June 2012 66390 Measuring Financial Capability and the Effectiveness of Financial Education OVERVIEW OF THE WORLD BANK TRUST FUND SUPPORTED BY THE RUSSIAN FEDERATION Measuring Financial Capability and the Effectiveness of Financial Education WORLD BANK TRUST FUND SUPPORTED BY THE RUSSIAN FEDERATION | FINANCIAL LITERACY & EDUCATION CONTENTS The World Bank Financial Literacy and Education Trust Fund Financial Capability Measurement Armenia Colombia Lebanon Malawi Mexico Namibia Papua New Guinea Tanzania iii Turkey Uruguay Zambia Monitoring and Evaluation Australia and New Zealand Brazil Dominican Republic India India and Kenya Kenya Malawi Mexico Nigeria South Africa Uganda Evaluation Toolkit for Financial Capability Programs in Low and Middle Income Countries vi The World Bank Financial Literacy and Education Trust Fund Motivations for the Trust Fund The increased role and complexity of �nancial markets and products has reinforced the need to improve the capacity of consumers to manage risks and effectively engage with the requirements of market economies, new �nancial products, emerging forms of social insurance programs and the associated need for individual life cycle planning, especially in low and middle income countries where �nancial inclusion is poor and educational attainments are low. To enable consumers 1 in these countries to understand, access, and make better �nancial decisions requires interventions to improve knowledge, skills, attitude and behavior. Despite the nearly universal nature of the need to develop programs aimed at improving these capabilities, there is not yet a consensus about how to best de�ne the knowledge and skills required to be �nancially capable, how to measure the level of capabilities among different groups and how to evaluate the effectiveness of programs designed to enhance these skills and capabilities. What the Trust Fund is Doing The primary focus of the Trust Fund is to assist with the development of a conceptual framework and operational toolkits for (1) measuring levels of �nancial capability and (2) evaluating the impact of �nancial capability enhancement programs. Financial support, intellectual guidance and technical assistance will be provided through the Trust Fund to develop innovative methods and test their application through support for country level programs. 2 Financial Capability Measurement Measuring Financial Capability in Low and Middle Income Countries Policy makers across the world are implementing �nancial education programs with the objective of improving people’s ability to make informed and effective �nancial decisions. This is particularly important given the increasing individual responsibility for life-long income planning and risk management, and the large number and extreme complexity of the products currently available in �nancial markets. Financial literacy is considered to be a crucial factor enabling The Trust Fund will provide oper- individuals to make good �nancial decisions and, as such, the ational guidance to design and 5 implement a survey instrument main target for �nancial education interventions. Nonetheless, to that can be used to assess the level date there is no clear consensus on what constitutes �nancial of �nancial capability in a popula- literacy. The term itself has been used in reference to different tion, to understand what areas concepts. Most of the research literature has used the term to of �nancial capability need to be indicate knowledge of �nancial concepts such as compound enhanced through interventions, and which segments of the popu- interest, inflation and risk, or awareness of the existence of lation are less �nancially capable. �nancial products or institutions. Researchers and practitioners, however, have come to understand that knowledge may not be suf�cient to determine �nancial well-being. One of the main reasons for this is that many factors may affect how people use or fail to use the knowledge they have acquired. Improving access to �nance and strengthening consumer protection mechanisms are two ways in which positive changes in the environment can enable individuals to make better �nancial decisions. At the same time, however, people also require certain skills, motivations, and attitudes to translate knowledge into behavior that generates positive �nancial outcomes. The Financial Literacy and Education Trust Fund supported by the Russian Federation, is undertaking a comprehensive, empirically rigorous effort to understand how this broad concept of �nancial capability can be measured in low and middle income countries. The approach adopted by the World Bank in implementing this work is based on the idea that �nancial capability cannot be de�ned a priori in the same way as it is conceptualized in developed countries, because individuals in these environments may require very different knowledge, skills, or attitudes to manage their resources well and to be considered �nancially capable. In order to understand these differences, the World Bank has Several countries are partici- supported and coordinated extensive qualitative research with pating in the program to help the general public in eight countries across the world. Through develop and test the survey these guided focus group conversations the research teams instrument through qualitative and expert group identi�ed a number of concepts that were research and national surveys: consistently mentioned across all countries as being elemental Armenia, Colombia, Lebanon, Malawi, Mexico, Namibia, Papua to any de�nition of �nancial capability. In order to measure these New Guinea, Tanzania, Turkey, concepts, questions from existing surveys were selected, where Uruguay, Zambia. 6 available, and new questions were designed with alternative formulations. In each country, there were then two rounds of cognitive interviews conducted in order to probe and test the questions to make sure that they were well understood, relevant, and informative. The World Bank has selected those questions that proved to work best across countries and drafted a pilot survey questionnaire that is now being implemented in about 10 countries, consisting of the original countries that participated in the qualitative study, plus three additional countries. The pilot surveys will be concluded by mid-2012. The data that is collected through this process will be used to explore the nuanced relationships between behavior, skills, knowledge, attitudes and environment in the area of �nancial capability. Ultimately, the process will serve to develop simple means of synthesizing this information into a measure of individual �nancial capability comprised of these different components. The �nal output of the program will be a set of diagnostic tools for policy makers to identify areas of �nancial capability and segments of their populations requiring particular �nancial capability interventions. