei, Isaac Osei- Johnson, Simon, and James Kwak. 2012. “Is Fi- Akoto, Christopher Udry. 2012. “Agricultural nancial Innovation Good for the Economy?” Decisions after Relaxing Credit and Risk Re- Innovation Policy and the Economy 12 (1): straints.” Unpublished working paper, Yale 1–16. University, New Haven, CT. Johnston, Don, and Jonathan Morduch. 2008. Karlan, Dean, Isaac Osei-Akoto, Robert Osei, and “The Unbanked: Evidence from Indonesia.” Christopher Udry. 2012. “Examining Underin- World Bank Economic Review 22 (3): 517–37. vestment in Agriculture: Measuring Returns to Kaboski, Joseph, and Robert M. Townsend, 2005. Capital and Insurance.” Innovations for Pov- “Policies and Impact: An Evaluation of Village- erty Action, New Haven, CT. Level Microfinance Institutions.” Journal of the Karlan, Dean, and Jonathan Zinman. 2009. “Ob- European Economic Association 3 (1): 1–50. serving Unobservables: Identifying Information ———. 2011. “A Structural Evaluation of a Large- Asymmetries with a Consumer Credit Field Scale Quasi-Experimental Microfinance Initia- Experiment.” Econometrica 77 (6): 1993–2008. tive.” Econometrica 79 (5): 1357–406. ———. 2010. “Expanding Credit Access: Using ———. 2012. “The Impact of Credit on Village Randomized Supply Decisions to Estimate the Economies.” American Economic Journal: Ap- Impacts.” Review of Financial Studies 23 (1): plied Economics 4 (2): 98–133. 433–64. Kanz, Martin. 2012. “What Does Debt Relief Do ———. 2011. “Microcredit in Theory and Prac- for Development? Evidence from India’s Bail- tice: Using Randomized Credit Scoring for Im- out Program for Highly Indebted-Rural House- pact Evaluation.” Science 332 (6035): 1278–84. holds.” Policy Research Working Paper 6258, Keeton, William R. 1979. Equilibrium Credit Ra- World Bank, Washington, DC. tioning. New York: Garland Press. Karim, Nimrah, Michael Tarazi, and Xavier Reille. ———. 1995. “Multi-Office Bank Lending to 2008. “Islamic Microfinance: An Emerging Small Businesses: Some New Evidence.” Fed- Market Niche” CGAP Focus Note 49, Con- eral Reserve Bank of Kansas City Economic sultative Group to Assist the Poor, Washington, Review 80 (2): 45–57. DC. Kempson, Elaine, Valeria Perotti, and Kinnon Karlan, Dean. 2007. “Social Connections and Scott. 2013. “Measuring Financial Capabil- Group Banking.” Economic Journal 117 (517): ity: Questionnaires and Implementation Guid- F52–84. ance for Low- and Middle-Income Countries.” Karlan, Dean, Edward Kutsoati, Margaret Mc- Working paper, Financial Literacy and Educa- Connell, Margaret McMillan, and Christopher tion, Russia Trust Fund, World Bank, Washing- Udry. 2012. “Savings Account Labeling in ton, DC. Ghana.” Abdul Latif Jameel Poverty Action Kendall, Jake, Natalia Mylenko, and Alejandro Lab, Massachusetts Institute of Technology, Ponce. 2010. “Measuring Financial Access GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 195 around the World.” Policy Research Working Kloeppinger-Todd, Renate, and Mahohar Sharma, Paper 5253, World Bank, Washington, DC. eds. 2010. “Innovations in Rural and Agricul- Kerr, William, Josh Lerner, and Antoinette Schoar. ture Finance.” Focus Note 18, International Forthcoming. “The Consequences of Entrepre- Food Policy Research Institute and the World neurial Finance: Evidence from Angel Financ- Bank, Washington, DC. ings.” Review of Financial Studies. Kugler, Maurice, and Reza Ofoghi. 2005. “Does Khandker, Shahidur. 2005. “Microfinance and Insurance Promote Economic Growth? Evi- Poverty: Evidence Using Panel Data from Ban- dence from the U.K.” Working paper, Univer- gladesh.” World Bank Economic Review 19 sity of Southampton, Southampton, U.K. (2): 263–86. Kumhof, Michael, and Romain Ranciere. 2010. Khandker, Shahidur, Hussain Samad, and Rubaba “Inequality, Leverage, and Crises.” IMF Work- Ali. 2013. “Does Access to Finance Matter in ing Paper 10/268, International Monetary Microenterprise Growth? Evidence from Ban- Fund, Washington, DC. gladesh.” Policy Research Working Paper 6333, ———. 2011. “Unequal = Indebted.” Finance and World Bank, Washington, DC. Development 48 (3): 25–27. Kim, Youngmi, and Michael Sherraden. 2011. Laeven, Luc. 2003. “Does Financial Liberaliza- “Do Parental Assets Matter for Children’s Edu- tion Reduce Financing Constraints?” Financial cational Attainment? Evidence from Mediation Management 32 (1): 5–34. Tests.” Children and Youth Services Review 33 Laeven, Luc, and Fabian Valencia. 2012. “Systemic (6): 969–79. Banking Crisis Database: An Update.” IMF King, Robert, and Ross Levine. 1993. “Finance and Working Paper 08/224, International Monetary Growth: Schumpeter Might Be Right.” Quar- Fund, Washington, DC. terly Journal of Economics 108 (3): 717–37. Laibson, David. 