Financial Inclusion and Capability Survey Report Financial Inclusion Practice Enhancing Financial Capability and Inclusion in Mozambique A Demand-Side Assessment August 2014 © 2013 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. 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, denominations, 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. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. i Acknowledgements This Financial Inclusion and Financial Capability Survey Report was prepared by a team led by Siegfried Zottel1 (Financial Sector Specialist) from the World Bank’s Financial Inclusion Practice, with contributions from Claudia Ruiz Ortega (Economist) and Sarah Yan Xu (Research Analyst). The team is grateful to the peer reviewers of this report - Samuel Munzele Maimbo (Lead Financial Sector Specialist), Johanna Jaeger (Financial Sector Specialist), and Aidan Coville (Economist) – for their valuable comments. Douglas Pearce (Manager, Financial Inclusion Practice), and Mazen Bouri (Senior Private Sector Development Specialist) provided overall guidance. In addition, survey preparation support provided by Tania Saranga (Survey Consultant) and design inputs provided by Sarah Fathallah (Analyst) are gratefully acknowledged. The team expresses its deepest appreciation to the Mozambican authorities, including the Banco de Mozambique (BdM) and the National Statistical Office (INE) for their cooperation and collaboration which made this project possible. The survey was carried out at the request of the BdM and was implemented in close collaboration with the BdM. In particular, the team wants to extend its sincere gratitude to the following officials and experts from BdM who provided invaluable support and strategic guidance throughout the project: Ms. Dra. Esselina Macome (Member of the Board and Head of the Directorate of Issuance, Payments and Accounts), Mr. Dias Macuacua (Director, Behavioral Supervision Department), Ms. Aurora Bila (Director, Payment Systems Department), Mr. Rafael Francisco (Assistant Director, Payment Systems Department), Mr. Emilio Rungo (Assistant Director, Behavioral Supervision Department), Mr. Jose Alfredo Lobato Zacarias (Specialist, Behavioral Supervision Department), Ms. Carla Fernandes (Technician, Payment Systems Department), and Ms. Bordina Muala (Technician, International Relations Department). The team’s sincere appreciation is further extended to the following INE officials and experts for their collaboration and technical support: Mr. Arao Balate (Director, Population Census Department) and Carlos Greva (Sampling specialist, Population Census Department). The team would also like to express its gratitude to Étude Économique Conseil (EEC Canada), a Montreal based survey firm, which was selected to undertake this survey. We are grateful to Fares Khoury, the president of ECC, as well as all supervisors and enumerators for their efforts and commitments to successfully complete this survey. Finally, the team owes particular appreciation to all Mozambican women and men who patiently responded to the survey. The Survey Report was prepared as part of the Swiss State Secretariat for Economic Affairs (SECO) Trust Fund on “Consumer Protection and Financial Literacy” and received complementary funding from the World Bank Africa Region Vice Presidency. 1 The corresponding lead author can be contacted at: szottel@worldbank.org ii Contents Preface .......................................................................................................................................................... 1 Key Findings .................................................................................................................................................. 2 Summary of Key Recommendations ............................................................................................................. 3 Executive Summary ....................................................................................................................................... 4 Financial Inclusion ..................................................................................................................................... 4 Recommendations ................................................................................................................................. 4 Financial Capability.................................................................................................................................... 6 Recommendations ................................................................................................................................. 6 Relationship between Financial Inclusion and Capability ......................................................................... 9 Recommendations ................................................................................................................................. 9 Financial Consumer Protection ................................................................................................................ 11 Recommendations ............................................................................................................................... 11 Background on the Mozambique Survey ..................................................................................................... 13 1. Financial Inclusion ................................................................................................................................... 15 1.1 Context ........................................................................................................................................ 15 1.2 Usage of Banks ........................................................................................................................... 17 1.3 Usage of Bank Products .............................................................................................................. 20 1.4 Usage of Nonbank Financial Institutions ...................................................................................... 23 1.5 Usage of Products of Nonbank Financial Institutions ................................................................... 25 1.6 Barriers to Formal Account Ownership ........................................................................................ 26 2. Financial Capability.................................................................................................................................. 28 2.1 Knowledge of Financial Concepts and Products................................................................................ 28 2.1.1 Knowledge of Financial Concepts ............................................................................................... 28 2.1.2 Knowledge of Financial Products ................................................................................................ 35 2.2 Financial Behavior and Attitudes ....................................................................................................... 37 3. Relationship between Financial Inclusion and Financial Capability ......................................................... 43 4. Financial Consumer Protection ................................................................................................................ 48 References .................................................................................................................................................. 53 Appendix...................................................................................................................................................... 55 iii A. Background on the Mozambique Survey ......................................................................................... 55 B. Financial Inclusion ........................................................................................................................... 57 C. Regression Tables ........................................................................................................................... 59 1. Financial Inclusion ....................................................................................................................... 59 2. Financial Capability...................................................................................................................... 69 iv Figures Figure 1: Knowledge and usage of commercial banks by location of respondent........................................ 17 Figure 2: Knowledge and usage of commercial banks by income quartile and variability of income ........... 18 Figure 3: Knowledge and usage of commercial banks by degree of media consumption ........................... 19 Figure 4: Media consumption by different sociodemographic groups .......................................................... 19 Figure 5: Bank account penetration in different regions in Mozambique ..................................................... 21 Figure 6: Percentage of Mozambicans currently holding a financial product from a bank by urban status . 21 Figure 7: Percentage of Mozambicans with a bank account by income quintiles ....................................... 22 Figure 8: Percentage of Mozambicans with bank credit by income quintiles .............................................. 22 Figure 9: Percentage of adults in urban areas with a mortgage by income quartiles .................................. 23 Figure 10: Percentage of Mozambicans that have ever used financial institutions ..................................... 24 Figure 11: Percentage of Mozambicans by the number of financial institutions that they have used ......... 24 Figure 12: Usage of financial products in rural and urban areas................................................................. 26 Figure 13: Percentage of Mozambicans with no formal accounts reporting they do not need this product . 26 Figure 14: Reasons for not having a formal account in rural and urban areas ........................................... 27 Figure 15: Distribution of financial literacy scores ....................................................................................... 31 Figure 16: Financial literacy quiz overview ................................................................................................. 32 Figure 17: Education levels of populations with low and high financial literacy scores ............................... 32 Figure 18: Reported awareness & understanding of financial terms ........................................................... 34 Figure 19: Comparison of reported understanding and financial literacy quiz results ................................. 34 Figure 20: Distribution of financial products awareness scores .................................................................. 35 Figure 21: Knowledge of financial products offered by different providers .................................................. 36 Figure 22: Percentage of Mozambicans that know about different providers by number of media used .... 37 Figure 23: Average financial capability scores ............................................................................................ 38 Figure 24: Average budgeting score by education levels in urban and rural areas ..................................... 41 Figure 25: Financial Capability in Choosing Financial Products (Left) and Being Far-sighted (Right) by Region ......................................................................................................................................................... 41 Figure 26: Average choosing financial products score by media consumed in urban and rural areas ........ 42 Figure 27: Financial products awareness score of Mozambicans with and without formal accounts .......... 44 Figure 28: Usage of financial products by awareness of financial products score ...................................... 44 Figure 29: Financial literacy scores of Mozambicans with and without formal accounts ............................. 45 Figure 30: Usage of financial products by financial literacy score............................................................... 45 Figure 31: Financial behaviors & attitudes of Mozambicans with and without formal accounts .................. 46 Figure 32: Financial behaviors & attitudes of Mozambicans with and without different financial products .. 47 Figure 33: Usage and satisfaction rates for different financial providers ..................................................... 49 Figure 34: Commercial bank satisfaction rates by financial literacy score .................................................. 50 Figure 35: Approaches to deal with financial service provider conflicts ...................................................... 50 Figure 36: Actions taken to redress conflicts with financial service providers ............................................. 51 Figure 37: Reasons for not solving conflicts with financial service providers .............................................. 52 v Figure 38: Estimated population break-down by urban/rural ...................................................................... 55 Figure 39: Estimated population break-down by different income groups .................................................. 55 Figure 40: Estimated Population Break-down by Male/Female .................................................................. 55 Figure 41: Estimated population break-down by age groups ...................................................................... 56 Figure 42: Estimated population break-down by education groups ............................................................ 56 Figure 43: Estimated division of stable/unstable income groups ................................................................ 56 Figure 44: Estimated population break-down by household size ................................................................. 56 Figure 45: Account at a formal financial institution across Sub-Saharan African countries ........................ 57 Figure 46: Loan from a financial institution in the last year across Sub-Saharan African countries ........... 58 Tables Table 1: International comparison of knowledge of basic financial concepts (in % of adults) ..................... 29 Table 2: Cross-country comparison of different financial capability scores ................................................. 39 Table 3: Probability of knowing about commercial banks on demographic and socioeconomic factors ..... 59 Table 4: Probability of having ever used commercial banks on demographic and socioeconomic factors . 60 Table 5: Probability of having ever used commercial bank services on village factors ............................... 61 Table 6: Probability of currently having a bank account on demographic and socioeconomic factors ........ 62 Table 7: Probability of currently having a bank loan on demographic and socioeconomic factors .............. 63 Table 8: Probability of having ever used insurance services on demographic and socioeconomic factors . 64 Table 9: Probability of having ever used MFI services on demographic and socioeconomic factors ......... 65 Table 10: Probability of having ever used money changers on demographic and socioeconomic factors .. 66 Table 11: Probability of having ever used money lenders on demographic and socioeconomic factors ..... 67 Table 12: Probability of having ever had a formal account on demographic and socioeconomic factors ... 68 Table 13: Probability of financial literacy and financial product knowledge scores on village factors ......... 69 Table 14: Probability of financial literacy score on demographic and socioeconomic factors ..................... 70 Table 15: Probability of financial knowledge score on demographic and socioeconomic factors ............... 71 Table 16: Capability of covering unexpected expenses on demographic and socioeconomic factors ........ 72 Table 17: Satisfaction rate on commercial banks on demographic and socioeconomic factors.................. 73 Table 18: Probability of using financial instruments on demographic and socioeconomic factors .............. 74 Table 19: Probability of using financial instruments on financial capability scores ...................................... 75 Boxes Box 1: The WB Financial Capability Survey in the context of the wider financial sector strategy ................ 16 Box 2: Media Consumption Overview .......................................................................................................... 19 Box 3: Financial Literacy Quiz ..................................................................................................................... 30 vi Abbreviations and Acronyms AFI Alliance for Financial Inclusion ASCAs Accumulating Savings and Credit Associations BdM Banco de Mocambique (Bank of Mozambique) CAPI Computer-Assisted Personal Interview EA Enumeration Areas EEC Étude Économique Conseil MFIs Microfinance Organizations MFSDS Mozambique Financial Sector Strategy NBFIs Nonbank Financial Institutions PCA Principal Component Analysis PPS Probability Proportional to Size PSUs Primary Sampling Units RTF Russia Trust Fund for Financial Literacy and Education vii Glossary2 Branchless Banking The delivery of financial services outside conventional bank branches through the use of retail agents and information and communications technologies, such as mobile phones, to transmit transaction details. Community Savings Savings and credit self-help groups such as ASCAs, OPEs, Groups Xitiques, and Conta Familias. Financial Capability The capacity to act in one’s best financial interest, given socioeconomic and environmental conditions. It encompasses knowledge (literacy), attitudes, skills and behaviors of consumers with respect to understanding, selecting, and using financial services, and the ability to access financial services that fit their needs. Financial Inclusion Financial Inclusion is defined as proportion of individuals who use financial services. Financial Institution Any public or private institution whose main function is the provision of financial services for its customers or members. Probably the most important financial service provided by financial institutions is acting as financial intermediaries. Financial Sector The totality of financial institutions that operate in Mozambique. This includes credit institutions and financial companies, as well as microfinance operators, which are under the supervision of the Mozambican central Bank (BdM); insurance companies, which are under the supervision of the Insurance Supervision institute / the Ministry of Finance; the stock exchange operators, which are under joint supervision of the BdM and the Mozambican Stock exchange; and the pension funds. Financial System In this report, the definition of financial system is equivalent to the financial sector. Formal Financial Financial institutions that are licensed by and prudentially Institutions supervised by the banking authorities in Mozambique, e.g. banks and licensed non-bank financial institutions. Informal Financial Financial institutions that are not registered with or officially 2Please note that this glossary is not meant to provide a legal definition of the terms used in this report. Different government agencies and stakeholders may have specific definitions of the term for the respective purposes of statistical information, government programs, incentive schemes, etc. viii Institutions recognized by any government authority, e.g. unregistered money lenders. Key Facts Statement A summary statement which provides consumers with simple and standard disclosure of key contractual information of a baking product or service, contributing to the consumers’ better understanding of the product or service. Key Fact Statements also allow consumers to easily compare offers provided by different banks before they purchase a banking product or service. Microfinance