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT ARMENIA Project Objectives The project is aimed at establishing a nationally representative baseline of �nancial capability in Armenia using the conceptual framework and the methodology developed by the Trust Fund. The project will provide the Government of Armenia with the information required to design �nancial education programs that will address the most important needs of different segments of the population. Such a baseline will also be important for allowing the government to be able to assess the effectiveness of certain �nancial education programs later on and Project Name decide which initiatives should be continued and perhaps scaled up and which should be modi�ed or discontinued. Armenia Financial Capability Survey Survey Implementation Partner Institution(s) The survey has been carried out as a standalone project and adhered Central Bank of Armenia; National to the Trust Fund guidelines. Correspondingly, the survey and the Statistical Service of the Republic of analysis focuses on �nancial capability as a broad concept, and Armenia; AM Partners Consulting uses the survey questionnaire developed by the Trust Fund. The 7 Company questionnaire has been translated and adapted into the Armenian language and context. Complementary questions from the survey Timeline developed by the World Bank’s Global Program on Consumer Protection and Financial Literacy have been added, following the August 2011 – June 2012 Trust Fund guidelines. The questionnaire (in Armenian) was piloted among 124 households living in urban and rural areas. Based on the results of the pilot tests conducted in Armenia and other countries, the Trust Fund made �nal improvements to the questionnaire. The revised write-up and questionnaire (in English) were reviewed and approved by the Trust Fund before the surveyors went to the �eld. The survey �eldwork was implemented by AM Partners in the capital (Yerevan) and ten regions of Armenia, from 10 April to 21 May 2012, using the survey methodology prescribed by the Trust Fund. The sample was nationally representative of the adult (18 years and older) population living in households. Overall, 2,000 valid interviews were completed during �eldwork from a total of 4,647 contacted households. A multi-stage strati�ed sampling method was used (strati�cation by urban and rural areas, by regions or Marzes). The unit of analysis was the individual (randomly chosen, using a Kish table). Overall, the team closely adhered to the instructions provided by the Trust Fund team to ensure cross-country comparability of the results and the achievement of a common framework that is suitable to be adopted in the widest possible range of countries. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Main Outputs and Dissemination The project will produce the following outputs as deliverables: Policy Impact • Inception report on proposed methodology for survey implementation; The Central Bank of Armenia (CBA) has dedicated substantial resources • Pilot test report including overall and module length (time), comments from interviewers and supervisors, problems encountered in different phases of to consumer protection and �nan- pilot, and solutions proposed/recommended changes to the questionnaire; cial education programs. However, • Revised write-up and survey questionnaire, both in English and Armenian; to date there has been no nationally representative baseline of �nancial • Cleaned datasets containing the entire data gathered through the interviews, including all codes in SPSS format and all corresponding capability levels of the Armenian documentation to the Trust Fund; population. Therefore, this project aims to provide CBA with detailed • Technical/statistical report including description of methodology, presentation of data structure and coding, discussion of any issues or and segmented information on the concerns encountered in the data collection process, and basic descriptive levels of �nancial capability of Arme- statistical analysis of the data (in Armenian and English). nians, help CBA identify and de�ne The survey results will be presented at a dissemination event in Armenia, which targeted programs of consumer will be held for all relevant stakeholders in government, �nancial, private and awareness and �nancial education, non-pro�t sectors. The full dataset and reports will be published on the website of the Global Program on Consumer Protection and Financial Literacy http:// and provide CBA with a baseline www.worldbank.org/consumerprotection. that will later allow measurement of overall outcomes of implementation of �nancial education and consumer protection measures. Further, the �ndings of the survey will provide demand-side inputs to 8 the FIRST-funded Financial Consumer Protection project in Armenia. These inputs will complement the �ndings of the diagnostic review of the Arme- nian legal and institutional frame- work for �nancial consumer protec- tion, conducted in early 2012 by the World Bank’s Global Program on Con- sumer Protection and Financial Lit- eracy. The survey and the diagnostic would then provide key inputs to the preparation of an Action Plan and the actual implementation of measures to improve consumer protection and �nancial capability in Armenia. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT COLOMBIA Financial Capability Measurement Project Objectives This project was undertaken by the Banco de la República and the World Bank with the objectives of: (i) understanding where the most important gaps in knowledge exist in the Colombian population, and Project Name for whom, with the goal of designing and targeting improved �nancial Colombia Financial Capabilities and education programs; (ii) benchmarking the level of �nancial capabilities Education Measurement Project of Colombians; (iii) supporting the design of public policies to enhance both the familiarity with and quality of �nancial services. Partner Institution(s) Banco de la República (Central Bank Focus Groups of Colombia); Invamer Colombia; A total of 10 focus group sessions were conducted, representing a mix Corporación Maloka of participants across: (i) low and middle income groups (ii) different 9 age groups and (iii) urban and rural populations (around Bogota, Timeline Bucaramanga, Cartagena, Neiva, and Pereira). Each focus group lasted July 2010 – September 2012 approximately two hours, which was the length of time for the average of eight focus group participants to complete their discussion of topics in the interview guide. In-Depth Interviews Individual interviews were conducted in two waves of 15 participants