1997. “Golden Eggs and Hyper- Kinnan, Cynthia, and Robert M. Townsend. 2012. bolic Discounting.” Quarterly Journal of Eco- “Kinship and Financial Networks, Formal Fi- nomics 112 (2): 443–77. nancial Access, and Risk Reduction,” American La Porta, Rafael, Florencio Lopez-de-Silanes, Economic Review Papers and Proceedings 102 Andrei Shleifer, and Robert Vishny. 1997. (3): 289–93. “Legal Determinants of External Finance.” Klapper, Leora. 2006. “The Role of Factoring for Journal of Finance 52 (3): 1131–50. Financing Small and Medium Enterprises.” Jour- nal of Banking and Finance 30 (11): 3111–30. ———. 1998. “Law and Finance.” Journal of Po- Klapper, Leora, Luc Laeven, and Raghuram Rajan. litical Economy 106 (6): 1113–55. 2006. “Entry Regulation as a Barrier to Entre- Larraín, Christian, and Jorge Quiroz. 2006. “Eval- preneurship.” Journal of Financial Economics uación de la Adicionalidad del Fondo de Ga- 82 (3): 591–629. rantías de Pequeñas Empresas.” Working paper, Klapper, Leora, and Inessa Love. 2011. “The Im- BancoEstado, Santiago, Chile. pact of the Financial Crisis on New Firm Regis- Lelarge, Claire, David Sraer, and David Thesmar. tration.” Economic Letters 113 (1): 1–4. 2008. “Entrepreneurship and Credit Con- Klapper, Leora, and Giogios Panos. 2011. “Finan- straints: Evidence from a French Loan Guar- cial Literacy and Retirement Planning: The Rus- antee Program.” Unpublished working paper, sian Case.” Journal of Pension Economics and National Bureau of Economic Research, Cam- Finance 10 (4): 599–618. bridge, MA. Klapper, Leora, and Simon Parker. 2011. “Gender Lerner, Josh, and Antoinette Schoar. 2005. “Does and the Business Environment for New Firm Legal Enforcement Affect Financial Transac- Creation.” World Bank Research Observer 26 tions?” Quarterly Journal of Economics 120 (2): 237–57. (1): 223–46. Klein, Daniel. 1992. “Promise Keeping in the Great Levine, Ross. 1998. “The Legal Environment, Society: A Model of Credit Information Shar- Banks, and Long-Run.” Journal of Money, ing.” Economics and Politics 4 (2): 117–36. Credit, and Banking 30 (3): 596–613. 196 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 ———. 1999. “Law, Finance, and Economic Evidence from Recent Financial Crises.” Jour- Growth.” Journal of Financial Intermediation nal of Financial Economics 83 (2): 453–69. 8 (1–2): 8–35. Love, Inessa, and Susana Sanchez. 2009. “Credit ———. 2002. “Bank-Based or Market-Based Fi- Constraints and Investment Behavior in Mex- nancial Systems: Which Is Better?” Journal of ico’s Rural Economy.” Policy Research Work- Financial Intermediation 11 (4): 398–428. ing Paper 5014, World Bank, Washington, ———. 2005. “Finance and Growth: Theory and DC. Evidence.” In Handbook of Economic Growth, Lusardi, Annamaria, Robert Clark, Jonathan Fox, Vol. 1, edited by Philippe Aghion and Steven John Grable, and Ed Taylor. 2010. “Promising Durlauf, 865–934. Amsterdam, the Nether- Learning Strategies, Interventions, and Delivery lands: North-Holland Elsevier. Methods in Financial Literacy Education: What Levine, Ross, and Sara Zervos. 1998. “Stock Mar- Techniques, Venues, Tactics, Mechanisms, kets, Banks, and Economic Growth.” American etc., Show the Most Promise to Promote and Economic Review 88 (3): 537–58. Achieve Financial Well-Being?” Colloquium Levy-Yeyati, Eduardo, Alejandro Micco, and Ugo paper, National Endowment for Financial Edu- Panizza. 2007. “A Reappraisal of State-Owned cation, Denver, CO. Banks.” Economica 7 (2): 209–47. Lusardi, Annamaria, and Olivia Mitchell. 2007. Lewis, Logan, Bernardo Morais, and Claudia Ruiz. “Baby Boomer Retirement Security: The Roles 2013. “Banking Competition and Collusion.” of Planning, Financial Literacy, and Housing Unpublished working paper, World Bank, Wealth.” Journal of Monetary Economics 54 Washington, DC. (1): 205–24. Li, Hongyi, Lyn Squire, and Heng-Fu Zou. 1998. ———. 2008. “Planning and Financial Literacy: “Explaining International and Intertemporal How Do Women Fare?” American Economic Variations in Income Inequality.” Economic Review 98 (2): 413–17. Journal 108 (446): 26–43. ———. 2011a. “Financial Literacy around the Li, Hongyi, Lixin Colin Xu, and Heng-Fu Zou. World: An Overview.” Journal of Pension Eco- 2000. “Corruption, Income Distribution, nomics and Finance 10 (4): 497–508. and Growth.” Economics and Politics 12 (2): ———. 2011b. “Financial Literacy and Retire- 155–82. ment in the United States.” Journal of Pension Lim, Youngjae, and Robert M. Townsend. 1998. Economics and Finance 10 (4): 509–25. “General Equilibrium Models of Financial Lusardi, Annamaria, and Peter Tufano. 