Financial institutions that target poor and low-income persons as Institutions their main market niche. Micro-insurance Protection of low-income people against specific perils in exchange for regular monetary payments (premiums) proportionate to the likelihood and cost of the risk involved. Money Changers A money changer is a person who exchanges the coins or currency of one country for that of another. Money Lenders A money lender is an informal lender, either person or a group which offers small personal loans at rather high interest rates (‘agiotas’). This category also includes friends, relatives, and neighbors who offer loans which need to be repaid. Nonbank Financial A Financial Institution that provides financial services without Institution meeting the legal definition of a bank, i.e. it does not hold a banking license. Examples are microfinance institutions, insurers, etc. Teachable Moments Times in people's lives when they are more likely to be receptive to new information as they can relate it directly to their own life events. ix Preface Financial capability, as defined by the World Bank and in this report, is the capacity to act in one’s best financial interest, given socioeconomic and environmental conditions. It encompasses knowledge (literacy), attitudes, skills and behavior of consumers with respect to understanding, selecting, and using financial services, and the ability to access financial services that fit their needs (World Bank 2013d). In this report, financial inclusion is defined as the proportion of individuals that use financial services. Financial capability has become a policy priority for policy makers seeking to promote responsible financial inclusion and to ensure financial stability and functioning financial markets. Today people are required to take increasing responsibility for managing a variety of risks over the life cycle. People who make sound financial decisions and who effectively interact with financial service providers are more likely to achieve their financial goals, hedge again financial and economic risks, improve their household’s welfare, and support economic growth. Boosting financial capability has therefore emerged as a policy objective that complements governments’ financial inclusion and consumer protection agendas. To this end, policy makers are increasingly using surveys as diagnostic tools to identify financial capability areas that need improvement and vulnerable segments of the population which could be targeted with specific interventions. In response to a request of the Banco de Mocambique (BdM), the World Bank has conducted a financial capability survey. This is a priority follow up to the Mozambique Financial Sector Strategy (MFSDS) 2013-2022, given i) that financial literacy/capability has been identified by the BdM as a priority area going forward, ii) the low levels of financial inclusion and the importance of financial capability in enabling people to take up and benefit from financial products and services, and iii) the lack of comprehensive, robust, and reliable data which has prevented policy makers so far from formulating specific policy actions and setting quantifiable and concrete targets. The key findings and recommendations presented in this report cover 4 main areas: 1. Financial Inclusion, 2. Financial Capability, 3. Relationship between Financial Inclusion and Capability, and 4. Financial Consumer Protection. The remaining chapters are structured as follows. Chapter 1 explores the financial inclusion landscape in Mozambique. Chapter 2 gives an overview of Mozambicans’ levels of financial capability, in particular about their financial knowledge, attitudes and behaviors. The relationship between financial capability and inclusion is discussed in chapter 3. The last chapter investigates if the products which financially included individuals use are effectively meeting their needs. 1 Key Findings 2 Summary of Key Recommendations Recommendations Responsible Term3 Introduce policies that promote a more BdM MT competitive and diverse financial sector Financial Promote branchless banking BdM MT Inclusion Encourage banks to introduce no-frills savings BdM MT and payment accounts with nil or very low minimum balance Develop a comprehensive financial education BdM, Ministry of Finance, MT strategy or action plan, based on the results of Ministry of Education, industry this financial capability survey associations, consumer associations, and other stakeholders Use a wide range of programs, including mass BdM, Ministry of Finance, MT media, comic books, trusted intermediaries, etc., Ministry of Education, industry to enhance financial knowledge, and change associations, consumer attitudes and behavior associations, and other stakeholders Financial Combine financial capability enhancing BdM, Ministry of Finance, MT Capability interventions with other interventions, such as Ministry of Education, industry text message reminders, to increase its associations, consumer effectiveness associations, and other stakeholders Combine financial capability-enhancing programs BdM, industry associations, MT with available financial products, which most market participants people can access, to promote beneficial participations in the financial markets Share this survey results with financial BdM ST institutions to help them develop tailored products to the needs of underserved population Require Key Fact Statements for financial BdM ST products and test consumer understanding of disclosure material Require financial institutions to disclose in all pre- BdM ST Financial contractual and contractual disclosure formats Consumer detailed information on the internal as well as Protection relevant external dispute resolution mechanisms Analyze data on consumer complaints submitted BdM MT by financial institutions periodically and use this information as input to supervisory and regulatory activities 3 ST, short term, indicates action can be undertaken in 0-6 months. MT, medium term, indicates 6 months-1 year. LT, long term, indicates 1+ years 3 Executive Summary Financial Inclusion The financial system in Mozambique is heavily dominated by banks, but only 52 percent of adults have ever used their products. The problem of lack of access to basic financial services is far more pronounced in rural areas where 42 percent of the population has ever used bank products as compared to 73 percent in urban locations. Within urban and rural communities, the data suggests that the development level of the area matters. As compared to areas with lower development and infrastructure levels, people are more likely to use bank services in areas with shorter distances to bank branches and better infrastructure. Other financially excluded segments are people living on low and fluctuating incomes. The survey data further suggests that around a third of the population is not being reached by any financial service providers and that a substantial overlap exists on the type of clients targeted by banks and other providers, including MFIs. As with banks, clients of money changers and insurance companies are concentrated at the highest income segment. In contrast, MFI clients are not from the lowest income segment, and it is more likely for Mozambican respondents above the median income to have used MFIs than for those below it. Only money lender clients seem to be more likely to be less educated and from a less favorable background. Among the financially included segments, bank accounts are the most common products. On average, 46 percent of urban residents currently own a bank account, as compared to 19 percent of rural dwellers. Usage of money transfer services, credit, both from formal4 and informal providers, and insurance products is not very common. More sophisticated savings products such as investments in stocks or private pensions are hardly used at all. Important barriers to account ownership are lack of money, affordability and lack of financial knowledge of financial products and services. One out of five of those without an account who live in rural areas report that they cannot afford it. Although this number is lower for urban populations, it is still the second most important reason for not having an account. Findings from the survey also suggest that lack of trust and financial knowledge of financial products hinders 19 and 26 percent of urban and rural Mozambican respondents without accounts from using basic financial services. Recommendations In order to close the identified gap between urban and rural populations in accessing financial services, it is recommended to harness the potential of branchless banking. Mobile or agent banking can dramatically reduce the costs of delivering financial services outside larger urban centers, in particular 4 This number includes credit from banks and MFIs. 4 in low-density and remote areas with prohibitively high costs of establishing traditional branch networks. Policies facilitating the introduction of these lower-cost technologies, such as the development of a legal framework, can help reach remote locations and rural populations that were previously excluded from financial services. Furthermore, the introduction of basic bank accounts could become an entrance door to the formal financial system to underserved parts of the population identified through this survey. It is suggested to encourage banks to introduce no-frills savings and payment accounts with nil or very low minimum balance because they can enable low income segments to transfer money and to store it at a safe place (World Bank, 2013a). However, international experience in countries such as India, the Philippines, or South Africa shows that policies related to the introduction of such products need to be complemented with public awareness campaigns, otherwise the uptake of these products might be very low. Advancing financial inclusion levels in Mozambique will also require a more competitive and diverse financial sector to make products affordable to larger parts of the population. In Mozambique, not only the financial sector is heavily concentrated in banks, but also, within the banking sector the three largest banks account for 85 percent of the sector’s assets (Mozambique Council of Ministers, 2013), suggesting low competition in the sector. In line with Mozambique’s Financial Sector Development Strategy (MFSDS) and with the findings of recent research (e.g. Love and Martinez Peria, 2012), introducing policies that promote competition could encourage lower prices and make products affordable for broader segments of the population. The substantial overlap of clients targeted by banks and other providers further indicate the need to support the development, formalization and expansion of Nonbank financial institutions providing microfinance and micro-insurance services to lower income segments and rural populations. 5 Financial Capability Survey results highlight that financial knowledge and awareness levels of basic financial concepts and products are a significant challenge in Mozambique, as well as in many countries across different income levels. Mozambican respondents demonstrate relatively high comfort levels in solving simple numeracy tasks, compared to respondents from economies with different income levels. However, only 28 percent of Mozambican respondents have good understanding of compound interest and inflation, which appears to be low from a cross-country perspective. Likewise, awareness of financial products other than those provided by banks, MFIs, and money lenders appears to be limited. Respondents who are the least familiar with financial products tend to live in rural areas and on low and irregular income streams. The need to manage low and uncertain income streams is a strong predictor of low awareness of financial products, in particular with insurance products which would allow them to deal with bad events when they occur. Policies targeting Mozambicans with fluctuating income may need to be of first order since 72 percent of adults in Mozambique report having a volatile income. An international comparison shows that Mozambican respondents are especially competent in managing day-to-day finances, but are among the most challenged in terms of putting money aside for future expenses and choosing appropriate financial products. While Mozambicans demonstrate strengths in budgeting and monitoring their expenses, compared to respondents from 9 other countries, they display relatively weaker performance in saving, putting money aside for unexpected and old age expenses, and in particular in choosing appropriate financial products. These results are of concern given their implications for people’s ability to smooth consumption, to cope with economic shocks, to generate lump sums for productive investments, to take advantage of available financial products, and ultimately for their long term wellbeing. Despite being especially capable in a number of financial capability areas, rural dwellers and people living on low and fluctuating incomes struggle in particular with setting aside funds for unexpected expenses. As compared to higher income segments, the ability of low income populations to cope with unexpected shocks seems to be limited by their scarce resources. Similarly, people living with fluctuating incomes and rural residents have more difficulties with setting aside funds for unexpected expenses than their respective counterpart groups. As compared to urban populations and people with stable incomes, rural dwellers and those with varying incomes also struggle more with budgeting, choosing financial products, and they tend to think less about the future. Consequently, daily hardship and the constant struggle with solving immediate problems seem to draw their attention away from their longer-term needs. Recommendations It is suggested that a comprehensive financial education strategy or action plan be developed based on the results of the financial capability survey. The survey identified numerous financial 6 capability issues across various segments of the population. Further, the report suggests a number of policy actions which could be undertaken to improve financial capabilities. In order to ensure that scarce resources are used in the most efficient way, prioritization of certain financial capability enhancing programs is essential. The development of a financial education strategy or action plan could help to identify key priorities. Such priorities could be based on a number of criteria, including i) the need, ii) desired and expected impacts, iii) costs, iv) opportunities to scale up and v) leverage on existing programs. Both, the development of a strategy or action plan and the setting of priorities would require a wide consultation process which includes various stakeholders from public, private and non-profit entities. This could help to facilitate a wider consensus building about the importance of this topic and to achieve better coordination of all stakeholders and available resources towards boosted financial capabilities in Mozambique. In light of overall low product awareness levels, mass media campaigns that provide information about key features of financial products may be an effective means to increase beneficial use of financial products. The survey results suggest that effective channels to reach populations who are the least familiar with financial products would be mobile phones, TV, or radio (see Box 1). Awareness campaigns can also be used to disseminate the introduction of more sophisticated products among Mozambicans. For instance, pension products to adults not covered by public pension plans, or longer saving products, such as investments in funds or bonds, which would benefit the development of long term finance, such as housing finance, in Mozambique. Innovative and interactive measures, and edutainment in particular, should be considered to reach the adult target audience and to ensure that increased financial awareness translates into actions. The scientific field of behavioral economics has documented a plethora of behavioral biases which can prevent people from translating their knowledge into action. For instance, people tend to be biased towards the status quo and to choose the default option. They may also suffer from self-control issues, procrastination, overconfidence, or systematically underestimate the time to complete tasks (Buehler et al 2002). Recent research has shown that innovation on delivery matters. Conveying financial messages through innovative ways such as using popular TV soap operas, films, videos or radio programs can be quite effective, not only in improving financial knowledge but also in altering savings and borrowing behavior (Berg and Zia 2013, Coville et al 2014). Edutainment programs are also presumed to be more effective if messages are delivered in an engaging and entertaining manner through appealing stories that stick to memories, and if they are repeated and reinforced over time. For instance, in Kenya, a popular television drama, ‘Makutano Junction’, incorporated financial education messages into some of its stories. These messages aim to encourage people to save regularly or to open a bank account, rather than to keep money under a mattress. Other examples of the use of entertainment education for finance are ‘Scandal!’ in South Africa or ‘Mucho Corazon’ in Mexico. As with other soap operas, people watch these edutainment dramas because they identify with the characters and enjoy the stories; but in the course of watching the shows, they benefit from the financial capability enhancing messages. 7 Publications can also be a useful means of conveying financial capability enhancing messages, since each copy can be read by several people and can be retained for future reference. A diverse range of publications can be used, including leaflets, booklets, fliers and posters. Articles in newspapers and magazines are also important tools, especially if contained within general sections of the newspaper or magazine, rather than in specialist financial ones. Comic books have been found to be particularly effective in several countries, such as Kenya, India, and South Africa, where literacy levels are also low. In such cases comic books effectively facilitate discussion within the family on topics related to financial literacy. Furthermore, financial education can be effectively delivered through trusted organizations and individuals with whom the target audience deals with on a regular (day-to-day) basis. Managing one's personal finances is an important aspect of everyday living. Many organizations have an interest in helping people to become financially knowledgeable and capable. Especially to reach remote communities in rural areas, and groups that are hard to engage, it could be considered to collaborate with community organizations and trusted intermediaries such as community leaders, social workers, and to support them with resources, training or funding, if necessary. A promising way to increase the effectiveness of financial capability enhancing programs is to combine it with other interventions, such as the use of reminders. Rigorous impact assessments in Boliva, Peru, and the Philippines provide evidence that coupling a financial capability enhancing intervention with reminders via text messages is a promising way to address some behavioral biases and to induce behavioral change (Karlan et al 2010). Since Mozambicans appear to struggle with long-term financial decision making, periodic reminder messages could induce them to attend to the benefits and tasks of saving regularly and putting money aside for unexpected and old expenses. These interventions are also quite cost-effective and could be taken rapidly to national scale. The delivery of financial capability enhancing programs should further take advantage of ‘teachable moments’. Research shows that financial education works best when delivered to adults during teachable moments (Yoko et al 2012). Teachable moments are times in people's lives when they are more likely to be receptive to new information as they can relate it directly to their own life events. In terms of financial education, key teachable moments when one may re-examine his/her personal finances include marriage, new employment, and the launch of a new business. 8 Relationship between Financial Inclusion and Capability While Mozambican respondents who do not participate in financial markets are less aware of the services of financial institutions, their knowledge of financial concepts is comparable to those actively using financial products. On the one hand, this result suggests that a substantial fraction of Mozambicans that is not being reached by financial products, nevertheless has a similar understanding of financial concepts as respondents with established relations with financial institutions. On the other hand, it suggests that both financially excluded Mozambicans as well as those who are included deserve policy attention. A high number of Mozambicans who are financially active, lack basic knowledge and skills to make sound financial decisions. The likelihood of a less financially literate Mozambican using formal savings or credit is very similar to the likelihood of a person with better understanding of financial concepts. However, lack of knowledge of basic concepts does relate to usage of informal finance and savings. Individuals with low financial literacy are more likely to use informal savings and informal credit than individuals with higher financial literacy. Similarly, as respondents increase their awareness of financial products they rely more on formal financial institutions. Financial behaviors and attitudes are not notably different between those with and without a formal account, but when comparing the financial capability scores of users of various products, differences are more pronounced. Credit users, especially from informal sources, are more likely to overspend, and less likely to live within their means than the average respondent. Those who save on the other hand, regardless of whether they save formally or not, are more disciplined in spending their money. Recommendations In order to enable financially included Mozambicans to benefit from the products they use, financial knowledge and capability-enhancing programs could be combined with available financial products most people can access. Financial education programs could be tied to existing formal accounts most people can access and use such as at the time when they open an account or take out a loan. These programs should not only help to close existing gaps in their customers’ understandi ng of financial concepts but inform about the need to build up savings cushions for unexpected financial shocks and old age expenses. However, it must be ensured that any educational materials are truly informative, clear, impartial, and most importantly free from marketing. BdM may consider sharing the results of this survey widely with different financial service providers, but in particular with banks and MFIs to potentially develop products tailored to the needs of underserved segments of the population. Since savers are more likely to control their 9 spending than non-savers, it could make good business sense for financial service providers to develop products which meet the needs of underserved populations and help them to reach personal savings goals. For example, savings products have design features that affect the extent of people’s use of the product, such as commitment savings account or labeled accounts. The former consists of accounts where a certain amount of funds is deposited and access to cash is relinquished for a period of time or until a goal has been achieved. The latter describes accounts created with explicit savings goals, such as the establishment or expansion of a business, a car purchase, housing, or education (World Bank 2013a). 10 Financial Consumer Protection The survey results suggest that financially illiterate respondents are more vulnerable to encounter a conflict and to purchase and being sold products that do not meet their needs. Respondents who struggle to understand basic financial concepts seem to be less satisfied with bank products as compared to those with better understanding, suggesting the need for basic protective measures to ensure that they obtain the information they need to adequately understand the products they use. In addition, they also seem to be more vulnerable to experiencing a financial service provider conflict. Another interesting finding in the area of financial consumer protection is that consumers of financial services do not widely report complaints or other type of conflicts with providers, nor do they try to solve conflicts they encounter. Only 13 percent of the surveyed respondents state that they experienced a conflict with a financial service provider in the past 3 years. Less than half of those who encountered a dispute took action to try to solve it. Only 40 percent of those who did not experience a conflict indicated that if they faced a conflict they would try to resolve the disputes. Regarding the actions taken to seek redress, redress systems such as BdM’s team in charge of consumer complaints handling or legal courts were not sought at all by those who experienced a dispute. That courts were not considered at all can most likely be explained by perceived high costs and lengthy time of proceedings. That consumers do not turn to the BdM may be due to the fact that financial services contracts typically do not specify what a consumer should do in the event that he or she has a complaint, and the possibility of recourse to the BdM. Lack of trust or lack of awareness of which government authorities can be approached in the event of a dispute are also the most frequently cited reasons for not trying to solve a conflict. Recommendations These findings highlight that delivering financial education is not sufficient and needs to be complemented with measures to strengthen the financial consumer protection framework, such as regulations in the area of consumer disclosure. It needs to be ensured that consumers are provided with sufficient information to allow them to select financial products that are the most affordable and suitable. Therefore, in line with the recommendations of the 2012 Diagnostic Review of Consumer Protection and Financial Literacy 5 (CPFL), financial institutions should be required to provide a standardized ‘key fact statement’ that explains in plain language the key terms and conditions for each product. It would be beneficial to undertake consumer testing of key fact statements in order to ensure that 5 The recommendations highlighted in this report on financial consumer protection are important in the context of this survey. It should be noted that for strengthening the overall financial consumer protection framework additional measures which are detailed in the CPFL review play an important role. The review is available at: http://responsiblefinance.worldbank.org/diagnostic- reviews. 11 the presented information is properly understood by consumers and that the format covers all necessary information. In addition, it is suggested that BdM uses a variety of channels to provide consumers with comparable information on costs and terms of similar products, including internet, newspapers, community leaders, and consumer associations (World Bank 2013c). Financial institutions should also be required to inform their customers about their right to complain and about their complaints handling procedures. Legal or regulatory provisions should require financial institutions to provide customers with information on internal complaints handling procedures. This information should not only be disclosed in their products’ terms and conditions but also be visibly posted in branches and online. In addition, consumers should also be informed about formal redress systems such as BdM or legal courts. BdM should analyze consumer complaints statistics submitted by financial service providers and use this information as inputs to their supervisory and regulatory activities. All financial institutions should be obliged to share their complaints data with BdM. Based on the analysis of the consumer complaints and inquiries, BdM could propose guidelines, instructions or conduct awareness campaigns that address the main problems identified in such analysis. 12 Background on the Mozambique Survey The financial capability questionnaire used for this survey has been extensively tested in the context of low- and middle income countries. The survey instrument used is based on a questionnaire developed with support by the Russia Financial Literacy and Education Trust Fund (RTF) and is tailored to measure financial capability in low and middle income countries, although it can also be used in high income countries. Extensive qualitative research techniques were used to develop this survey instrument, including about 70 focus groups and more than 200 cognitive interviews in eight countries to identify the concepts that are relevant in low- and middle-income settings, and to test and adapt the questions to ensure that they are well understood and meaningful across income and education levels. The instrument is currently used or planned to be used in 14 countries in Latin America, Africa, Middle East and East Asia and the Pacific. The survey instrument used allows financial capability, financial inclusion, and consumer protection issues to be assessed and measured. Financial capability is measured by knowledge of financial concepts and products, and by attitudes, skills and behavior related to day-to-day money management, planning for the future, choosing financial products and staying informed. In order to jointly analyze financial capability and inclusion, the survey instrument captures information on usage of different kind of financial products and providers. The financial consumer protection section gathers information on incidence of conflicts with financial service providers and levels of satisfaction with financial products offered by different financial institutions. The survey instrument has been further customized to the policy priorities of BdM, through adding specific questions, for example relating to usage of money lenders and levels of satisfaction with products they provide. The Mozambique survey is nationally representative of the financially active population and comprises a total sample of 3,000 adults6. To fulfill the requirement of a scientifically sound survey which allows inferences to the whole universe of financially active adults in Mozambique who are either responsible for personal or household finances, probability sampling techniques were used to select a sample of 3,000 adults. Thereby, the most recent 2007 Mozambique Census of Population and Housing, kindly provided by the national statistical office (INE), was used as a sampling frame. The population was divided into 21 strata: 11 regions (Niassa, Cabo Delgado, Nampula, Zambezia, Tete, Manica, Sofala, Inhambane, Gaza, Maputo Provincia, and Maputo Cidade) and each region, except Maputo Cidade, was further divided into urban and rural strata. The sample was selected through a three stage cluster sampling. The primary sampling units (PSUs) selected at the first stage were enumeration areas (EA) delineated for the 2007 Mozambique census which were selected with probability proportional to size (PPS). The measure of size for each EA was based on the number of households in the sampling frame. Following the first stage selection of EAs, a household 6 Population aged 18 and older 13 listing was conducted in the chosen EAs. In each selected PSU, a sample of 20 households was selected from this list at the second stage, out of which 15 were targeted for surveying and 5 were reserve households for replacement purpose only. Finally, within each selected household, eligible adults either responsible for personal or household finances were randomly drawn by means of the Kish grid. Proper individual weights were calculated and used in the following analysis to adjust for varying probabilities of selection (design weights).7 Between August and December 2013, a Canadian survey firm implemented the survey using computer-assisted personal interview methods (CAPI). Étude Économique Conseil (EEC) Canada, a Montreal based survey firm, was hired to conduct the Financial Capability Survey in Mozambique. To ensure highest data quality and avoid common errors associated with paper-and-pencil surveys, an electronic version of the questionnaire including consistency checks were programmed and the survey was administered from tablet computers. Due to extensive efforts and different strategies used (e.g. training of enumerators on refusal conversion strategies, letters which were sent in advance to inform respondents about the surveys’ objectives, 5 contact attempts, etc.) the total non-response rate was less than 9 percent of the total targeted households. The adult population to which the results of this survey are meant to extrapolate has the following key characteristics: A large majority of the population (69 percent) live in rural areas, while the remaining 31 percent live in urban environments (see figure 38,). Slightly more than half of the population are women (51 percent, see figure 40). Ranking all individuals by their reported household income and dividing them into 4 groups, a third fall in the lowest income segment (less than 4500 MZN per month), 26 percent in the second lowest (between 4501 MZN and 6000 MZN), 24 percent in the second highest (between 6001 MZN and 9000 MZN), and 18 percent in the highest income group (more than 9000 MZN, see figure 39). Around 42 percent of the population is younger than 35, 46 percent ages from 35 to 55, and 12 percent of the population is older than 55 (see figure 41). In terms of the education attained, 10 percent of the population has tertiary education; 33 percent has some or completed secondary schooling, which includes lower and higher secondary degrees; 25 percent has completed primary schooling, while 32 percent has no schooling (see figure 42). Only 29 percent of the population is characterized as earning stable income, while the remaining 71 percent is facing irregular and uncertain income flows (see figure 43). The average number of adults per household is 4, whereas an average sized household comprises 6 people. As shown figure 44 in Appendix A, 59 percent of the respondents live in households with 4 to 6 members, around a quarter in households comprising 7 or more members. 7A sampling note is available upon request which entails detailed information about the sampling approach and the computation of weights which have been used in the subsequent analysis. 14 1. Financial Inclusion 1.1 Context Over the past two decades, Mozambique has been successfully implementing a series of reforms to its financial system which substantially improved the sector’s stability and depth. One of the most relevant ones has been the privatization of the financial sector. Starting in 2003, the role of private banks has gradually been increasing, currently representing about 95 percent of the total financial system assets (Mozambique Council of Ministers, 2013). Notable financial sector reforms that have been implemented range from regulation regarding the operation of the financial system to the establishment of units in BdM to increase supervision and transparency in the sector. These reforms, together with stable macroeconomic conditions, have resulted in increased financial sector’s assets and a steady decline in the fraction of non- performing loans of the system. However, financial sector growth does not necessarily translate in more financial inclusion. If there are barriers preventing financial products to reach groups of certain demographics, there might be scope for policy makers to work on this area. Compared to sixteen countries of Sub-Saharan Africa, Mozambique ranks fourth in the fraction of adults with an account at a formal financial institution8. According to 2011 Global Findex data, only in Mauritius, South Africa and Kenya more adults have access to these accounts. The Global Findex database further indicates that in Mozambique, Zimbabwe and Angola, on average 40 percent of adults report having an account at a formal financial institution. However, the gap between female and male access to formal accounts is wider in Mozambique than in Angola or Zimbabwe 9 . A similar pattern is observed in access to formal accounts by income distribution. In Mozambique, the difference in access between adults at the top 60 percent of the income distribution and those at the bottom is higher than in Angola or Zimbabwe (see figure 45 in Appendix). In terms of access to credit, 6 percent of adults in Mozambique report having a loan from a financial institution in the past year. While the gap between genders is not large (1 percentage point), the gap between adults at the top and the bottom of the income distribution is one of the widest of the region. 8.8 percent of the wealthiest adults have a formal loan, compared to only 1.8 percent of the poorest having one. Only in Kenya and Mauritius this gap is wider (see figure 46 in Appendix). In this report, financial inclusion is defined as the proportion of individuals that use financial services. As stated in the Global Financial Development Report 2014 (World Bank, 2013a), lack of usage of financial products does not necessarily mean lack of access. While some people may have access to financial services at affordable prices and may decide not to use them, others may lack access because of 8 This indicator from the Global Findex database includes accounts at a bank, credit union, cooperative, post office, or MFI. 9InMozambique, there is a 10 percentage point gap between female and male adults in access to formal accounts. 45 percent of male adults have a formal account, while only 35.5 percent of females do. In Zimbabwe and Angola the gap between male and female adults is of 5 and 1 percentage points respectively. 15 constraints such as excessively high costs, or unavailability of the services due to regulatory barriers or other factors. This chapter explores access to finance and the financial inclusion landscape in Mozambique, acknowledging that financial inclusion and access to finance are different issues. See Box 1 on how the results of this survey link to the wider financial sector strategy. Box 1: The WB Financial Capability Survey in the context of the wider financial sector strategy Following a broad and inclusive preparation process, the Council of Ministers approved in April 2013 the Mozambique Financial Sector Development Strategy (MFSDS) for 2013-2022. The objective of the MFSDS is to promote the development of a sound, diverse, competitive, and inclusive financial sector which provides citizens and businesses with convenient access to a range of appropriate and high quality financial services at affordable prices. In order to increase financial inclusion, the government has also included an emphasis on financial literacy and consumer protection in the MFSDS. In particular, the MFSDS includes: i) rapidly expanding financial literacy for all types of financial services to increase the public’s understanding of how financial se rvices can improve livelihoods, and its ability to access financial services; ii) putting in place a consumer protection framework both to protect consumers and to encourage new consumers to enter the market. In addition to the MFSDS, BdM’s financial inclusion commitment to the Alliance for Financial Inclusion (AFI) set out a substantial reform agenda for financial inclusion, as well as for financial literacy and consumer protection. The BdM’s commitment made at the AFI meetings in Cape Town on September 28, 2012, was to a Financial Inclusion Strategy (or action plan) that would cover financial inclusion, financial stability, financial literacy and consumer protection, and financial inclusion indicators. As an initial follow up to the MFSDS a diagnostic review of Consumer Protection and Financial Literacy has been conducted in 2013. The review provides a detailed assessment of the institutional, legal and regulatory framework for consumer protection in two segments of the financial sector in Mozambique: banking and non-bank credit institutions. The review was undertaken in response to a request received for WB technical assistance in the field of financial consumer protection made by BdM in November 2011. The WB Financial Capability Survey is a further priority follow up to the MFSDS, given i) that financial literacy/capability and improving financial access has been identified by the BdM as a priority area going forward, ii) the low levels of financial inclusion and the importance of financial capability in enabling people to take up and benefit from financial products and services, and iii) the lack of comprehensive, robust, and reliable data which has prevented policy makers so far from formulating specific policy actions and setting quantifiable and concrete targets. 