2009. Systems: Theory and Measurement in Village “Debt Literacy, Financial Experiences and Economies.” Review of Economic Dynamics 1 Overindebtedness.” NBER Working Paper (1): 59–118. 14808, National Bureau of Economic Research, Love, Inessa, and María Soledad Martínez Pería. Cambridge, MA. 2012. “How Bank Competition Affects Firms’ Lützenkirchen, Cédric, and Christian Weistroffer. Access to Finance.” Policy Research Working 2012. “Microfinance in Evolution: An Indus- Paper 6163, World Bank, Washington, DC. try between Crisis and Advancement.” Current Love, Inessa, María Soledad Martínez Pería, and Issues. Deutsche Bank Research, Frankfurt, Sandeep Singh. 2013. “Collateral Registries Germany. for Movable Assets: Does Their Introduction Lyman, Timothy, Mark Pickens, and David Por- Spur Firms’ Access to Bank Finance?” Policy teous. 2008. “Regulating Transformational Research Working Paper 6477, World Bank, Branchless Banking: Mobile Phones and Other Washington, DC. Technology to Increase Access to Finance.” Love, Inessa, and Nataliya Mylenko. 2003. “Credit Focus Note 43, Consultative Group to Assist Reporting and Financing Constraints.” Policy the Poor, World Bank, Washington, DC. Research Working Paper 3142, World Bank, Maddedu, Oscar. 2010. “The Status of Informa- Washington, DC. tion Sharing and Credit Reporting Infrastruc- Love, Inessa, Lorenzo Preve, and Virginia Sartia- ture in the Middle East and North Africa.” Allende. 2007. “Trade Credit and Bank Credit: World Bank, Washington, DC. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 197 Mader, Philip. 2013. “Rise and Fall of Micro- Group to Assist the Poor, World Bank, Wash- finance in India: The Andhra Pradesh Crisis in ington, DC. Perspective.” Strategic Change 22 (1–2): 47–66. McKenzie, David, and Christopher Woodruff. Maimbo, Samuel Munzele, ed. 2006. “Remit- 2008. “Experimental Evidence on Returns to tances and Economic Development in Somalia: Capital and Access to Finance in Mexico.” An Overview.” Social Development Paper 38, World Bank Economic Review 22 (3): 457–82. World Bank, Washington, DC. ———. Forthcoming. “What Are We Learning Mandell, Lewis. 2008. The Financial Literacy of from Business Training and Entrepreneurship Young American Adults: Results of the 2008 Evaluations around the Developing World?” National Jump$tart Coalition Survey of High World Bank Research Observer, World Bank, School Seniors and College Students. Washing- Washington, DC. ton, DC: Jump$tart Coalition for Personal Fi- McMillan, John, and Christopher Woodruff. 1999. nancial Literacy. “Interfirm Relationships and Informal Credit in Mandell, Lewis, and Linda Schmid Klein. 2007. Vietnam.” Quarterly Journal of Economics 114 “Motivation and Financial Literacy.” Financial (4): 1285–320. Services Review 16 (2): 105–16. Meador, Mark. 1982. “The Effects of Mortgage ———. 2011. “Financial Literacy around the Laws on the Home Mortgage Rates.” Journal World: An Overview.” Journal of Pension Eco- of Economics and Business 34 (2): 143–48. nomics and Finance 10 (4): 497–508. Melecký, Martin, and Anca Maria Podpiera. 2012. Marotta, Giuseppe. 2005. “Is Trade Credit More “Institutional Structures of Financial Sector Su- Expensive Than Bank Credit Loans? Evidence pervision, Their Drivers, and Emerging Bench- from Italian Firm-Level Data.” Applied Eco- mark Models.” World Bank, Washington, DC. nomics 37 (4): 403–16. Merton, Robert. 1992. “Financial Innovation and Martin, Matthew. 2007. “Literature Review on the Economic Performance.” Journal of Applied Effectiveness of Financial Education.” Working Corporate Finance 4 (4): 12–22. Paper 07–03, Federal Reserve Bank of Rich- Merton, Robert, and Zvi Bodie. 1995. “A Con- mond, Richmond, VA. ceptual Framework for Analyzing the Financial Mas, Ignacio. 2009. “The Economics of Branchless Environment.” In The Global Financial System: Banking.” Innovations, MIT Press 4 (2): 57–76. A Functional Perspective, edited by D. B. Crane, Mas, Ignacio, and Kabir Kumar. 2008. “Banking K. A. Froot, S. P. Mason, A. Perold, R. C. Mer- on Mobiles: Why, How, for Whom?” Focus ton, Z. Bodie, E. R. Sirri, and P. Tufano, 3–31. Note 48, Consultative Group to Assist the Poor, Boston, MA: Harvard Business School Press. World Bank, Washington, DC. ———. 2004. “The Design of Financial Systems: Massoud, Nadia, Anthony Saunders, and Barry Towards a Synthesis of Function and Structure.” Scholnick. 2007. “Who Makes Credit Card NBER Working Paper 10620, National Bureau Mistakes.” Unpublished working paper, Phila- of Economic Research, Cambridge, MA. delphia Federal Reserve, Philadelphia, PA. Mester, Loretta, Leonard Nakamura, and Mi- Maurer, Klaus, and Justyna Pytkowska. 2010. cheline Renault. 2007. “Transactions Accounts “Indebtedness of Microcredit Clients in Bosnia and Loan Monitoring.” Review of Financial and Herzegovina: Results from a Comprehen- Studies 20 (3): 529–56. sive Field Study.” Responsible Finance Studies, Miller, Calvin, and Linda M. Jones. 2010. Agricul- European Fund for Southeast Europe and Mi- tural Value Chain Finance: Tools and Lessons. crofinance Center, Warsaw, Poland. Rome: Food and Agriculture Organization of Mbiti, Isaac, and David Weil. 2011. “Mobile the United Nations; Warwickshire, U.K.: Practi- Banking: The Impact of M-Pesa in Kenya.” cal Action. NBER Working Paper 17129, National Bureau Miller, Margaret. 2003. Credit Reporting Systems of Economic Research, Cambridge, MA. and the International Economy. Cambridge, McKee, Kate, Estelle Lahaye, and Antonique Kon- MA: MIT Press. ing. 2011. “Responsible Finance: Putting Prin- Miller, Margaret, Julia Reichelstein, Christian ciples to Work.” Focus Note 73, Consultative Salas, and Bilal Zia. 2013. “Can You Teach 198 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Someone to Be Financially Capable? A Meta- NEDA (National Economic and Development Au- analysis of the Literature.” Background paper, thority of the Philippines). 2011. “Philippine World Bank, Washington, DC. Development Plan 2011–2016.” NEDA, Pasig Mischel, Walter, Yuichi Shoda, and Monica Ro- City, the Philippines. http://devplan.neda.gov driguez. 1989. “Delay of Gratification in Chil- .ph/. dren.” Science May (26): 933–38. Nenova, Tititana, Cecile Thioro Nang, and Anjum Mobarak, Ahmed, and Mark Rosenzweig. 2012. Ahmad. 2009. Bringing Finance to Pakistan’s “Selling Formal Insurance to the Informally In- Poor: Access to Finance for Small Enterprises sured.” Economics Department Working Paper and the Underserved. Washington, DC: World 97, Yale University, New Haven, CT. Bank. Mohieldin, Mahmoud, Zamir Iqbal, Ahmed Ros- OECD (Organisation for Economic Co-operation tom, and Xiaochen Fu. 2011. “The Role of Is- and Development). 2005. Oslo Manual: Guide- lamic Finance in Enhancing Financial Inclusion lines for Collecting and Interpreting Innovation in Organization of Islamic Cooperation Coun- Data. Paris: OECD. tries.” Policy Research Working Paper 4435, ———. 2011a. Financing High-Growth Firms: World Bank, Washington, DC. The Role of Angel Investors. Paris: OECD. Morduch, Jonathan. 1999. “The Role of Subsidies ———. 2011b. G20 High Level Principles on Fi- in Microfinance: Evidence from the Grameen nancial Consumer Protection. OECD, Paris. Bank.” Journal of Development Economics 60 Ongena, Steven, and Ilkay Sendeniz-Yuncu. 2011. (1): 229–48. “Which Firms Engage Small, Foreign, or State Morduch, Jonathan, and Dean Karlan. 2009. “Ac- Banks? And Who Goes Islamic? Evidence from cess to Finance.” In Handbook of Develop- Turkey.” Journal of Banking and Finance 35 ment Economics, vol. 5, edited by Dani Rodrik (12): 3213–24. and Mark Rosenzweig, 4704–84. Amsterdam: Osmani, Lutfun. 2007. “A Breakthrough in Wom- North-Holland. en’s Bargaining Power: The Impact of Micro- Moscarini, Guiseppe, and Fabien Postel-Vinay. credit.” Journal of International Development Forthcoming. “The Contribution of Large and 19 (5): 695–716. Small Employers to Job Creation in Times of Padilla, Jorge, and Marco Pagano. 2000. “Sharing High and Low Unemployment.” American Default Information as a Borrower Discipline Economic Review. Device.” European Economic Review 44 (10): Mulkay, Benoît, and Jacques Mairesse. 2013. “The 1951–80. R&D Tax Credit in France: Assessment and Pagano, Marco, and Tullio Jappelli. 1993. “Infor- Ex-Ante Evaluation of the 2008 Reform.” mation Sharing in Credit Markets.” Journal of NBER Working Paper 19073, National Bureau Finance 48 (5): 1693–718. of Economic Research, Cambridge, MA. Pande, Rohini, Shawn Cole, Anitha Sivasankaran, Munshi, Kaivan, and Mark Rosenzweig. 2005. Gautam Gustav Bastian, and Katherine Dur- “Why Is Mobility in India So Low? Social In- lacher. 2012. Does Poor People’s Access to surance, Inequality, and Growth.” BREAD Formal Banking Services Raise their Incomes? Working Paper 97, Bureau for Research and London, U.K.: EPPI-Centre, Social Science Re- Economic Analysis of Development, Durham, search Unit, Institute of Education, University NC. of London. Muravyev, Alexander, Dorothea Schäfer, and Olek- Paweenawat, Archawa, and Robert M. Townsend. sandr Talavera. 2009. “Entrepreneurs’ Gender 2012. “Village Economic Accounts: Real and and Financial Constraints: Evidence from Inter- Financial Intertwined.” American Economic national Data.” Journal of Comparative Eco- Review Papers and Proceedings 102 (3): nomics 37 (2): 270–86. 