16 1.2 Usage of Banks While banks dominate the financial sector in Mozambique, bank penetration is by no means homogeneous across the country, with rural populations in particular being excluded from bank services. Access to banks services has substantially improved over the last years, both in geographical and in demographical terms. In geographical terms, access to bank services improved from an average of 2.9 bank branches per 10,000 km2 in 2005 to 6.6 in 2012. Likewise, in 2012 Mozambique had on average 4.1 bank branches per 100,000 adults, as compared to the year 2005, when the country average was 2.2 branches of banks per 100,000 adults. However, the most significant improvement was registered in urban areas (BdM, 2013). In rural areas the problem of financial access is far more acute with an average of only 0.6 bank branches per 100,000 adults (Mozambique Council of Ministers, 2013). In a predominantly rural country, this gap is substantial. These disparities are reflected in the location of adults who know and use bank products. As figure 1 indicates, 74 percent of adults in Mozambique are familiar with the products offered by banks but only 52 percent of them report ever having used them. When examining regional patterns, people from urban areas are more likely to know and use bank products. Regression analysis (see tables 3 and 4) show that even after controlling for other socioeconomic and demographic factors, living in an urban neighborhood is strongly correlated with both the knowledge of these institutions and the usage of their products. Within rural and urban communities, the data suggests that differences in economic development of the area matters to explain the likelihood of using bank products. Holding constant the urban status of a neighborhood, people are more likely to use bank services in areas with shorter distances to MFIs branches and with better infrastructure (see table 5). Figure 1: Knowledge and usage of commercial banks by location of respondent Source: WB Financial Capability Survey, Mozambique 2013 Regarding usage of bank products, income is another characteristic that strongly predicts who uses these financial institutions, even after controlling for a set of demographic and socioeconomic factors (see table 4). As shown in figure 2, bank customers are more likely to be high and medium income. 17 Importantly, of all respondents with a fluctuating income, only 44 percent of them report having ever used banks. This result suggests there may be scope to provide some type of financial product to protect Mozambicans against income fluctuations and allow them to better smooth their consumption and plan their investments, especially since 72 percent of adults in Mozambique report having a volatile income. Figure 2: Knowledge and usage of commercial banks by income quartile and variability of income Source: WB Financial Capability Survey, Mozambique 2013 Mozambicans who use print, broadcast or internet media at a regular basis are also more likely to know and use bank products. According to regression analysis (see tables 3 and 4), even after holding income and other characteristics constant, more active media consumers are substantially more likely to know and have ever used bank services. Figure 3 presents the relation between usage and knowledge of bank products with an index of media consumption, defined as the number of media elements frequently used by respondents 10 . As seen in figure 3, knowledge about commercial banks and their use monotonically increases as individuals rely on more media elements. Interestingly, the relation between the media index and the fraction of people that have ever used banks is steeper than the relation of the index with the proportion of Mozambicans who know about bank services. Of all Mozambican respondents who do not use any kind of media element, while 56 percent of them know about bank services, only 25 percent have ever used banks. On the other extreme, 94 percent of the most active media users know and have ever used bank services. This pattern suggests that Mozambicans who are more likely to use media are not only more informed and literate Mozambicans, but also, with more means to obtain bank products. The types of media that respondents were asked about were newspapers (national and local), radio, TV, the internet, and 10 mobile phones. 18 Figure 3: Knowledge and usage of commercial banks by degree of media consumption11 Source: WB Financial Capability Survey, Mozambique 2013. Box 2: Media Consumption Overview While mobile phones, TV and radio are widely used in Mozambique, the penetration of print media and internet is biased towards more affluent, urban and highly educated segments of the population. Figure 4 reveals that even those at the bottom of the pyramid widely use TV and mobile phones. For instance, compared to 70 percent of urban dwellers, 58 percent of rural residents regularly watch TV. Mobile phones are even more popular, with penetration rates of 69 percent in rural areas as compared to 76 percent in urban areas. Similar differences in mobile phone usage can be observed between lowest and highest income earners and people with lowest and those with highest educational attainment. Print media and internet show most variation across different segments of the population and are hardly used at all by those with low educational attainment, rural dwellers, and people living on low and fluctuating incomes. Figure 4: Media consumption by different sociodemographic groups 11 Media consumption index refers to the number of media sources regularly used by respondents. 19 Source: WB Financial Capability Survey, Mozambique 2013. Media consumption index refers to the number of media sources regularly used by respondents. 1.3 Usage of Bank Products Among the range of products and services banks offer in Mozambique, the most commonly used products are bank accounts, followed by loans. As seen in figure 6, substantial disparities in the usage of bank products arise between urban and rural areas. In urban areas, 46 percent of Mozambicans currently have a deposit, saving or checking account with a bank, whereas in rural areas, this percentage drops to 19 percent. Despite this gap between urban and rural populations, figure 5 reveals notable regional differences in bank account usage. While the account penetration in southern provinces of Maputo, Gaza, and Inhambane averages 56 percent, only 21 percent of adults living in the northern provinces of Nampula, Niassa, and Capo Delgado have a bank account. The lowest account penetration is found in the central provinces of Sofala, Manica, Tete, and Zambezia, where only 16 percent of the population has an account with a bank. Disparities in the fraction of Mozambicans using bank loans can also be observed between urban and rural living environments. Only about 15 percent of adults in urban areas currently have a loan with a bank, compared to 7 percent in rural areas. 20 Figure 5: Bank account penetration in different regions in Mozambique Source: WB Financial Capability Survey, Mozambique 2013 Figure 6: Percentage of Mozambicans currently holding a financial product from a bank by urban status Source: WB Financial Capability Survey, Mozambique 2013 Regression analysis suggests that Mozambican respondents are substantially more likely to have bank accounts and bank loans at higher levels of income, even after controlling for community characteristics such as urban status, and other socio-demographic factors (see tables 6 and 7). As figures 7 and 8 show, both in urban and rural places, the use of bank accounts and credit loans is concentrated among the group of Mozambicans at the wealthiest income levels, particularly from urban areas. The regression analysis on the probability of having a bank account suggests that the development and infrastructure level of the community helps explain where bank accounts are concentrated. People are more likely to use bank services in areas with shorter distances to bank branches and with better water 21 supply. While more developed communities also have the expected positive relation on the probability of having a bank loan, these are not statistically different from zero. Figure 7: Percentage of Mozambicans with a bank account by income quintiles Source: WB Financial Capability Survey, Mozambique 2013 Figure 8: Percentage of Mozambicans with bank credit by income quintiles Source: WB Financial Capability Survey, Mozambique 2013 Regarding long-term finance, mortgages are rarely used by Mozambican adults, even in urban areas less than 5 percent of people report currently having one (see figure 9). According to the MFSDS, the lack of mortgage products offered by banks has resulted in a major shortage of affordable and adequate homes in urban areas of Mozambique (Mozambique Council of Ministers, 2013). The lack of finance for the acquisition of houses is driven by several factors, including supply-side issues such as limited coverage. The great majority of mortgages are provided for properties in the capital city of Maputo, in neighboring Matola, and in surrounding suburbs, with hardly any provided in other parts of the country. Demand-side constraints such as the lack of minimum amount of savings and prohibitive high costs for housing finance, on the other hand explain why, as seen in figure 9, mortgages in urban areas are 22 significantly concentrated among the wealthiest individuals. In urban areas, while 10 percent of adults in the fourth quartile of the income distribution have housing loans, only 4 percent of those in the third quartile have a mortgage. Figure 9: Percentage of adults in urban areas with a mortgage by income quartiles Source: WB Financial Capability Survey, Mozambique 2013 1.4 Usage of Nonbank Financial Institutions The financial system in Mozambique is heavily dominated by banks, but 48 percent of adults have never used their products. Banks however, constitute the most common financial provider of the country, followed by money lenders and microfinance institutions (MFIs), with 41 and 37 percent of adults having used these. Interestingly, bank users are more likely to have used products from other types of financial providers than non-bank users (see figure 10). Mozambicans that have never used bank products are also less likely to have purchased products and services provided by MFIs, insurance companies, money changers and even money lenders. Out of the 48 percent of the adult population that has never used banks, only 20 percent has used MFIs and less than 10 percent has used insurance products or money changers. In contrast, out of the 52 percent of the population of bank users, more than half has used MFIs or money lenders, and about 30 percent has used insurance products and money changers (see figure 10). Furthermore, even though on average Mozambican respondents have used at least one financial provider, this number masks significant differences. While 19 percent of adults have used at least three different financial institutions, more than a third of the population has never used any financial provider12 (see figure 11). The data suggest two patterns. First, an important fraction of the population in Mozambique does not use financial institutions, either because they choose not to or because they find it difficult to do so. Second, while bank users may be more active in the financial market than non-bank users, banks and other financial institutions may also be targeting similar clients, and those who are excluded from banks are not being reached by other financial providers. 12 The financial providers that are included in this statistic are banks, insurance companies, MFIs and money changers. 23 Figure 10: Percentage of Mozambicans that have ever used financial institutions Source: WB Financial Capability Survey, Mozambique 2013 Figure 11: Percentage of Mozambicans by the number of financial institutions that they have used Source: WB Financial Capability Survey, Mozambique 2013 The data suggests substantial overlap on the type of clients targeted by banks and by certain financial institutions in Mozambique, such as insurance companies, money changers and to some extent, MFIs. Regression analysis shows that income, access to media and community characteristics are the factors that best predict who the users of different financial providers are (see table 8 – table 11 ). As with banks, clients of money changers and insurance companies are concentrated at the top of the income distribution. In contrast, MFI clients are not from the lowest income levels, but it is more likely for Mozambicans above the median income to have used MFIs than for adults below it. Different from the other financial providers, clients of money lenders are not associated with higher income. If anything, money lender clients are more likely to be less educated and from a less favorable background. Interestingly, individuals with fluctuating income are less likely to have ever used the products of insurance companies or of any other financial institution. This result suggests there is room for policies to target this population, who currently do not use any financial instrument to protect against volatile income. 24 In terms of geographic characteristics of clients, banks, insurance companies and money changers concentrate their services among urban clients from better-off communities. By contrast, clients of MFIs are more likely to be both from urban and peri-urban communities rather than inner city neighborhoods. Not surprisingly, time of commute matters - distance to the closest MFI branch is negatively associated with the likelihood of using these institutions. Mozambicans that use money lenders are more likely to be from areas with less infrastructure and lower population, and where the commute to the closest bank branch takes longer. 1.5 Usage of Products of Nonbank Financial Institutions Financial providers in Mozambique offer a range of services that span from investment products to pension plans, insurance, wire transfers, etc. (see figure 12). The type of saving products Mozambicans currently use also varies substantially between urban and rural areas (panel 1 of figure 12). People in urban areas are compared to rural dwellers more likely to save in formal institutions or to use community based savings methods13. Even though it is less likely for rural adults to save formally, 47 percent of both rural and urban Mozambicans report saving at home. This suggests that relative to urban areas, in rural areas there may be various barriers that prevent people to save at formal institutions. More sophisticated saving products, such as pensions or investment in stocks are not very common among the population. Private pensions are only used by 1 and 4 percent of rural and urban adults, which is consistent with the fact that only few private pension companies exist in Mozambique (Mozambique Council of Ministers, 2013). Similarly, Mozambique’s stock market is in its early development, with only 3 percent of urban Mozambicans reporting having invested in stocks, bonds or funds. As with savings, usage of credit- both from formal and informal providers-, insurance products and money transfer services is higher in urban areas (panels 2, 3 and 4 of figure 12). This may partially reflect that urban Mozambicans have a higher demand for financial products, but also that low penetration of financial institutions in rural areas translates in lower usage of their services. 13Community based savings methods refer to ASCAs (Accumulating Savings and Credit Associations), OPEs, Xitiques, and Conta Familias. 25 Figure 12: Usage of financial products in rural and urban areas Source: WB Financial Capability Survey, Mozambique 2013. 1.6 Barriers to Formal Account Ownership Whereas 29 percent of urban Mozambicans do not have basic formal accounts, in rural areas, 58 percent of adults lack one. While lack of formal accounts does not necessarily mean lack of access, only 5 percent of urban residents who lack accounts state that they do not need these products (see figure 13). Much less, only 1 percent of rural dwellers without account report that they do not need one. Figure 13: Percentage of Mozambicans with no formal accounts reporting they do not need this product Source: WB Financial Capability Survey, Mozambique 2013 26 The most important reason reported for not having an account is that people do not have enough money to use them. The Financial Capability Survey asked respondents without a formal account to report why they do not have an account at a financial institution. As figure 14 indicates, the most common reason in urban areas is that Mozambicans do not have enough money to use them. Regression analysis (see table 12) indicates that it is more likely for adults living in inner cities and lower education to state this reason. Other relevant reasons mentioned by urban respondents include that formal accounts are too expensive and interestingly, that they are too far away. The same reasons were mentioned by respondents in rural areas, but relative to their urban neighbors, rural respondents complain more about accounts being too far away and too expensive. Another important barrier to formal account ownership is affordability, which is a much greater barrier in rural areas. Fixed fees and costs of opening and maintaining an account can make small transactions unaffordable for large segments of the population. As can be seen in figure 14, 20 percent of those without an account who live in rural areas report that they cannot afford it. Although this number is lower for urban populations, it is still the second most important reason for not having an account. High costs and fees most likely reflect lack of competition as well as underdeveloped physical and institutional infrastructure. Taken together, even 19 and 26 percent of urban and rural respondents without formal accounts state that they do not know how to open them or that they do not trust them. The national average of those without formal account who reportedly do not know how to open an account or lack trust in them is 25 percent. These results suggest that financial illiteracy may be a significant barrier in the financial market of the country, and that information campaigns that familiarize Mozambicans with financial products may be one approach towards financial inclusion. Figure 14: Reasons for not having a formal account in rural and urban areas Source: WB Financial Capability Survey, Mozambique 2013 27 2. Financial Capability Financial Capability is the internal capacity to act in one’s best financial interest, given socioeconomic environmental conditions. It therefore encompasses the knowledge, attitudes, skills, and behaviors of consumers with regard to managing their resources and understanding, selecting, and make use of financial services that fit their needs. 