441–46. Nair, Praveen, and Avneesh Kaicker. 2009. “Sec- Phelan, Christopher, and Robert M. Townsend. ondary Trading Exchange: Feasibility Study for 1991. “Computing Multiperiod Information- India.” Working paper, SP Jain Center of Man- Constrained Optima.” Review of Economic agement, Singapore. Studies 58 (5): 853–81. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 199 Pickens, Mark, David Porteous, and Sarah Rot- ties and Challenges of the Market at the Bottom man. 2009. “Banking the Poor via G2P Pay- of the Pyramid. New York, NY: McGraw Hill. ments.” Focus Note 58, Consultative Group to Rocha, Roberto, Subika Farazi, Rania Khouri, Assist the Poor, World Bank, Washington, DC. and Douglas Pearce. 2010. “The Status of Pitt, Mark, and Shahidur Khandker. 1998. “The Bank Lending to SMEs in the Middle East and Impact of Group-Based Credit on Poor House- North Africa Region: The Results of a Joint holds in Bangladesh: Does the Gender of Partic- Survey of the Union of Arab Banks and the ipants Matter?” Journal of Political Economy World Bank.” Working paper, World Bank, 106 (5): 958–96. Washington, DC. Porteous, David. 2005. Making Financial Markets Roodman, David. 2011. Due Diligence: An Im- Work for the Poor. Johannesburg, South Af- pertinent Inquiry into Microfinance. Center for rica: FinMark Trust. Global Development, Washington, DC. Prescott, Edward, and Robert M. Townsend. ———. 2012. “Latest Impact Research: Inching 1984a. “General Competitive Analysis in an Toward Generalization.” Consultative Group Economy with Private Information.” Interna- to Assist the Poor, World Bank, Washington, tional Economic Review 25 (1): 1–20. DC. http://www.cgap.org/blog/latest-impact ———. 1984b. “Pareto Optima and Competitive -research-inching-towards-generalization?utm Equilibria with Adverse Selection and Moral _source=feedburner&utm_medium=feed Hazard.” Econometrica 52 (1): 21–45. &utm_campaign=Feed%253A%20cgap%252 Qian, Jun, and Philip Strahan. 2007. “How Laws FUaRp%20%2528Prod%20-%20CGAP%20 and Institutions Shape Financial Contracts: The Microfinance%20Blog%2529. Case of Bank Loans.” The Journal of Finance Roodman, David, and Jonathan Morduch. 2011. 62 (6): 2803–34. “The Impact of Microcredit on the Poor in Quercia, Roberto, and Jonathan Spader. 2008. Bangladesh: Revisiting the Evidence.” Work- “Does Homeownership Counseling Affect the ing Paper 174, Center for Global Development, Prepayment and Default Behavior of Afford- Washington, DC. able Mortgage Borrowers?” Journal of Policy Rosenzweig, Mark, and Kenneth Wolpin. 1993. Analysis and Management 27 (2): 304–25. “Credit Market Constraints, Consumption Rajan, Raghuram G. 2010. Fault Lines: How Hid- Smoothing, and Accumulation of Durable den Fractures Still Threaten the World Econ- Production Assets in Low-Income Countries: omy. Princeton, NJ: Princeton University Press. Investments in Bullocks in India.” Journal of Rajan, Raghuram G., and Luigi Zingales. 1998. Political Economy 101 (2): 223–44. “Financial Dependence and Growth.” Ameri- Rotman, Sarah, David Ferrand, and Stephen Ras- can Economic Review 88 (3): 559–86. mussen. 2012. “The Jipange KuSave Experi- Rajan, Uday, Amit Seru, and Vikrant Vig. Forth- ment in Kenya.” Brief (October), Consultative coming. “The Failure of Models That Predict Group to Assist the Poor, World Bank, Wash- Failure: Distance, Incentives, and Defaults.” ington, DC. Journal of Financial Economics. Rotman, Sarah, Kabir Kumar, and Marcia Parada. Rashid, Shahidur, Alex Winter-Nelson, and Philip 2013. “An Overview of the G2P Payments Garcia. 2010. “Purpose and Potential for Com- Sector in Pakistan.” Consultative Group to As- modity Exchanges in African Economies.” sist the Poor, World Bank, Washington, DC. IFPRI Discussion Paper 01035, International Ruiz, Claudia. 2013. “From Pawn Shops to Banks: Food Policy Research Institute, Washington, The Impact of Formal Credit on Informal DC. Households.” Policy Research Working Paper Ravallion, Martin, and Gaurav Datt. 1996. “How 6634, World Bank, Washington, DC. Important to India’s Poor Is the Sectoral Com- Saadani, Youssef, Zsofia Arvai, and Roberto position of Economic Growth?” World Bank Rocha. 2011. “A Review of Credit Guarantee Economic Review 10 (1): 1–25. Schemes in the Middle East and North Africa Rhyne, Elizabeth. 2009. Microfinance for Bankers Region.” Policy Research Working Paper 5612, and Investors: Understanding the Opportuni- World Bank, Washington, DC. 200 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 Safavian, Mehnaz, and Aban Haq. 2013. Are Pak- Shapiro, Jeremy. 2012. “Weather Insurance and In- istan’s Women Entrepreneurs Being Served by vestment Choice.” Unpublished working paper, the Microfinance Sector? Directions in Develop- Massachusetts Institute of Technology, Cam- ment: Finance series, World Bank, Washington, bridge, MA. DC. Shui, Haiyan, and Lawrence Ausubel. 2004. “Time Safavian, Mehnaz, and Joshua Wimpey. 2007. Inconsistency in the Credit Card Market.” Dis- “When Do Enterprises Prefer Informal Credit?” sertation, University of Maryland, College Policy Research Working Paper 4435, World Park, MD. Bank, Washington, DC. Smith, Brent. 2011. “Stability in Consumer Credit Saldaña, Cesar. 2000. “Assessing the Economic Scores: Level and Direction of FICO Score Drift Value of Credit Guarantees.” Journal of Philip- as a Precursor to Mortgage Default and Pre- pine Development 27 (1): 27–68. Payment.” Journal of Housing Economics 20 Samila, Sampsa, and Olav Sorenson. 2011. “Ven- (4): 285–98. ture Capital, Entrepreneurship, and Economic Spader, Jonathan, Janneke Ratcliffe, Jorge Mon- Growth.” Review of Economics and Statistics toya, and Peter Skillern. 2009. “The Bold and 93 (1): 338–49. the Bankable: How the Nuestro Barrio Tele- Schicks, Jessica, and Richar Rosenberg. 2011. “Too novela Reaches Latino Immigrants with Finan- Much Microcredit? A Survey of the Evidence cial Education.” Journal of Consumer Affairs on Over-Indebtedness.” Occasional Paper 19, 43 (1): 56–79. Consultative Group to Assist the Poor, World Stango, Victor, and Jonathan Zinman. 2009. Bank, Washington, DC. “Exponential Growth Bias and Household Fi- Schreiner, Mark. 2003. “Scoring: The Next Break- nance.” Journal of Finance 64 (6): 2807–49. through in Microcredit?” http://www.cgap Stein, Jeremy. 2002. “Information Production .org/sites/default/files/CGAP-Occasional-Paper and Capital Allocation: Decentralized vs. Hi- -Scoring-The-Next-Breakthrough-in-Micro erarchical Firms.” Journal of Finance 57 (5): credit-Jan-2003.pdf. 1891–921. ———. 2004. “Benefits and Pitfalls of Statistical Stein, Peer, Tony Goland, and Robert Schiff. 2010. Credit Scoring for Microfinance.” Savings and “Two Trillion and Counting: Assessing the Development 28 (1): 63–86. Credit Gap for Micro, Small and Medium- Schultz, Theodore. 1980. “Nobel Lecture: The Size Enterprises in the Developing World.” Economics of Being Poor.” Journal of Political McKinsey & Company and International Finance Economy 88 (4): 639–51. Corporation, World Bank, Washington, DC. Schumpeter, Joseph. 1934. The Theory of Eco- Stein, Peer, Bikki Randhawa, and Nina Bilandzic. nomic Development. Cambridge, MA: Harvard 2011. “Toward Universal Access: Address- University Press. ing the Global Challenge of Financial Inclu- Sekita, Shizuka. 2011. “Financial Literacy and Re- sion.” Chapter 10 in Postcrisis Growth and tirement Planning in Japan.” Journal of Pension Development: A Development Agenda for the Economics and Finance 10 (4): 637–56. G-20, edited by Shahrokh Fardoust, Yongbeom Seshan, Ganesh, and Dean Yang. 2012. “Transna- Kim, and Claudia Sepúlveda. Washington, DC: tional Household Finance: A Field Experiment World Bank. on the Cross-Border Impacts of Financial Edu- Stephanou, Constantinos, and Camila Rodriguez. cation for Migrant Workers.” Draft working 2008. “Bank Financing to Small and Medium- paper, Innovations for Poverty Action, New Sized Enterprises in Colombia.” Policy Re- Haven, CT. search Working Paper 4481, World Bank, Shane, Scott A. 2005. “Angel Investing: A Re- Washington, DC. port Prepared for the Federal Reserve Banks of Stiglitz, Joseph. 1990. “Peer Monitoring and Atlanta, Cleveland, Kansas City, Philadelphia, Credit Markets.” World Bank Economic Re- and Richmond.” October 1, Department of view 4 (3): 351–66. Economics, Case Western Reserve University, Stiglitz, Joseph, and Andrew Weiss. 1981. “Credit Cleveland, OH. Rationing in Markets with Imperfect Infor- GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 201 mation.” American Economic Review 71 (3): finance: Is it Worth It?” International Journal of 393–410. Finance and Economics 17 (2): 103–23. Straub, Stéphane. 2005. “Informal Sector: The Van Order, Robert, Simon Firestone, and Peter Credit Market Channel.” Journal of Develop- Zorn. 2007. “The Performance of Low Income ment Economics 78 (2): 299–321. and Minority Mortgages.” Ross School of Busi- Suryahadi, Asep, Daniel Suryadarma, and Sudarno ness Paper 1083, University of Michigan, Ann Sumarto. 2009. “The Effects of Location and Arbor, MI. Sectoral Components of Economic Growth on Van Rooij, Maarten, Annamaria Lusardi, and Rob Poverty: Evidence from Indonesia.” Journal of Alessie. 