2.1 Knowledge of Financial Concepts and Products The recent global financial crisis has highlighted the importance of financial knowledge and skills (financial literacy) for peoples’ ability to take sound financial decisions and to benefit from the financial services they use. It is a well-accepted hypothesis that limitations in consumers’ ability to fully understand the financial products and risks they had taken on, contributed significantly to the worst financial crises since the great depression (Geradi et al. 2010; Klapper et al. 2012). Due to increased availability of credit in Mozambique, the continuous growth of microfinance, and the development of branchless banking networks, financial products and services are becoming available to populations which have been formerly disconnected from the formal financial system. While these developments provide benefits, they also bear risks which may be unfamiliar to existing and new customers. To be able to benefit from these new opportunities without being exposed to undue risks, a certain level of financial knowledge and skills is required. In addition, limited financial knowledge can constrain the take up of financial products and services. While lack of money, affordability and long distances were the most cited reasons for not having an account, 25 percent of Mozambicans without a formal account state that they do not know how to open an account or that they do not trust these products. This chapter explores demand side constraints in uptake and beneficial use of financial services. In particular, it tries to identify gaps in financial knowledge that need policy attention as well as vulnerable groups that display limited knowledge and understanding of financial concepts and products. 2.1.1 Knowledge of Financial Concepts Financial knowledge levels of fundamental concepts are a significant challenge in Mozambique, as well as in many countries across different income levels. Table 1 shows for 21 countries the proportion of adults with understanding of basic concepts such as inflation, simple and compound interest and who are able to perform simple divisions. While survey respondents from Mozambique demonstrate relatively high comfort levels in solving simple numeracy tasks, compared to respondents from economies with different income levels, main areas for improvement, such as understanding of the working of compound interest and what inflation is, appear to be more of a challenge from a cross-country perspective. 28 Table 1: International comparison of knowledge of basic financial concepts (in % of adults) Simple Compound Simple Country Year Inflation Interest Interest division Albania 2011 61 40 10 89 Armenia 2010 83 53 18 86 Colombia 2012 69 19 26 86 Czech Republic 2010 80 60 32 93 Estonia 2010 86 64 31 93 Germany 2010 61 64 47 84 Hungary 2010 78 61 46 96 Ireland 2010 58 76 29 93 Lebanon 2012 69 66 23 88 Malaysia 2010 62 54 30 93 Mexico 2012 55 30 31 80 Mongolia 2012 39 69 58 97 Mozambique 2013 28 78 28 93 Norway 2010 87 75 54 61 Peru 2010 63 40 14 90 Poland 2010 77 60 27 91 South Africa 2010 49 44 21 79 Turkey 2012 46 28 18 84 Tajikistan 2012 17 35 56 97 United Kingdom 2010 61 61 37 76 Uruguay 2012 82 50 N/A 86 Source: WB Financial Capability Surveys and OECD National Financial Literacy and Inclusion Surveys To assess respondent’s financial knowledge and their basic numeracy skills, 7 questions were added to the 2013 Mozambique Financial Capability Survey, covering basic calculus and financial concepts such as interest rates, inflation, compound interest, risk diversification, and the main purpose of insurance products. These questions have been asked because they capture financial concepts and skills which are widely considered as being crucial for informed savings and borrowing decisions as well as for being able to manage risks more effectively and or to take advantage of investment opportunities. We construct a financial literacy index based on the number of correct responses provided by each survey participant to the seven financial literacy questions. This index ranges from 0 to 7, whereby 0 indicates respondents who struggle the most with correctly answering any of these questions, while a score of 7 indicates survey participants with good understanding of fundamental financial concepts and the ability to perform simple mathematical calculations. 29 Box 3: Financial Literacy Quiz Question 1 Imagine that five brothers are given a gift of 10,000 MZN. If the brothers have to divide the money equally, how much does each one get? Question 2 Now, imagine that the five brothers have to wait for one year to get their part of the 10,000MZN and inflation stays at 10%. In one year’s time will they be able to buy:  More with their share of money than they could today  The same amount  Less than they could buy today  It depends on the types of things that they want to buy (do not read out this option) Question 3 Suppose you put 10,000 MZN into a savings account with a guaranteed interest rate of 2% per year. You don’t make any further payments into this account and you don’t withdraw any money. How much would be in the account at the end of the first year, once the interest payment is made? Question 4 How much would be in the account at the end of five years? Would it be:  More than 11,000 MZN  Exactly 11,000 MZN  Less than 11,000 MZN  It is impossible to tell from the information given Question 5 Let’s assume that you saw a TV-set of the same model on sales in two different shops. The initial retail price of it was 10,000 MZN. One shop offered a discount of 1,500 MZN, while the other one offered a 10% discount. Which one is a better bargain, a discount of 1,500 MZN of 10%?  A discount of 1,500 MZN  They are the same  A 10% discount Question 6 Which of the following statements best describes the primary purpose of insurance products?  To accumulate savings  To protect against risks  To make payments or send money  Other Question 7 Suppose you have money to invest. Is it safer to buy stocks of just one company or to buy stocks of many companies?  Buy stocks of one company  Buy stocks of many companies 30 The survey results suggest that on average respondents were able to correctly answer 3.7 out of 7 questions on financial literacy. As shown in figure 15, the majority of survey participants were able to provide between 3 and 5 (around two thirds of the sample) correct answers. Giving correct responses to 6 or more questions seemed, however, to be a challenging task which was only achieved by around 9 percent, while only slightly more than 1 percent was able to provide correct responses to all 7 financial literacy questions. A more concerning finding is, that a significant proportion of respondents, around one fifth (18 percent), was not able to provide more than 2 correct answers, while 9 percent of the sample struggled in answering more than 1 financial literacy question correctly. Figure 15: Distribution of financial literacy scores Source: WB Financial Capability Survey, Mozambique 2013 One can conclude that whilst most can perform simple financial calculations, they may lack the specific knowledge required to make sensible savings and borrowing decisions. Figure 16 reveals that almost all respondents were able to perform simple divisions (93 percent) and around three quarters were comfortable with simple interest rate calculations. By contrast, understanding of fundamental financial concepts is more challenging for the majority of the sample. 43 percent of the surveyed population understands the main purpose of insurance products. A similar proportion of the sample (42 percent) understands that holding stocks from different companies implies less risky returns than holding stocks from a single company). Slightly less, around 38 percent, seems to be comfortable in solving simple numeracy task in order to identify better bargains. The most notable knowledge gaps which deserve most policy attention are that only around a quarter of the survey participants understand the working of compound interest (28 percent) and how inflation affects their savings (27 percent). 31 Figure 16: Financial literacy quiz overview Source: WB Financial Capability Survey, Mozambique 2013 Figure 17: Education levels of populations with low and high financial literacy scores Source: WB Financial Capability Survey, Mozambique 2013 Vulnerable groups who struggle the most in understanding basic financial concepts are more likely to have low educational attainment, less likely to be formally employed, and they live further away from the next bank branch. As shown in figure 17, survey participants who provided 2 or less correct responses to the financial literacy quiz-type questions appear to be more likely to have lower educational attainment than their counterpart group with better understanding of financial concepts. For instance, while 12 percent of those who achieved a score higher than 2 on the financial literacy quiz have tertiary education, only 2 percent of those who answered 2 or less questions correctly can be found in the group with the highest educational attainment. Similarly, those who are challenged with answering more than 2 of the financial literacy questions correctly are also less likely to be formally employed compared to those who 32 reach higher financial literacy scores. Interestingly, proximity to banks matters too and those who live further away from the next bank branch are more likely to struggle in understanding basic financial concepts. Likewise, regression analysis shows (see table 13) that even after controlling for socioeconomic and demographic factors the availability of a primary school, such as distance to school, is associated with better understanding of financial concepts. However, the magnitude of these two infrastructural effects appears to be rather small. In contrast, a better understanding of financial concepts is strongly correlated with individual characteristics such as higher education, formal employment and living in urban habitats. As may be expected, regression analysis reveals (see table 14), that higher financial literacy scores correlate with higher educational attainment, even after controlling for other characteristics. Likewise, the knowledge gap between those who are formally employed and those who are not employed is significant. Compared to rural dwellers, those who live in urban areas have significantly better understanding of financial concepts. Notably, even though income levels are not significantly correlated with higher financial literacy scores, income uncertainty is. Those who live on varying incomes score on average significantly higher on the financial literacy quiz-type questions than their counterparts whose income is not subject to income fluctuations. Moreover, higher levels of financial knowledge relate to enabling environmental factors such as the proximity to banks and the development level of the location people live in. The longer it takes to get to the next bank branch, the less likely it is that survey participants are familiar with financial concepts, even after controlling for other characteristics by means of regression analysis (see table 13). Similarly, indicators which proxy for the socio-economic and infrastructural development of the location, such as low criminal rates or proximity to hospitals, seem to matter for better understanding of financial concepts. As compared to those residents who live in safe areas with basic health infrastructure, those who live in areas with high criminal rates or larger distances to the next hospital are less likely to answer more financial literacy questions correctly. This may be due to the influence that these socio-economic and infrastructural factors exert on the decision of banks to open their branches in certain areas. Areas for improvement identified through the objective financial knowledge quiz are also reflected in participants’ self-assessment of their levels of awareness and understanding of financial concepts. In order to compare the objective findings of the financial literacy quiz into the context of subjective education needs, respondents were also asked to self-assess their awareness and understanding of financial terms and concepts such as interest rates, insurance products, exchange rates and inflation. As seen in figure 18, reported awareness and understanding of interest rates is wide-spread. While only around 7 percent of the respondents stated that they hav e never heard about the term ‘interest rate’, the overwhelming majority of respondents (79 percent) indicated that they have not only heard about it, but also know what it means. The latter number is fairly close to the proportion of adults who provided a correct answer to the quiz question on simple interest (78 percent). Similarly to the results of the financial literacy quiz, respondents’ self-assessment acknowledges that they struggle to understand the main purpose of insurance products or what inflation is. 33 The high correlation between what respondents know and what they think they know is indicative of peoples’ readiness to accept their areas for improvement and thus creates demand for efficient financial education. While most of those who stated having knowledge about financial products and terms also answered correctly the financial knowledge quiz-type questions, a relatively small fraction of respondents who stated that they lack understanding managed to score on the respective objective measures (see figure 19). Notably, for the concept of inflation a wide gap between self-perceived and actual understanding appears to exist. However, these results prove an indication that people are very well aware of their areas for improvement which may lay the foundations and create demand for efficient financial education programs. Figure 18: Reported awareness & understanding of financial terms Source: WB Financial Capability Survey, Mozambique 2013 Figure 19: Comparison of reported understanding and financial literacy quiz results Source: WB Financial Capability Survey, Mozambique 2013 34 2.1.2 Knowledge of Financial Products In order to assess survey participants’ awareness levels of financial products the financial capability survey captured information on peoples’ familiarity with products offered by different financial service providers. In particular, survey participant were asked if they are familiar with products offered by banks, MFIs, insurance companies, community savings groups, money changers, money lenders, and brokerage houses. We construct a financial products awareness index based on the number of financial products known, as indicated by each survey participant. This index ranges from 0 to 7, whereby 0 indicates respondents who are not familiar with products offered by any type of provider. Respondents with a score of 7 on the other hand stated familiarity with products offered by the seven providers that the survey asked about. Figure 20: Distribution of financial products awareness scores Source: WB Financial Capability Survey, Mozambique 2013 As far as the average number of financial products known is concerned, respondents are familiar with products provided by 3.3 different types of providers. Figure 20 shows that around two thirds of the sample indicated to be familiar with between 3 to 5 products, whereas around 9 percent are familiar with financial products provided by 6, and only around 2 percent with products offered by 7 different providers. The most concerning fact is, however, that even 22 percent are not aware of financial products provided by any type of provider. A deeper exploration into the type of financial products known reveals that survey participants are mainly familiar with bank products (74 percent), followed by products offered by MFIs (71 percent) and money lenders (64 percent). Much less, around two fifth of the sample indicated that they are familiar with the products offered by community savings groups (43 percent) and insurance companies (38 percent). Money changers and their products are known by a third of the respondents (see figure 21), whereas brokerage houses and their services are only known by 7 percent of the sample which can be 35 explained by the fact that the capital market in Mozambique is currently in a nascent stage, as is the case in many countries at a similar income levels (Mozambique Council of Ministers, 2013). Figure 21: Knowledge of financial products offered by different providers Source: WB Financial Capability Survey, Mozambique 2013 Respondents who are the least familiar with financial products offered by different providers tend to live in rural neighborhoods and on low and irregular income streams. The need to manage low and erratic income flows as well as living in rural environments is a strong predictor for being less familiar with products from a variety of financial service providers, in particular with insurance products. Insurance products could, however, enable them to better deal with bad events when they occur. Even though the insurance sector in Mozambique is still under-developed, this result suggests that improving awareness of insurance is the key to increasing demand and should complement any policy actions to target this population and to grow the market for insurance. Another pattern which emerges is that better financial knowledge of products and services highly correlates with the regular consumption of different types of media. Not surprisingly, people who try to stay informed by using different types of media at a regular basis are substantially more likely than less active media consumers to demonstrate familiarity with all types of financial products, not only bank products (see figure 22). 