2011. “Financial Literacy and Stock Development Economics 89 (1): 109–17. Market Participation.” Journal of Financial Tazhibayeva, Kamilya, and Robert M. Townsend. Economics 101 (2): 449–72. 2012. “The Impact of Climate Change on Rice ———. 2012. “Financial Literacy, Retirement Yields: Heterogeneity and Uncertainty.” Un- Planning, and Household Wealth.” Economic published paper, Massachusetts Institute of Journal 122 (560): 449–78. Technology, Cambridge, MA. Vercammen, James. 1995. “Credit Bureau Policy Tax, Sol. 1953. Penny Capitalism: a Guatemalan and Sustainable Reputation Effects in Credit Indian Economy. Washington, DC: U.S. Gov- Markets.” Economica 62 (248): 461–78. ernment Printing Office. Visaria, Sujata. 2009. “Legal Reform and Loan Townsend, Robert M. 1982 “Optimal Multiperiod Repayment: The Microeconomic Impact of Debt Recovery Tribunals in India.” American Contracts and the Gain from Enduring Rela- Economic Journal: Applied Economics 1 (3): tionships under Private Information.” Journal 59–81. of Political Economy 90 (6): 1166–86. Ward, Damian, and Ralf Zurbruegg. 2000. “Does ———. 1994. “Risk and Insurance in Village Insurance Promote Economic Growth? Evi- India.” Econometrica 62 (3): 539–91. dence from OECD Countries.” Journal of Risk ———. Forthcoming. “Accounting for the Poor: and Insurance 67 (4): 489–506. The T.W. Schultz Lecture.” American Journal Warnock, Veronica, and Frank Warnock. 2008. of Agricultural Economics. “Markets and Housing Finance.” Journal of Townsend, Robert M., and Jacob Yaron. 2001. Housing Economics 17 (3): 239–51. “The Credit Risk Contingency System of an Wenner, Mark, Navajas Sergio, Carolina Trivelli, Asian Development Bank.” Economic Perspec- and Alvaro Tarazona. 2007. “Managing Credit tives 25 (3): 31–48. Risk in Rural Financial Institutions in Latin Tufano, Peter, Timothy Flacke, and Nicholas May- America.” Sustainable Development Depart- nard. 2010. “Better Financial Decision Making ment Best Practices Series, Inter-American among Low Income and Minority Groups.” Development Bank, Washington, DC. Working paper, Rand Corporation, Santa Mon- WFE (World Federation of Exchanges). 2011. “Al- ica, CA. ternative Financing for SMEs.” FOCUS 226, Udry, Christopher, and Santosh Anagol. 2006. WFE, Paris, France. “The Return to Capital in Ghana.” American Williamson, Steven. 1987. “Costly Monitoring, Economic Review 96 (2): 388–93. Loan Contracts, and Equilibrium Credit Ra- UNCITRAL (United Nations Commission on tioning.” Quarterly Journal of Economics 102 International Trade Law). 2010. Legislative (1): 135–46. Guide on Secured Transactions. New York: Willis, Lauren. 2006. “Decision-Making and the United Nations. Limits of Disclosure: The Problem of Predatory Urquizo, Jacqueline. 2012. “The Financial Behav- Lending: Price.” Maryland Law Review 65: ior of Rural Residents: Findings from Five Latin 707–838. American Countries.” Acción report, Washing- Woodward, Susan. 2003. “Consumer Confusion ton, DC. in the Mortgage Market.” Research Paper, Sand Van Gool, Joris, Wouter Verbeke, Piet Sercu, and Hill Econometrics, Palo Alto, CA. http://www Bart Baesens. 2012. “Credit Scoring for Micro- .sandhillecon.com/pdf/consumer_confusion.pdf. 202 BIBLIOGRAPHY GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 World Bank. 2006a. “The Rural Investment Cli- ———. 2012c. World Development Report 2013: mate: It Differs and It Matters.” Report 36543- Jobs. Washington, DC: World Bank. GLB, World Bank, Washington, DC. ———. 2012d. Information and Communications ———. 2006b. “Expanding Post-Harvest Finance for Development: 2012 Maximizing Mobile. through Warehouse Receipts and Related In- Washington, DC: World Bank. struments.” Issue 8, Agricultural and Rural ———. 2012e. “Innovation in Retail Payments Development Notes, World Bank, Washington, Worldwide: A Snapshot.” Financial Infrastruc- DC. ture Series, World Bank, Washington, DC. ———. 2007a. World Development Report 2008: ———. 2012f. “Measuring Financial Capability Agriculture for Development. Washington, DC: and the Effectiveness of Financial Education.” World Bank. Financial Literacy and Education, Russia Trust ———. 2007b. “Bank Financing to Small and Me- Fund, World Bank, Washington, DC. dium Enterprises: Survey Results from Argen- ———. 2012g. Good Practices for Consumer Fi- tina and Chile.” Working paper, World Bank, nancial Protection. Washington, DC: World Washington, DC. Bank. ———. 2007c. “Bank Lending to Small and ———. 2012h. “Developing a Comprehensive Na- Medium Enterprises: The Republic of Serbia.” tional Retail Payments Strategy.” Financial In- Unpublished working paper, World Bank, frastructure Series: Payment Systems Policy and Washington, DC. Research, World Bank, Washington, DC. ———. 2009a. “Russian Federation: Key Findings ———. 2012i. “Financial Inclusion Strategies and Recommendations.” Vol. 1 of Diagnostic Reference Framework.” Prepared for the G20 Review of Consumer Protection in Financial Mexico Presidency, World Bank, Washington, Services. Private and Financial Sector Develop- DC. ment Department, Europe and Central Asia Re- ———. 2012j. “From Remittances to M-Payments: gion, World Bank, Washington, DC. Understanding ‘Alternative’ Means of Payment ———. 2009b. “Azerbaijan: Technical Note on within the Common Framework of Retail Pay- Consumer Protection and Financial Literacy.” ments System Regulation.” Financial Infra- Private and Financial Sector Development De- structure Series, World Bank, Washington, DC. partment, Europe and Central Asia Region, ———. 2012k. “General Guidelines for the Devel- World Bank, Washington, DC. opment of Government Payment Programs.” ———. 2010. “Latvia: Key Findings and Recom- Financial Infrastructure Series, World Bank, mendations.” Vol. 1 of Diagnostic Review of Washington, DC. Consumer Protection and Financial Capabil- ———. 2012l. “A Practical Guide for Retail Pay- ity. Private and Financial Sector Development ments Stocktaking.” Financial Infrastructure Department, Europe and Central Asia Region, Series, World Bank, Washington, DC. World Bank, Washington, DC. ———. 2013a. “Making Sense of Financial Capa- ———. 2011a. “General Principles for Credit Re- bility Surveys around the World: A Review of porting.” World Bank, Washington, DC. Existing Financial Capability and Literacy Mea- ———. 2011b. Migration and Remittances Fact- surement Instruments.” Working paper, World book 2011. Washington, DC: World Bank. Bank, Washington, DC. ———. 2011c. “Payment Systems Worldwide: A ———. 2013b. Growing Africa: Unlocking the Snapshot—Outcomes of the Global Payment Potential of Agribusiness. Washington, DC: Systems Survey 2010.” Financial Infrastructure World Bank. Series, World Bank, Washington, DC. ———. 2013c. “An Analysis of Trends in the Av- ———. 2012a. Global Financial Development Re- erage Total Cost of Migrant Remittance Ser- port 2013: Rethinking the Role of the State in vices.” Remittance Prices Worldwide Policy Finance. Washington, DC: World Bank. Note 7 (September), World Bank, Washington, ———. 2012b. “Good Practices for Financial DC. Consumer Protection.” Washington, DC: ———. 2013d. “Migration and Development World Bank. Brief.” World Bank, Washington, DC. GLOBAL FINANCIAL DEVELOPMENT REPORT 2014 BIBLIOGRAPHY 203 Wurgler, Jeffrey. 2000. “Financial Markets and for Rural Finance.” Lead theme paper, U.S. the Allocation of Capital.” Journal of Financial Agency for International Development, Wash- Economics 58 (1–2): 187–214. ington, DC. Xu, Lisa, and Bilal Zia. 2012. “Financial Lit- Zhan, Min, and Michael Sherraden. 2011. “Assets eracy around the World: An Overview of the and Liabilities, Education Expectations, and Evidence with Practical Suggestions for the Children’s College Degree Attainment.” Chil- Way Forward.” Policy Research Working Paper dren and Youth Services Review 33 (6): 846–54. 6107, World Bank, Washington, DC. Zia, Bilal H. 2008. “Export Incentives, Financial ———. Forthcoming. “Financial Literacy Pro- Constraints, and the (Mis)allocation of Credit: grams in the Developing World: An Overview of Micro-level Evidence from Subsidized Export Evidence with Practical Suggestions for the Way Forward.” World Bank Research Observer. Loans.” Journal of Financial Economics 87 (2): Zavatta, Roberto. 2008. “Financing Technology 498–527. Entrepreneurs and SMEs in Developing Coun- Zollman, Julie, and Daryl Collins. 2010. “Finan- tries.” infoDev, World Bank, Washington, DC. cial Capability and the Poor: Are We Missing Zeller, Manfred. 2003. “Models of Rural Fi- the Mark?” Insights paper, Financial Sector nancial Institutions. Paving the Way Forward Deepening Kenya, Nairobi, Kenya. ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. • 11 trees Global Financial Development Report • 5 million British 2014 was printed on recycled paper with 30 thermal units of total percent postconsumer fiber in accordance energy with the recommended standards for paper • 956 pounds of net usage set by the Green Press Initiative, a greenhouse gases nonprofit program supporting publishers in • 5,188 gallons of using fiber that is not sourced from endan- waste water gered forests. For more information, visit • 347 pounds of http://www.greenpressinitiative.org. solid waste