36 Figure 22: Percentage of Mozambicans that know about different providers by number of media used Source: WB Financial Capability Survey, Mozambique 2013 2.2 Financial Behavior and Attitudes Even if people possess the knowledge of basic financial concepts and products they may struggle to translate it into action. To identify the role that attitudes play in individuals' financial decisions and to see how attitudes translate into financial behavior, the survey contains questions on different aspects (components) of financial capability that include attitudes/motivations and behaviors. This chapter gives an overview of strengths and areas for improvements surveyed Mozambicans show regarding relevant financial behaviors and attitudes. In the Mozambique data set, 10 main components of financial capability can be identified, some of which refer to behaviors, and others to attitudes or motivations. Each financial capability component is measured through a combination of relevant questions. These are identified by using a statistical technique called principal component analysis (PCA). PCA is a data reduction method that finds a small number of linear combinations of those variables that explain most of the variance in the data. The method is used to aggregate the variables that measure different nuances of the same component in order to obtain a single indicator (or score) for that component. Each component score ranges between 0 (lowest score) and 100 (highest score). The following eight components measure behaviors related to financial capability: budgeting, not overspending, living within means, monitoring expenses, saving, planning for unexpected expenses, making provisions for old age, and choosing products. More specifically, ‘budgeting’ measures the extent to which people plan how to use their money and whether they adhere to the plan; ‘not overspending’ assesses whether people refrain from spending their income on non-essentials or on things 37 they cannot afford; ‘living within means’ measures the level of borrowing and whether people borrow to buy food and other essentials; ‘monitoring expenses’ measures the ability to follow planned budgets and expenses; ‘saving’ measures whether people see themselves as trying to save for the future, trying to save for emergencies, and trying to save even if a small amount; ‘planning for unexpected expenses’ indicates whether people could cover an unexpected expense equivalent to a month's income and whether they worry about it; ‘making provisions for old age’ indicates whether people have strategies in place that allow them to cover for expenses in old age; and ‘choosing products’ indicates whether people search for alternatives, check terms and conditions, get information before selecting financial products, and search until they found the best products for their needs. The last score is only calculated for those who have personally chosen a financial product in the past 5 years. Two financial capability components refer to attitudes and motivations, such as farsightedness, and attitude towards information. In particular, ‘farsightedness’ measures whether people agree or disagree with statements such as ‘I live for today’, ‘The future will take care of itself’, ‘I only focus on the short term’; the measure ‘attitudes towards information’ is a combination of getting information and advice before making financial decisions, learning from others, and having many aspirations. Figure 23: Average financial capability scores Source: WB Financial Capability Survey, Mozambique 2013 Compared to other aspects of financial capability, survey participants show strengths in areas related to day-to-day money management. Figure 23 shows all scores for different aspects of financial capability in increasing order. As can be seen, the highest average score (74) is obtained for controlled 38 budgeting, followed by scores which are related to managing day-to-day finances, such as living within means (64), not overspending (63), and monitoring expenses (61). Financial attitudes and behaviors, on the other hand, which relate to thinking about the future, and putting money aside for unexpected or old age expenses, seem to be a major challenge in Mozambique. Figure 23 also reveals that compared to day-to-day money management, respondents score on average much lower with regard to coping with unexpected shocks (45), saving regularly, even if only a little (42), thinking of the future (40) and making provisions for old age expenses (40). These low scores are worrying given their implications for people’s ability to smooth consumption, to cope with economic shocks, to generate lump sums for productive investments, and ultimately for their long-term wellbeing. The lowest score is, however, found for financial behaviors that relate to choosing financial products and services (34). The lowest overall financial capability score indicates that consumers’ ability to take advantage of available financial services may be limited. The survey results show that slightly less than half of those who have chosen a financial product in the past 5 years were searching for information from a range of resources or searched until they found the best product for their needs, while only 37 percent considered many alternatives before deciding which product to get. An international comparison to survey participants in 9 countries confirms that Mozambicans respondents are especially good in managing day-to-day finances, but are among the most challenged in terms of putting money aside for unexpected and old age expenses and choosing financial products. Table 2 compares the financial capability scores Mozambicans achieved in 10 different areas to the ones of respondents in various countries in which a similar survey has been conducted. While survey participants from Mozambique demonstrate strengths in budgeting and monitoring how they had spent their money, they display relatively weaker performance in saving, putting money aside for unexpended and old age expenses, and in particular in choosing financial products. Table 2: Cross-country comparison of different financial capability scores Planning for Controlled Not over- Monitoring Using Country unexp. budgeting spending expenses information expenses Armenia 74 84 63 69 64 Colombia 80 79 36 80 59 Lebanon 40 70 44 71 73 Mexico 52 70 41 72 64 Mongolia 65 71 N/A 65 N/A Mozambique 74 63 61 56 45 Nigeria 78 71 48 N/A 71 Tajikistan 81 94 N/A 87 N/A Turkey 60 66 50 69 68 Uruguay 71 84 48 76 55 39 Planning for one’s Choosing financial Country Saving Far-sightedness own future products Armenia 46 28 100 59 Colombia 45 37 67 57 Lebanon 40 55 71 63 Mexico 57 35 65 59 Mongolia 62 60 N/A 49 Mozambique 42 40 40 34 Nigeria 55 N/A N/A N/A Tajikistan 66 84 N/A N/A Turkey 30 50 72 52 Uruguay 44 35 60 N/A Source: WB Financial Capability Surveys In Mozambique, low-income populations and high-income earners seem to have complementary skills. As compared to higher income segments, low income groups are mastering the task of monitoring their expenses and seem to be quite farsighted. However, as would be expected, their ability to cope with unexpected shocks seems to be limited by their scarce resources. Those respondents who live on higher incomes are more inclined to build cushions for unexpected expenses in comparison to the lowest income group, even after controlling for other demographic and socioeconomic factors (see table 16). Income matters not only in terms of levels, but even more if it is subject to fluctuations. Regression analysis also show, that despite being especially good in living within their means, those with varying incomes struggling more with setting up a budget and adhering to it, monitoring their expenses, and choosing financial products. In addition, they tend to think less about the future and have more difficulties with setting aside funds for unexpected expenses than their counterparts with stable incomes. Consequently, daily hardship and the constant struggle with solving immediate problems seem to draw their attention away from their longer-term considerations and needs. Another important personal characteristic which correlates with higher scores in several financial capability areas is higher education. As can be seen in figure 24, those with higher educational attainment appear to outperform those with lower education levels in terms of budgeting. Similarly, those with higher education levels also appear to experience less difficulties in living within their means, are more inclined to save regularly, even if only a little, and to cover for old age expenses than their counterparts with lower educational attainment. However, as compared to those without school, those with higher education levels appear to be more challenged with choosing financial products which fit their needs best. 40 Figure 24: Average budgeting score by education levels in urban and rural areas Source: WB Financial Capability Survey, Mozambique 2013 Living in an urban neighborhood seems to be an important enabler in terms of sound financial- decision taking. Residents living in rural areas display relative strengths in monitoring their expenses but struggle more than urban dwellers in a number of areas. As compared to urban populations, rural dwellers have more difficulties in terms of budgeting, not overspending, covering for unexpected shocks and choosing products that meet their needs. Rural residents also seem to have a less forward-looking attitude than their counterparts living in urban environments. Figure 25: Financial Capability in Choosing Financial Products (Left) and Being Far-sighted (Right) by Region Source: WB Financial Capability Survey, Mozambique 2013 Noticeable differences in financial capability scores across regions can only be observed in a few areas. As can be seen in Figure 25, those who live in southern parts of the country appear to be more 41 inclined to think about the future and demonstrate strengths in choosing financial products and services which fit their needs best. Better results in the area of choosing financial products, in which survey participants score lowest, are related to greater regular consumption of different types of media. The particularly low average scores in the area of choosing financial products among those who have personally selected a financial product in the past 5 years raise the question which personal factors can be associated with a better beneficial use of financial products. Figure 26 reveals that the regular use of different media types is an important personal characteristic which highly correlates with the ability to select appropriate products, especially in rural areas. Figure 26: Average choosing financial products score by media consumed in urban and rural areas Source: WB Financial Capability Survey, Mozambique 2013 42 3. Relationship between Financial Inclusion and Financial Capability There is little doubt that financial capability and financial inclusion influence each other. While lack of knowledge about financial products may hinder their use, it may also be the case that as people begin using financial services, they become more familiarized with them and knowledgeable about them, in a “learning by doing” fashion. While disentangling a causal link between financial inclusion and financial capability is beyond the scope of this report, this chapter presents an overview of who the financially excluded in Mozambique are and how their financial capabilities compare to those financially included. Results of the Financial Capability Survey indicate that in Mozambique, low understanding and lack of trust are excluding a substantial fraction of adults from using basic financial products, such as formal accounts. As outlined in chapter 1, 25 percent of Mozambicans state that they do not have a formal account because they do not know how to open them or because they do not trust them. In rural areas, these two reasons are more frequently mentioned than the service being too costly, the branch being too far away, or not having enough money to use the account. In urban areas, not having enough money to use the account is the only reason that is more common than lack of trust or knowledge about the product. Mozambicans without formal accounts are also less aware of the services of different financial providers, and thus, less likely to use them. According to the financial products awareness index described in chapter 2, even after controlling for a set of factors ranging from socioeconomic, demographic and village characteristics, awareness of financial providers is an important predictor of activity in the financial market. Compared to those with high awareness about financial products, Mozambicans who are less knowledgeable of the services of financial institutions are less likely to have a formal account and to use different financial instruments - such as credit, insurance or savings- from various financial providers (see figures 27 and 28). This however, may simply reflect that people gain awareness of financial institutions and their products as they become more active in the financial market. A substantial proportion, one out of five Mozambican respondents display lack of knowledge about any of the financial products assessed in the survey. Of these, none of them has a formal account or any other type of financial product. In contrast, Mozambicans who are more likely to know about the services of different financial providers are consistently more active in the financial market. 43 Figure 27: Financial products awareness score of Mozambicans with and without formal accounts Source: WB Financial Capability Survey, Mozambique 2013 Even though the usage of informal providers is high, as Mozambicans increase their awareness of financial products they rely more on formal financial institutions. As figure 28 indicates, the fraction of adults saving at a bank steadily increases with the awareness index. In contrast, usage of informal savings slightly decreases. Likewise, the fraction of adults who use bank credit increases at a steeper rate than the proportion of Mozambicans using informal finance. Figure 28: Usage of financial products by awareness of financial products score Source: WB Financial Capability Survey, Mozambique 2013 44 While Mozambicans who do not participate in the financial market are less aware of the services of different financial institutions, their financial literacy level is comparable to those actively using financial products. According to the financial literacy index discussed in chapter 2, Mozambicans with and without formal accounts have comparable knowledge levels on basic calculus and fundamental financial concepts (see figure 29). Figure 29: Financial literacy scores of Mozambicans with and without formal accounts Source: WB Financial Capability Survey, Mozambique 2013 A similar surprising finding is that the least financially literate Mozambicans are as likely to use formal and community based savings (e.g. ASCAs) as the most literate ones (see figure 30). However, as financial literacy increases, the likelihood of using informal savings mechanisms, such as saving money under the mattress, declines. A similar pattern is found regarding credit. More financially literate adults are less likely to rely in credit, especially if coming from an informal provider. On one hand, these findings suggest that financial products are not reaching Mozambicans who in terms of financial understanding and ability are at least as able as Mozambicans with established relations with financial institutions. On the other hand, these patterns also indicate that policymakers need to pay attention not only to the financially excluded population, but also to Mozambicans who despite being financially active, lack basic financial knowledge and skills to make informed and sound decisions on their savings, loans or insurance plans. Figure 30: Usage of financial products by financial literacy score 45 Source: WB Financial Capability Survey, Mozambique 2013 Regarding behaviors and attitudes, they are overall not too different between those respondents with and without a formal account. Using the same financial capability scores as described in chapter 2, survey results suggest that Mozambicans who do not have a formal account are as likely to live within their means, save for the future, not overspend, or plan for unexpected expenses as those with formal accounts. However, figure 31 indicates that Mozambicans who have a formal account are marginally more likely to monitor their expenses and obtain information and advice before making financial decisions. Figure 31: Financial behaviors & attitudes of Mozambicans with and without formal accounts Source: WB Financial Capability Survey, Mozambique 2013 Interesting patterns appear when examining the financial capability scores of Mozambicans who use different financial products (tables 18 and 19 present the regression results). Those who save, regardless of whether they save with formal providers, use community based savings methods or save at home, are more disciplined with their spending than non-savers (see figure 32, panel A). More disciplined Mozambicans are actually more likely to be saving in more formal places, even after controlling for a range of socio-economic, demographic and location variables. Mozambicans who formally save at a bank are also better in monitoring and planning their expenses (see figure 32, panel D). The budgeting behavior and financial capability scores of credit users are substantially different from the rest of the population. As indicated in figure 32 (panel C), Mozambicans who currently have credit are significantly less likely to live within their means than Mozambicans without credit. This may 46 reflect that Mozambicans are using credit when their income flows are not enough to cover their expenses. Similar to the pattern observed in savings, as Mozambicans borrow from more informal providers their budgeting and financial capability scores decrease. Those who borrow from informal lenders are the least likely to live within their means, and as seen in figure 32 (panel B), informal lender borrowers are also substantially less likely to make provisions for old age and more likely to over spend. Figure 32: Financial behaviors & attitudes of Mozambicans with and without different financial products Source: WB Financial Capability Survey, Mozambique 2013 47 4. Financial Consumer Protection In addition to peoples’ ability to take sounds financial decisions, the recent financial crisis has highlighted the importance of financial consumer protection to protect consumers from abusive sale practices and to level the playing field between providers and consumers of financial services. Financial consumer protection is about ensuring a fair interaction between providers and consumers of financial services. A consumer protection regime is essential in counterbalancing the inherent disadvantage of financial service consumers vis-à-vis the power, information, and resources of their providers. Without basic protective measures, consumers can be challenged or may find it costly to obtain sufficient information or adequately understand the financial products that they use. Thus, financial consumer protection is necessary to ensure stable financial markets in Mozambique while ensuring that expanded access benefits consumers and the overall economy. Given the relatively low levels of financial inclusion in Mozambique, a number of initiatives are planned or already underway to increase financial sector outreach to formally excluded populations (see Mozambique Council of Ministers, 2013). Increased access to finance can result in substantial positive effects, both on the macro level as well as on the level of individuals. However, it can be harmful if inexperienced consumers are not protected against fraud or unfair business practices. In addition, effective financial consumer protection frameworks are critical for instilling confidence in the financial system and for promoting financial sector outreach. A high incidence of conflicts with providers of financial services or low levels of satisfaction with financial products used could undermine the trust in the financial system. Despite making existing consumers worse off, it can also discourage potential new consumers to enter the market, which may partially explain why 12 percent of those without a formal account mentioned lack of trust as the main barrier for not having an account (see chapter 1). This section investigates if the products financially included individuals use are effectively meeting their needs. In particular, it identifies segments which are more likely to have encountered a financial service provider conflict in the past three years and who seem less likely to be satisfied with the products they use. In order to measure whether products financially included individuals use are effectively meeting their needs, the financial capability survey captured if survey participants’ were in general satisfied with different types of products they ever used. More specifically, respondents were asked if they are satisfied with products they used from by banks, MFIs, insurance companies, community savings groups, money changers, money lenders, and brokerage houses. Despite serving more customers than other types of providers, banks and their products seem to meet their customers’ needs on a moderate level, similar to those of providers with lowest penetration rates. While 52 percent of the population have been using bank products at some point in their lives (see chapter 1), only 49 percent of their clients indicate to be satisfied with their products. As shown in figure 33, which presents the satisfaction rates for different providers and orders them by their 48 level of penetration; this is a much lower satisfaction rate than the following four providers with the widest outreach achieve. Only insurance companies and brokerage houses whose products reach a small fraction of adults appear to have a similar proportion of customers who claim to be satisfied with the products they offer. A potential explanation for the rather low satisfaction rate banks face may be the lack of competition in the banking sector. This lack of competition does not only result in high costs, commissions and fees (see World Bank 2013), but may also lower incentives to provide appropriate products that satisfy their customers’ needs. Figure 33: Usage and satisfaction rates for different financial providers Source: WB Financial Capability Survey, Mozambique 2013 Delving deeper into the issue of low bank satisfaction rates revealed that financially illiterate respondents are more vulnerable to purchasing and being sold products that do not meet their needs. Those who are struggling more with understanding basic financial concepts also appear to be less satisfied with bank products (see figure 34), even after controlling for other demographic and socioeconomic factors (see table 17). This result suggests that those with lower financial literacy levels are more vulnerable to selecting and being sold inappropriate or even harmful products. Consequently, they require basic measures which help them understand costs and key terms of (bank) products they use and protect them from misleading sale practices. 49 Figure 34: Commercial bank satisfaction rates by financial literacy score Source: WB Financial Capability Survey, Mozambique 2013 Another interesting finding is that consumers of financial services do not widely report complaints or other types of conflicts with providers, neither do they try to solve conflicts they encounter. As shown in figure 35, only 13 percent of the sampled respondents state, that they experienced a conflict with a financial service provider in the past three years. Less than half of those who encountered a dispute (47 percent) took actions to try to solve it. Only 40 percent of those who did not experience a conflict stated that if they faced a conflict they would try to solve it. Figure 35: Approaches to deal with financial service provider conflicts Source: WB Financial Capability Survey, Mozambique 2013 Looking at the characteristics of those who faced a dispute, the survey results suggest that among the groups most vulnerable to having encountered a conflict are those who struggle to understand basic financial concepts. Scoring low on the financial literacy quiz is not only a good predictor for being 50 less satisfied with bank products as compared with those who score higher, but also for being more vulnerable to having encountered a financial service provider conflict. Regression analysis reveal, that holding other demographic and socioeconomic characteristics constant, factors which correlate with a higher probability of having faced a conflict are earning stable incomes, consuming more different types of media regularly, and living in an urban neighborhood. This is due to the fact that these characteristics are also highly correlated with usage of products from a wide range of providers, which increases the probability of having encountered a conflict. Regarding the actions taken to seek redress, redress systems such as the respective regulatory government agency or legal courts were not sought at all by those who experienced a dispute with their financial service provider. In the event that a consumer complaint is not resolved within the financial institution’s own internal procedures, the customer has currently two possible options. One is to turn to the BdM and its team inside of the Department for Strategic Planning, Communication and Image in charge of consumer complaint handing and financial education; the other one is to proceed to court. Figure 36 presents the approaches followed in trying to resolve the conflict. As can be seen, the most common approach was to submit a grievance to the company who sold the product (41 percent), followed by approaching the service provider through friends and family (40 percent). 31 percent of the respondents preferred to stop using the service before the contract expired. Figure 36 further reveals that neither BdM and its team in charge of consumer complaint handling, nor legal courts were sought as systems of redress. That courts were not considered at all can most likely be explained by perceived high costs and lengthy time of proceedings. That consumers do not turn to the BdM may on the other hand be due to the fact that financial services contracts typically do not specify what a consumer should do in the event that he or she has a complaint, and the possibility of recourse to the BdM in the event that the complaint is not resolved to the consumer’s satisfaction following completion of the financial institution’s internal complaints procedures (World Bank 2013c). Figure 36: Actions taken to redress conflicts with financial service providers Source: WB Financial Capability Survey, Mozambique 2013 51 The main reasons for not trying to solve a conflict are either lack of trust in or lack of awareness of respective government authorities which can be approached in the event of a dispute. More than half of those who did not take any actions to solve a dispute stated lack of trust in the effectiveness of government authorities as main reason for their inertia (see figure 37), followed by 48 percent who indicated that they are not aware of any government agencies they can approach for help. Around 40 percent of those who did not try to solve a conflict mentioned that they did not take any actions because they perceive financial institutions as too powerful, while a quarter stated that the law does not adequately protect them. Figure 37: Reasons for not solving conflicts with financial service providers Source: WB Financial Capability Survey, Mozambique 2013 52 References  Atkinson, Adele and Flore-Anne Messy. 2012. “Measuring Financial Literacy: Results of the OECD/International Network on Financial Education (INFE) Pilot Study.” OECD Working Papers on Finance, Insurance, and Privat Pensions, No. 15, OECD Publishing.  Banerjee, Abhijit, Esther Duflo, Rachel Glennerster, and Cynthia Kinnan. 2013. “The miracle of microfinance? Evidence from a randomized evaluation.” National Bureau of Economic Research Working Paper No. 18950.  Banco de Mocambique (BdM). 2013. “Challenges to Financial Inclusion in Mozambique. A Supply-Side Approach.”  Berg, Gunhild and Bilal Zia. 2013. “Financial Literacy through Mainstream Media: Evaluating the Impact of Financial Messages in a South African Soap Opera.” World Bank Working Paper, Washington, DC.  Buehler, R., Griffin, D. and Ross, M. 2002. Inside the planning fallacy: The causes and consequences of optimistic time predictions. Pp. 250-270 in Gilovich, T., Griffin, D. and Kahneman, D. (eds.) Heuristics and Biases: The Psychology of Intuitive Judgment. Cambridge, U.K.: Cambridge University Press.  Demirguc-Kunt, Asli and Leora Klapper, 2012. “Measuring Financial Inclusion: The Global Findex Database.” World Bank Working Paper No. 6025, Washington, DC.  Coville, Aidan, Vincenzo Di Maro, Siegfried Zottel and Felipe Alexander Dunsch. 2013. “The Impact of Financial Literacy through Feature Films: Evidence from a randomized experiment in Nigeria.” Financial Literacy & Education, Russia Trust Fund.  FinMark Trust. 2009. “FinScope 2009 – Mozambique Survey report.”  Gerardi, Kristopher, Lorenz Goette, and Stephan Meier. 2010. “Financial Literacy and SubprimeMortgage Delinquency: Evidence from a Survey Matched to Administrative Data.” Federal Reserve Bank of Atlanta Working Paper Series 2010‐10.  Mozambique Council of Ministers. 2013. “Mozambique’s Financial Sector Development Strategy 2013- 2022.”  Love, Inessa, and Maria Soledad Martinez Peria. 2012. “How Bank Competition Affects Firms’ Access to Finance.” Policy Research Working Paper 6163, World Bank, Washington, DC.  Karlan, Dean, Margaret McConnel, Sendhil Mullainathan, and Jonathan Zinman. 2010. “Getting to the top of mind: How reminders can increase Saving.” National Bureau of Economic Research Working Paper No. 16205.  Klapper, Leora, Anna Maria Lusardi, and Georgios A. Panos. 2012. “Financial Literacy and the Financial Crisis.” World Bank Working Paper No. 5980, Washington, DC.  Yoko Doi, David McKenzie and Bilal Zia. 2012. “Who you train Matters: Identifying Complementary Effects of Financial Education on Migrant Households.” World Bank Working Paper No. WPS6157, Washington, DC. 53  World Bank Group. 2013a. “Global Financial Development Report 2014: Financial Inclusion.” World Bank, Washington, DC.  World Bank Group. 2013c. “Mozambique: Diagnostic Review of Consumer Protection and Financial Literacy. Volume I – Key findings and Recommendations.” World Bank, Washington, DC.  World Bank Group. 2013d. “Financial Capability Surveys Around the World: Why Financial Capability is important and how Surveys can help.” World Bank, Washington, DC. 54 Appendix A. Background on the Mozambique Survey Figure 38: Estimated population break-down by urban/rural Source: WB Financial Capability Survey, Mozambique 2013 Figure 39: Estimated population break-down by different income groups Source: WB Financial Capability Survey, Mozambique 2013 Figure 40: Estimated Population Break-down by Male/Female Source: WB Financial Capability Survey, Mozambique 2013 55 Figure 41: Estimated population break-down by age groups Source: WB Financial Capability Survey, Mozambique 2013 Figure 42: Estimated population break-down by education groups Source: WB Financial Capability Survey, Mozambique 2013 Figure 43: Estimated division of stable/unstable income groups Source: WB Financial Capability Survey, Mozambique 2013 Figure 44: Estimated population break-down by household size Source: WB Financial Capability Survey, Mozambique 2013 56 B. Financial Inclusion Figure 45: Account at a formal financial institution across Sub-Saharan African countries Account at a formal financial institution by gender (% age 15+) Senegal Sudan Congo, Rep. Cameroon Gabon Tanzania Mauritania Uganda Ghana Nigeria Mauritius Zimbabwe Mozambique Kenya South Africa Angola 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Male Female Account at a formal financial institution by income (% age 15+) Senegal Sudan Congo, Rep. Cameroon Gabon Tanzania Mauritania Uganda Ghana Nigeria Mauritius Zimbabwe Mozambique Kenya South Africa Angola 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 By income, top 60% By income, bottom 40% Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex.Demirguc-Kunt and Klapper, 2012 57 Figure 46: Loan from a financial institution in the last year across Sub-Saharan African countries Loan from a financial institution in the past year by gender (% age 15+) Senegal Sudan Congo, Rep. Cameroon Gabon Tanzania Mauritania Uganda Ghana Nigeria Mauritius Zimbabwe Mozambique Kenya South Africa Angola 0.0 5.0 10.0 15.0 20.0 25.0 Male Female Loan from a financial institution in the past year by income (% age 15+) Senegal Sudan Congo, Rep. Cameroon Gabon Tanzania Mauritania Uganda Ghana Nigeria Mauritius Zimbabwe Mozambique Kenya South Africa Angola 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 By income, top 60% By income, bottom 40% Source: Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org/globalfindex.Demirguc-Kunt and Klapper, 2012 58 C. Regression Tables 1. Financial Inclusion Table 3: Probability of knowing about commercial banks on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00951*** 0.00889** 0.00742* 0.00710* (0.00332) (0.00370) (0.00377) (0.00398) Male -0.0306 -0.0992 -0.0862 -0.127 (0.0708) (0.0710) (0.0729) (0.0781) Primary school -0.195** -0.171* -0.135 -0.143 (0.0843) (0.0894) (0.0920) (0.0970) Secondary school -0.296** -0.187 -0.0621 -0.0579 (0.141) (0.146) (0.151) (0.162) Tertiary school -0.527*** -0.433*** -0.290* -0.292* (0.148) (0.157) (0.159) (0.169) Read/write Portuguese 0.157 0.0249 -0.128 -0.252 (0.135) (0.138) (0.142) (0.155) Hhd head 0.198** 0.279*** 0.287*** 0.301*** (0.0798) (0.0877) (0.0892) (0.0954) 2nd income quantile 0.0140 0.00495 0.0278 (0.0806) (0.0826) (0.0896) 3rd income quantile 0.121 0.0571 0.0249 (0.0832) (0.0856) (0.0918) 4th income quantile 0.292*** 0.186* 0.127 (0.105) (0.107) (0.111) Unemployed 0.259 0.277 0.405** (0.164) (0.170) (0.189) Formally employed 0.0739 0.0834 0.000803 (0.120) (0.121) (0.135) Informally employed 0.0436 0.0650 0.118 (0.115) (0.117) (0.125) Self-employed -0.109 -0.101 -0.0495 (0.103) (0.106) (0.112) Retired 0.146 0.0480 0.0989 (0.331) (0.343) (0.391) Urban village 0.620*** 0.527*** (0.0790) (0.0820) One media used 0.458*** (0.146) Two media used 0.641*** (0.152) Three media used 0.681*** (0.163) Four media used 0.774*** (0.205) Five media used 1.192*** (0.302) Six media used 1.254*** (0.401) HH size 0.00690 (0.0223) 1 = if income stable 0.243*** (0.0791) Saved as a Child 0.0707 (0.133) Constant 0.345** 0.327* 0.248 -0.395 (0.146) (0.177) (0.187) (0.282) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 25 F 8.871 5.152 7.594 5.769 59 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Table 4: Probability of having ever used commercial banks on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.0105*** 0.0109*** 0.00978*** 0.00981*** (0.00272) (0.00311) (0.00314) (0.00345) Male 0.0719 0.0595 0.0847 0.00635 (0.0662) (0.0676) (0.0701) (0.0747) Primary school -0.208** -0.180** -0.141 -0.145 (0.0831) (0.0856) (0.0856) (0.0893) Secondary school -0.363*** -0.259* -0.115 -0.0993 (0.129) (0.131) (0.134) (0.145) Tertiary school -0.463*** -0.372** -0.200 -0.103 (0.150) (0.158) (0.158) (0.178) Read/write Portuguese 0.357*** 0.179 -0.00187 -0.213 (0.119) (0.122) (0.122) (0.132) Hhd head 0.0827 0.163* 0.175** 0.244*** (0.0776) (0.0828) (0.0854) (0.0926) 2nd income quantile 0.0278 0.0184 0.114 (0.0775) (0.0782) (0.0774) 3rd income quantile 0.343*** 0.274*** 0.275*** (0.0787) (0.0833) (0.0871) 4th income quantile 0.682*** 0.574*** 0.546*** (0.0987) (0.103) (0.105) Unemployed 0.100 0.110 0.207 (0.143) (0.149) (0.150) Formally employed 0.0622 0.0585 -0.144 (0.114) (0.118) (0.125) Informally employed -0.0469 -0.0302 0.120 (0.113) (0.116) (0.124) Self-employed -0.156 -0.158 -0.0198 (0.0945) (0.0999) (0.105) Retired 0.532* 0.401 0.398 (0.294) (0.317) (0.405) Urban village 0.734*** 0.588*** (0.0753) (0.0781) One media used 0.610*** (0.140) Two media used 0.838*** (0.155) Three media used 0.883*** (0.142) Four media used 1.112*** (0.213) Five media used 1.559*** (0.271) Six media used 1.702*** (0.390) HH size 0.0698*** (0.0209) 1 = if income stable 0.599*** (0.0773) Constant -0.388*** -0.501*** -0.639*** -1.912*** (0.119) (0.161) (0.169) (0.250) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 8.580 10.13 14.68 11.72 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 60 Table 5: Probability of having ever used commercial bank services on village factors (1) Urban location 0.541** (0.220) Peri-urban location 0.296 (0.222) Rural location -0.0982 (0.213) Distance in min to primary school -0.00401* (0.00215) Distance in min to clinic/hospital 0.00212 (0.00163) Distance in min to bank -0.00205 (0.00197) Distance in min to MFI -0.0123*** (0.00195) Most homes have electricity inside property -0.292* (0.149) Most homes have piped water inside property 0.462*** (0.163) Water supply a problem to some extent 0.0675 (0.0953) Water supply is not a problem 0.432*** (0.155) Unemployment a problem 0.175 (0.119) Life in location has not changed from five years ago 0.202** (0.102) Life in location is worse than five years ago 0.125 (0.119) Normal dress standards in location -0.0431 (0.104) Middle income location (perceived) 0.142 (0.135) Low income location (perceived) 0.102 (0.145) Constant 0.477* (0.253) Observations 2,625 df_m 17 F 37.90 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 61 Table 6: Probability of currently having a bank account on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00126 0.00112 -0.000703 -0.00200 (0.00313) (0.00327) (0.00326) (0.00344) Male 0.0735 0.0507 0.0758 0.0184 (0.0653) (0.0694) (0.0723) (0.0802) Primary school -0.218** -0.234** -0.193* -0.184* (0.103) (0.102) (0.104) (0.111) Secondary school -0.414*** -0.346** -0.216 -0.213 (0.135) (0.136) (0.141) (0.155) Tertiary school -0.523*** -0.488*** -0.329** -0.265 (0.158) (0.158) (0.165) (0.183) Read/write Portuguese 0.525*** 0.413*** 0.242* 0.0715 (0.117) (0.119) (0.123) (0.134) Hhd head 0.0458 0.0716 0.0933 0.112 (0.0798) (0.0825) (0.0820) (0.0898) 2nd income quantile 0.00981 0.00422 0.157 (0.0916) (0.0936) (0.0972) 3rd income quantile 0.316*** 0.255** 0.321*** (0.0997) (0.107) (0.107) 4th income quantile 0.675*** 0.573*** 0.570*** (0.107) (0.108) (0.112) Unemployed 0.124 0.120 0.170 (0.148) (0.153) (0.165) Formally employed 0.222** 0.220** -0.0862 (0.0946) (0.0956) (0.107) Informally employed -0.0462 -0.0435 0.107 (0.100) (0.101) (0.110) Self-employed -0.0192 -0.0210 0.116 (0.0890) (0.0924) (0.100) Retired 0.607** 0.511* 0.381 (0.245) (0.266) (0.279) Urban village 0.714*** 0.564*** (0.0711) (0.0797) One media used -0.142 (0.155) Two media used -0.0544 (0.146) Three media used -0.0521 (0.148) Four media used 0.312 (0.203) Five media used 0.463** (0.225) Six media used 1.319*** (0.340) HH size 0.0397* (0.0201) 1 = if income is stable 0.675*** (0.0802) Constant -0.742*** -0.925*** -1.070*** -1.413*** (0.152) (0.182) (0.184) (0.230) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 4.633 7.872 12.57 11.25 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 62 Table 7: Probability of currently having a bank loan on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00639* 0.00474 0.00356 0.00293 (0.00344) (0.00382) (0.00388) (0.00404) Male 0.0735 0.0363 0.0518 0.0465 (0.0865) (0.0925) (0.0936) (0.0983) Primary school -0.110 -0.108 -0.0852 -0.0840 (0.106) (0.118) (0.120) (0.123) Secondary school -0.340** -0.332** -0.259 -0.304* (0.148) (0.162) (0.168) (0.172) Tertiary school -0.606*** -0.600*** -0.528*** -0.526*** (0.168) (0.185) (0.185) (0.196) Read/write Portuguese 0.378*** 0.316** 0.206 0.0932 (0.130) (0.141) (0.148) (0.148) Hhd head 0.120 0.218** 0.230** 0.202* (0.0931) (0.103) (0.103) (0.115) 2nd income quantile 0.0378 0.0426 0.133 (0.116) (0.120) (0.117) 3rd income quantile 0.270** 0.233* 0.266** (0.128) (0.135) (0.133) 4th income quantile 0.317*** 0.235** 0.207* (0.112) (0.115) (0.115) Unemployed -0.0452 -0.0450 -0.0880 (0.176) (0.180) (0.207) Formally employed -0.154 -0.172 -0.339*** (0.122) (0.124) (0.122) Informally employed -0.130 -0.133 -0.0552 (0.123) (0.125) (0.135) Self-employed -0.212* -0.224** -0.163 (0.108) (0.110) (0.120) Retired 0.457* 0.398 0.388 (0.265) (0.276) (0.299) Urban village 0.426*** 0.328*** (0.0944) (0.107) One media used -0.0388 (0.157) Two media used 0.0899 (0.160) Three media used -0.0164 (0.168) Four media used 0.326 (0.243) Five media used 0.498** (0.221) Six media used 0.530** (0.264) HH size 0.0134 (0.0203) 1 = if income is stable 0.437*** (0.103) Constant -1.652*** -1.568*** -1.642*** -1.855*** (0.161) (0.220) (0.220) (0.294) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 6.305 3.929 5.677 4.807 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 63 Table 8: Probability of having ever used insurance services on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00468 0.000921 -7.49e-05 -0.000571 (0.00355) (0.00406) (0.00404) (0.00381) Male 0.00603 0.0233 0.0348 -0.0483 (0.0826) (0.0832) (0.0876) (0.0940) Primary school -0.297*** -0.343*** -0.310*** -0.326*** (0.100) (0.107) (0.110) (0.112) Secondary school -0.263* -0.335** -0.246 -0.160 (0.145) (0.158) (0.162) (0.172) Tertiary school -0.540*** -0.689*** -0.587*** -0.447** (0.178) (0.189) (0.195) (0.207) Read/write Portuguese 0.635*** 0.573*** 0.454*** 0.185 (0.118) (0.130) (0.134) (0.143) Hhd head -0.0447 0.0439 0.0571 0.0925 (0.0929) (0.105) (0.108) (0.106) 2nd income quantile -0.586*** -0.594*** -0.261** (0.104) (0.106) (0.107) 3rd income quantile -0.130 -0.185* -0.0923 (0.0937) (0.100) (0.102) 4th income quantile 0.275*** 0.189** 0.240** (0.0889) (0.0926) (0.0971) Unemployed -0.168 -0.176 -0.0616 (0.153) (0.162) (0.164) Formally employed 0.207* 0.203 -0.220 (0.119) (0.124) (0.140) Informally employed -0.516*** -0.532*** -0.355** (0.124) (0.127) (0.147) Self-employed -0.377*** -0.397*** -0.178 (0.105) (0.110) (0.122) Retired 0.139 0.0309 -0.212 (0.284) (0.279) (0.265) Urban village 0.505*** 0.271*** (0.0744) (0.0852) One media used 0.379 (0.244) Two media used 0.585** (0.238) Three media used 0.757*** (0.229) Four media used 0.970*** (0.278) Five media used 1.115*** (0.253) Six media used 0.567* (0.337) HH size 0.0687*** (0.0180) 1 = if income is stable 1.355*** (0.0710) Constant -1.215*** -0.674*** -0.769*** -2.279*** (0.161) (0.202) (0.201) (0.340) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 9.790 13.24 15.76 25.96 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 64 Table 9: Probability of having ever used MFI services on demographic and socioeconomic factors (1) (2) (3) (4) Age -0.000990 -0.00202 -0.00277 -0.00288 (0.00291) (0.00322) (0.00317) (0.00324) Male 0.132* 0.110 0.119 0.0690 (0.0700) (0.0717) (0.0731) (0.0686) Primary school -0.100 -0.0216 -0.000286 0.0321 (0.0834) (0.0908) (0.0901) (0.0946) Secondary school -0.0175 0.0916 0.154 0.235 (0.126) (0.132) (0.134) (0.151) Tertiary school -0.294** -0.244 -0.170 -0.0232 (0.146) (0.159) (0.160) (0.180) Read/write Portuguese 0.113 -0.0161 -0.0960 -0.294** (0.113) (0.116) (0.117) (0.129) Hhd head -0.0900 0.00734 0.0140 0.0790 (0.0787) (0.0865) (0.0882) (0.0847) 2nd income quantile -0.190** -0.196** -0.0763 (0.0804) (0.0800) (0.0865) 3rd income quantile 0.204** 0.171** 0.187** (0.0815) (0.0839) (0.0845) 4th income quantile 0.282*** 0.223** 0.227** (0.0995) (0.104) (0.103) Unemployed -0.229 -0.233 -0.113 (0.164) (0.167) (0.163) Formally employed -0.109 -0.114 -0.391*** (0.131) (0.136) (0.135) Informally employed -0.228** -0.224** -0.0576 (0.111) (0.113) (0.112) Self-employed -0.184* -0.184* -0.0383 (0.0996) (0.101) (0.106) Retired 0.0251 -0.0317 -0.0946 (0.265) (0.264) (0.263) Urban village 0.329*** 0.177** (0.0728) (0.0680) One media used 0.319** (0.145) Two media used 0.478*** (0.153) Three media used 0.693*** (0.159) Four media used 0.691*** (0.188) Five media used 0.645*** (0.203) Six media used -0.0427 (0.277) HH size 0.0297* (0.0179) 1 = if income is stable 0.820*** (0.0672) Constant -0.323** -0.197 -0.252 -1.115*** (0.132) (0.176) (0.176) (0.262) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 2.827 3.519 4.354 10.08 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 65 Table 10: Probability of having ever used money changers on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00814** 0.00631* 0.00550 0.00642 (0.00344) (0.00373) (0.00373) (0.00393) Male 0.0252 0.0216 0.0348 0.00163 (0.0763) (0.0805) (0.0843) (0.0813) Primary school 0.0369 0.0451 0.0736 0.129 (0.0984) (0.101) (0.103) (0.108) Secondary school 0.0157 0.0579 0.138 0.251* (0.128) (0.129) (0.131) (0.135) Tertiary school -0.219 -0.235 -0.143 -0.0186 (0.145) (0.152) (0.151) (0.156) Read/write Portuguese 0.385*** 0.296** 0.194* -0.0713 (0.112) (0.114) (0.116) (0.123) Hhd head -0.0166 0.0470 0.0578 0.0789 (0.0914) (0.0984) (0.103) (0.0981) 2nd income quantile -0.153* -0.159* 0.0205 (0.0893) (0.0922) (0.0893) 3rd income quantile 0.0515 0.00525 0.0430 (0.0969) (0.103) (0.103) 4th income quantile 0.353*** 0.272*** 0.274** (0.101) (0.103) (0.107) Unemployed 0.0272 0.0316 0.115 (0.164) (0.170) (0.172) Formally employed -0.0118 -0.0117 -0.359*** (0.116) (0.119) (0.126) Informally employed -0.213* -0.211* -0.0709 (0.117) (0.120) (0.127) Self-employed -0.129 -0.129 0.0130 (0.0978) (0.101) (0.107) Retired 0.228 0.159 0.128 (0.252) (0.257) (0.294) Urban village 0.418*** 0.226** (0.0790) (0.0900) One media used 0.544*** (0.198) Two media used 0.621*** (0.191) Three media used 0.714*** (0.197) Four media used 1.108*** (0.231) Five media used 0.936*** (0.235) Six media used 1.597*** (0.300) HH size 0.0149 (0.0194) 1 = if income is stable 0.757*** (0.0858) Constant -1.328*** -1.186*** -1.271*** -2.252*** (0.162) (0.202) (0.203) (0.334) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 4.650 3.743 6.470 11.38 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 66 Table 11: Probability of having ever used money lenders on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.00471* 0.00439 0.00377 0.00422 (0.00284) (0.00302) (0.00300) (0.00310) Male 0.00823 0.00116 0.00868 -0.0619 (0.0692) (0.0726) (0.0731) (0.0684) Primary school -0.271*** -0.245** -0.225** -0.230** (0.0905) (0.100) (0.0999) (0.108) Secondary school -0.309** -0.272** -0.211 -0.196 (0.119) (0.131) (0.130) (0.140) Tertiary school -0.640*** -0.633*** -0.564*** -0.524*** (0.151) (0.160) (0.155) (0.171) Read/write Portuguese 0.204* 0.143 0.0645 -0.108 (0.106) (0.110) (0.107) (0.120) Hhd head 0.0653 0.103 0.109 0.114 (0.0793) (0.0866) (0.0873) (0.0888) 2nd income quantile -0.272*** -0.277*** -0.124 (0.0717) (0.0714) (0.0766) 3rd income quantile -0.0558 -0.0895 -0.0773 (0.0805) (0.0844) (0.0867) 4th income quantile 0.0220 -0.0375 -0.109 (0.0868) (0.0914) (0.0941) Unemployed -0.0527 -0.0518 0.0854 (0.165) (0.168) (0.178) Formally employed 0.117 0.116 -0.252* (0.128) (0.131) (0.133) Informally employed -0.217** -0.212* -0.0402 (0.109) (0.110) (0.114) Self-employed -0.194** -0.193* -0.00959 (0.0974) (0.0982) (0.0959) Retired -0.313 -0.370 -0.483* (0.240) (0.241) (0.247) Urban village 0.302*** 0.101 (0.0706) (0.0726) One media used 0.506*** (0.146) Two media used 0.564*** (0.144) Three media used 0.565*** (0.151) Four media used 0.748*** (0.196) Five media used 1.057*** (0.212) Six media used 1.494*** (0.299) HH size 0.0255 (0.0204) 1 = if income stable 0.993*** (0.0673) Constant -0.311** -0.0740 -0.123 -1.063*** (0.129) (0.166) (0.165) (0.223) Observations 3,000 2,758 2,758 2,572 df_m 7 15 16 24 F 5.280 5.493 5.991 15.09 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 67 Table 12: Probability of having ever had a formal account on demographic and socioeconomic factors (1) (2) (3) (4) Age 0.0130*** 0.0138*** 0.0130*** 0.0136*** (0.00276) (0.00337) (0.00344) (0.00376) Male 0.0716 0.0831 0.108 0.0300 (0.0617) (0.0656) (0.0681) (0.0731) Primary school -0.210** -0.194** -0.156* -0.155* (0.0862) (0.0904) (0.0907) (0.0934) Secondary school -0.326** -0.233* -0.0911 -0.0760 (0.129) (0.136) (0.138) (0.154) Tertiary school -0.480*** -0.414*** -0.251 -0.139 (0.150) (0.158) (0.158) (0.185) Read/write Portuguese 0.356*** 0.195 0.0174 -0.189 (0.118) (0.121) (0.122) (0.138) Hhd head 0.0576 0.0995 0.110 0.181** (0.0735) (0.0781) (0.0805) (0.0888) 2nd income quantile 0.0145 0.00581 0.0954 (0.0761) (0.0759) (0.0789) 3rd income quantile 0.318*** 0.253*** 0.253*** (0.0792) (0.0820) (0.0865) 4th income quantile 0.551*** 0.448*** 0.379*** (0.0972) (0.100) (0.111) Unemployed 0.202 0.212 0.329** (0.145) (0.150) (0.154) Formally employed 0.177 0.173 -0.0474 (0.108) (0.110) (0.113) Informally employed 0.00200 0.0163 0.182 (0.108) (0.107) (0.118) Self-employed -0.103 -0.107 0.0452 (0.0883) (0.0908) (0.0975) Retired 0.371 0.228 0.314 (0.278) (0.293) (0.368) Urban village 0.717*** 0.571*** (0.0733) (0.0770) One media used 0.564*** (0.144) Two media used 0.813*** (0.151) Three media used 0.890*** (0.145) Four media used 1.163*** (0.204) Five media used 1.284*** (0.219) Six media used 0.0639*** (0.0185) HH size 0.646*** (0.0801) 1 = if income stable -0.513*** -0.661*** -0.806*** -2.062*** (0.123) (0.166) (0.175) (0.256) Constant 3,000 2,758 2,758 2,525 7 15 16 23 Observations 10.03 8.656 14.25 11.23 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 68 2. Financial Capability Table 13: Probability of financial literacy and financial product knowledge scores on village factors (1) (2) Urban location 0.00816 -0.0308 (0.0520) (0.0934) Peri-urban location -0.0314 0.118 (0.0977) (0.107) Rural location -0.0350 -0.173* (0.0546) (0.0961) Distance in min to primary school 0.000841 -0.00126 (0.000597) (0.00110) Distance in min to clinic/hospital -0.00106*** 0.00111 (0.000387) (0.000924) Distance in min to bank -0.00111** -5.62e-05 (0.000515) (0.00105) Distance in min to MFI 0.00173*** -0.00725*** (0.000515) (0.00109) Most homes do not have electricity inside 0.0337 -0.0427 property (0.0392) (0.0593) Most homes do not have piped water inside -0.0247 0.124 property (0.0457) (0.0932) Water supply is a problem to some extent 0.0305 -0.0165 (0.0205) (0.0459) Water supply is not a problem -0.0220 0.0185 (0.0397) (0.0790) Unemployment is a problem to some extent -0.0155 0.0595 (0.0289) (0.0629) Crime is a problem to some extent -0.114*** 0.0616 (0.0264) (0.0510) Crime is not a problem 0.0461 0.197*** (0.0321) (0.0715) Life in location has not changed from 5 years -0.0454* 0.0648 ago (0.0261) (0.0581) Life in location is worse than 5 years ago -0.0366 0.0494 (0.0426) (0.0678) Normal dress standards in location 0.00368 -0.0438 (0.0234) (0.0471) Middle income location (perceived) 0.0105 -0.265*** (0.0333) (0.0761) Low income location (perceived) 0.0510 -0.280*** (0.0385) (0.0786) Constant 1.336*** 1.829*** (0.0512) (0.138) Observations 2,534 2,625 df_m 19 19 F 6.031 32.65 Estimates of the poisson model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 69 Table 14: Probability of financial literacy score on demographic and socioeconomic factors (1) (2) (3) (4) age 0.00167 0.00221* 0.00213* 0.00215* (0.00106) (0.00119) (0.00117) (0.00117) male 0.0175 0.0171 0.0179 0.0176 (0.0193) (0.0200) (0.0200) (0.0201) Primary school 0.0433 0.0199 0.0221 0.0185 (0.0288) (0.0328) (0.0327) (0.0327) Secondary school 0.141*** 0.133*** 0.139*** 0.127*** (0.0409) (0.0423) (0.0423) (0.0421) Tertiary school 0.252*** 0.239*** 0.246*** 0.236*** (0.0448) (0.0477) (0.0480) (0.0470) Read/write Portuguese -0.0312 -0.0155 -0.0231 -0.0138 (0.0380) (0.0367) (0.0362) (0.0354) Hhd head 0.00224 -0.0132 -0.0122 -0.0145 (0.0271) (0.0295) (0.0294) (0.0298) 2nd income quantile 0.0261 0.0242 0.0228 (0.0250) (0.0249) (0.0249) 3rd income quantile 0.0340 0.0296 0.0287 (0.0271) (0.0267) (0.0266) 4th income quantile 0.0314 0.0243 0.0169 (0.0287) (0.0278) (0.0287) Unemployed 0.0175 0.0165 0.0169 (0.0410) (0.0412) (0.0419) Formally employed 0.104*** 0.106*** 0.102*** (0.0346) (0.0346) (0.0351) Informally employed -0.0242 -0.0246 -0.0211 (0.0319) (0.0318) (0.0318) Self-employed 0.00386 0.00315 0.00447 (0.0258) (0.0259) (0.0262) Retired -0.105 -0.108 -0.106 (0.0982) (0.0995) (0.0993) 1 = if income is stable -0.131*** -0.139*** -0.124*** (0.0264) (0.0269) (0.0277) Urban village 0.0364* 0.0374* (0.0212) (0.0218) Saved as a child 0.0563* (0.0307) One media used 0.0566 (0.0398) Two media used 0.0158 (0.0359) Three media used 0.0332 (0.0386) Four media used 0.0410 (0.0553) Five media used 0.0699 (0.0504) Six media used 0.139* (0.0767) Constant 1.172*** 1.159*** 1.156*** 1.070*** (0.0484) (0.0642) (0.0643) (0.0781) Observations 2,898 2,488 2,488 2,488 df_m 7 16 17 24 F 11.57 5.836 5.501 4.438 Estimates of the poisson model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 70 Table 15: Probability of financial knowledge score on demographic and socioeconomic factors (1) (2) (3) (4) age 0.00542*** 0.00439*** 0.00385*** 0.00403*** (0.00132) (0.00128) (0.00125) (0.00123) male 0.0168 -0.00895 -0.00391 -0.00657 (0.0346) (0.0302) (0.0309) (0.0304) Primary school -0.0919** -0.0747* -0.0611 -0.0569 (0.0396) (0.0379) (0.0379) (0.0380) Secondary school -0.116* -0.0176 0.0151 0.0170 (0.0598) (0.0594) (0.0596) (0.0599) Tertiary school -0.246*** -0.184*** -0.143** -0.136* (0.0715) (0.0700) (0.0681) (0.0692) Read/write Portuguese 0.229*** 0.0669 0.0218 -0.0441 (0.0587) (0.0589) (0.0591) (0.0610) Hhd head 0.0351 0.0766** 0.0835** 0.0837*** (0.0364) (0.0312) (0.0325) (0.0316) 2nd income quantile -0.00976 -0.0180 -0.0138 (0.0358) (0.0362) (0.0370) 3rd income quantile 0.0792** 0.0554 0.0467 (0.0359) (0.0370) (0.0358) 4th income quantile 0.177*** 0.137*** 0.115*** (0.0370) (0.0377) (0.0369) Unemployed 0.0548 0.0488 0.0509 (0.0554) (0.0584) (0.0586) Formally employed -0.144*** -0.125*** -0.121** (0.0486) (0.0476) (0.0471) Informally employed -0.0218 -0.0265 -0.0198 (0.0487) (0.0486) (0.0476) Self-employed -0.0760* -0.0828* -0.0838** (0.0427) (0.0432) (0.0423) Retired -0.110 -0.131* -0.0784 (0.0801) (0.0750) (0.0723) 1 = if income is stable 0.440*** 0.400*** 0.358*** (0.0270) (0.0255) (0.0257) Urban village 0.200*** 0.157*** (0.0287) (0.0270) Saved as a child -0.0411 (0.0368) One media used 0.280*** (0.0943) Two media used 0.365*** (0.0939) Three media used 0.427*** (0.0986) Four media used 0.498*** (0.0982) Five media used 0.522*** (0.103) Six media used 0.682*** (0.104) Constant 0.915*** 0.854*** 0.839*** 0.579*** (0.0645) (0.0747) (0.0754) (0.126) Observations 3,000 2,572 2,572 2,572 df_m 7 16 17 24 F 9.109 25.66 25.81 23.18 Estimates of the poisson model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 71 Table 16: Capability of covering unexpected expenses on demographic and socioeconomic factors (1) (2) (3) (4) age 0.117* 0.135* 0.114 0.0853 (0.0646) (0.0736) (0.0739) (0.0759) male 1.255 1.099 1.293 1.498 (1.500) (1.656) (1.638) (1.718) Primary school -1.767 -0.793 -0.306 -0.857 (2.200) (2.223) (2.210) (2.356) Secondary school 0.0463 1.093 2.432 1.764 (3.342) (3.604) (3.582) (3.638) Tertiary school 0.636 1.645 3.270 2.141 (3.977) (4.253) (4.252) (4.329) Read/write Portuguese 1.882 0.868 -0.815 -1.425 (2.841) (3.057) (3.076) (3.123) Hhd head 2.465 2.188 2.408 2.785 (1.632) (1.939) (1.902) (1.987) 2nd income quantile 6.330*** 5.930*** 5.530*** (1.850) (1.861) (1.911) 3rd income quantile 7.044*** 6.046*** 5.640*** (2.081) (2.076) (2.104) 4th income quantile 4.503* 2.896 2.221 (2.325) (2.398) (2.448) Unemployed 1.730 1.478 0.925 (3.591) (3.589) (3.615) Formally employed -1.176 -0.538 -1.253 (2.813) (2.836) (2.923) Informally employed -0.115 -0.239 -1.155 (2.476) (2.501) (2.587) Self-employed 3.199 3.003 2.537 (2.192) (2.245) (2.357) Retired 6.275 5.391 7.136 (6.613) (6.323) (6.564) 1 = if income is stable 5.586*** 3.956* 4.230* (2.082) (2.039) (2.232) Urban village 8.236*** 7.424*** (1.500) (1.596) Financial literacy score -0.189 (0.473) Saved as a child 3.939 (3.045) One media used 2.488 (3.246) Two media used 5.659* (3.108) Three media used 2.645 (3.521) Four media used 7.455 (4.906) Five media used 9.829* (5.245) Six media used 11.07 (7.528) Constant 54.43*** 47.08*** 46.50*** 42.31*** (3.148) (3.903) (3.891) (5.417) Observations 2,996 2,568 2,568 2,484 df_m 7 16 17 25 F 2.407 3.055 5.494 4.412 Estimates of the regression model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 72 Table 17: Satisfaction rate on commercial banks on demographic and socioeconomic factors (1) age 0.00239 (0.00399) male 0.0751 (0.0996) Primary school -0.0300 (0.132) Secondary school -0.0134 (0.195) Tertiary school 0.0658 (0.222) Read/write Portuguese 0.153 (0.181) Hhd head 0.0160 (0.113) 2nd income quantile 0.152 (0.125) 3rd income quantile 0.127 (0.115) 4th income quantile 0.0760 (0.124) Unemployed 0.217 (0.183) Formally employed 0.0618 (0.166) Informally employed -0.0456 (0.148) Self-employed 0.184 (0.138) Retired 0.779** (0.348) 1 = if income is stable -0.212** (0.106) Urban village 0.0231 (0.0815) Financial literacy score 0.0608** (0.0289) Saved as a child -0.243** (0.113) One media used 0.141 (0.259) Two media used -0.0413 (0.266) Three media used -0.0870 (0.282) Four media used 0.350 (0.300) Five media used -0.000557 (0.305) Six media used 0.679* (0.353) Constant -0.414 (0.365) Observations 1,458 df_m 25 F 3.092 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 73 Table 18: Probability of using financial instruments on demographic and socioeconomic factors (1) (2) (3) (4) Insurance products Bank credit Bank savings Money transfers Financial literacy index -0.0109 -0.00364 0.0162 0.0280 (0.0252) (0.0332) (0.0288) (0.0269) Awareness of financial products index 0.271*** 0.244*** 0.262*** 0.365*** (0.0236) (0.0329) (0.0233) (0.0209) Budgeting 0.000895 -0.00169 -0.00133 -0.00330*** (0.00142) (0.00139) (0.00146) (0.00119) Living within means -0.000381 -7.82e-05 0.00144 -0.000749 (0.00134) (0.00137) (0.00135) (0.00127) Not over-spending 0.000481 0.000529 0.00170* 0.00121 (0.000995) (0.00115) (0.000919) (0.000941) Monitor expenses -0.00254** 0.00440** 0.00335** -0.00123 (0.00128) (0.00191) (0.00154) (0.00139) Attitudes towards info -0.00257 -0.000764 -0.00195 0.00144 (0.00180) (0.00189) (0.00199) (0.00151) Planning unexpected -0.000400 -0.000883 0.00160 0.000757 (0.00120) (0.00124) (0.00132) (0.00104) Saving for the future 6.20e-05 0.000580 0.000718 0.00133 (0.00118) (0.00144) (0.00130) (0.00130) Far sightedness 0.000612 0.00354* 0.00176 -0.00582*** (0.00167) (0.00201) (0.00194) (0.00168) Provisions for old age -0.000893 -0.000626 0.00176 0.00199** (0.00113) (0.00146) (0.00112) (0.00100) Age -0.00117 -0.00231 -0.00325 0.0166*** (0.00404) (0.00508) (0.00491) (0.00438) Male -0.120 0.0388 0.101 -0.00284 (0.0930) (0.111) (0.0975) (0.0884) Primary school 0.205* 0.0908 -0.213 0.128 (0.124) (0.146) (0.132) (0.103) Secondary school 0.248 0.00907 -0.225 -0.0795 (0.172) (0.186) (0.194) (0.159) Tertiary school 0.311 -0.264 -0.100 -0.0841 (0.189) (0.212) (0.232) (0.195) Read/write Portuguese -0.270* -0.113 0.181 -0.207 (0.140) (0.153) (0.150) (0.151) Hhd head 0.0627 0.299** -0.176 0.198** (0.109) (0.129) (0.114) (0.0938) (0.108) (0.123) (0.129) (0.112) Constant -1.603*** -2.780*** -2.837*** -4.194*** (0.355) (0.430) (0.366) (0.513) Observations 2,482 2,482 2,474 2,482 df_m 35 35 34 35 F 7.576 5.159 11.30 16.48 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 74 Table 19: Probability of using financial instruments on financial capability scores (5) (6) (7) (8) Credit from MFI Community Informal credit Informal savings based savings Financial literacy index 0.0523* -0.0293 -0.0261 -0.0352 (0.0289) (0.0271) (0.0302) (0.0220) Awareness of financial products index 0.248*** 0.290*** 0.196*** 0.361*** (0.0215) (0.0184) (0.0208) (0.0172) Budgeting 0.000870 -0.00140 -0.00118 0.00155 (0.00151) (0.00142) (0.00113) (0.00103) Living within means -0.00416** 0.000319 -0.00439*** -0.000225 (0.00165) (0.00117) (0.00117) (0.00117) Not over-spending 0.00193* 0.00163** -0.00200** 0.00181** (0.00105) (0.000812) (0.000932) (0.000723) Monitor expenses -6.64e-05 -0.00100 -0.000301 0.00100 (0.00161) (0.00133) (0.00138) (0.00116) Attitudes towards info -0.00309* -0.00121 0.000471 -0.000245 (0.00187) (0.00175) (0.00153) (0.00137) Planning unexpected 0.000218 0.000770 -0.000919 7.80e-05 (0.00136) (0.00118) (0.00134) (0.00110) Saving for the future 0.000947 -0.000912 0.00219* -0.000889 (0.00176) (0.00120) (0.00119) (0.00112) Far sightedness -0.000345 -0.00163 0.00410** -0.00182 (0.00190) (0.00182) (0.00166) (0.00156) Provisions for old age 0.000987 0.000685 -0.00483*** -0.00208** (0.00122) (0.00102) (0.00116) (0.000805) Age -0.00845 -0.00989** -0.00486 0.00118 (0.00567) (0.00395) (0.00409) (0.00371) Male -0.128 -0.00643 -0.226** -0.162** (0.117) (0.0848) (0.0892) (0.0765) Primary school -0.0513 -0.0534 -0.252** -0.0605 (0.131) (0.103) (0.119) (0.0968) Secondary school 0.0165 0.106 -0.210 0.0768 (0.195) (0.172) (0.178) (0.143) Tertiary school -0.189 -0.0364 -0.430** -0.144 (0.232) (0.196) (0.205) (0.162) Read/write Portuguese -0.187 -0.466*** -0.116 -0.128 (0.173) (0.163) (0.142) (0.124) Hhd head -0.0317 0.0244 0.0402 0.0748 (0.131) (0.104) (0.105) (0.0834) (0.113) (0.119) (0.113) (0.0964) Constant -1.849*** -1.404*** -0.717** -0.839*** (0.481) (0.426) (0.355) (0.312) Observations 2,425 2,482 2,482 2,482 df_m 33 35 35 35 F 10.41 12.10 13.42 17.22 Estimates of the probit model. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 75