each to probe elements of their �nancial practices in greater detail and test question structure and wording. The participants in the �rst wave of these interviews had taken part in focus groups, while the participants in the second wave of interviews were new to the topics addressed in the questionnaire. Survey Implementation These newly-developed questions on �nancial capabilities are being applied in a nationally representative survey of approximately 1500 individuals. This survey, which will cover key areas such as daily money management, �nancial planning, and knowledge of basic �nancial concepts will build on a survey on �nancial burden, capability, and education (IEFIC for its initials in Spanish) that was developed by Banco The Russia Trust Fund for Financial Literacy and Education de la República and DANE in 2008 and initially applied only to the was established in October 2008 to support the advancement banked population in Bogota. of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Main Outputs and Dissemination The main outputs of the project will include synthesis of focus groups and Policy Impact in-depth interviews, a report on Stage 1 �ndings (conclusion of qualitative research), a quantitative baseline dataset on �nancial capabilities in Colombia, The results of the study will feed and a presentation of key �ndings from the quantitative data. directly into the Colombian national strategy and policy for �nancial edu- cation, in which Banco de la Repub- lica plays a key coordinating role. Greater knowledge of the �nancial practices of the population will sup- port better targeting of �nancial education initia-tives to distinct seg- ments of the pop-ulation. Ultimately, improved knowl-edge of �nancial issues in different segments of the population should permit Colom- bian citizens to achieve higher levels of individual and family welfare. 10 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT LEBANON Financial Capability Measurement Project Objectives The project aims to implement the �rst national survey that will measure �nancial literacy and capabilities in Lebanon. It will inform the government’s policy in various areas including but not limited to the Project Name design and development of a national strategy for promoting economic Measuring Financial Capability and and �nancial education of all Lebanese citizens as well as reforms in the Literacy in Lebanon social insurance system. The results of the survey will be used to create awareness of the existing level of �nancial literacy in Lebanon and a Partner Institution(s) baseline dataset to be used for further policy-oriented research including the evaluation of interventions that seek to improve �nancial literacy. Basil Fuleihan Institute of Finance (IOF); Consultation and Research Institute Survey Implementation 11 The topics covered in the survey will include: day to day money Timeline management, planning for the future, key motivations behind �nancial July 2011 – August 2012 behavior, understanding basic concepts of public �nance, and the level of �nancial literacy. For the �rst four components, the team will rely on a robust questionnaire developed by the technical team of the Russia Trust Fund. In addition, public �nance and �nancial literacy have been identi�ed as important topics to consider in the case of Lebanon and thus speci�c questions have been added to the questionnaire without affecting the way the core questions are administered. The survey will be implemented through household visits with a sample size of 1,200 individuals. The sample will be representative of the national adult (18+) population, probability sampling will be used and replacement of households and individuals will be documented. The 2004 and 2007 Household Survey issued by the Central Administration of Statistics (CAS) of Lebanon will be used as a reference document for the distribution of the questionnaire across all governorates. A pre-test of the questionnaire on 10–15 individuals will take place to ensure that the questions are well understood and solicit the desired information. A pilot test including 100 individuals will take place to pilot the questionnaire and methodology. The pilot will include households/ individuals from different regions, age groups and characteristics. Based on the results, the formulation of the questions and/or content The Russia Trust Fund for Financial Literacy and Education may be further adapted. was established in October 2008 to support the advancement of �nancial literacy and capability, through the development In order to ensure a high �eldwork quality, the quality of the survey of methods and best practices for the assessment of �nancial is expected to be consistently tested through a random sample capability and for the evaluation of �nancial education programs. drawn from the returned completed surveys. The control process will The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. consist of ensuring that all responses obtained have been entered in a correct manner and that the surveyor has compiled fully with the Trust Fund’s instructions. In order to be analyzed, the dataset will include sampling weights, incorporating the sample design weights, compensating for factors such as Policy Impact unequal probabilities of selection, non-response and to ensure the weighted sample matches a known population distribution. Financial Education and Awareness Lebanon is considered amongst the �rst Main Outputs and Dissemination countries in the MENA region realizing The outputs of the project are the following: the importance of �nancial literacy and education and the necessity of educating • completed and clean dataset of survey results; basic statistical description its citizens in economic and �nancial mat- of survey results; ters. In fact, the Lebanese Government has included related theories and concepts into • detailed methodology report. its public school curriculum; however, it has Moreover, the World Bank team will complete a �nal report summarizing the been noticeable that Lebanese youth faces results of the project, in terms of survey process and basic summary of results. dif�culties in understanding the main theo- ries and dealing with the practical implica- The results of the survey will be communicated to Government of�cials involved tions of public and/or private money man- in the various policy areas through continues policy dialogue and World Bank agement. As a result, this survey will make analytical and operational work. The survey data will also be published in raw it possible to assess the level of economic form. The Institute of Finance will produce a user-friendly citizen awareness guide and �nancial capabilities on a national distributed freely at a national level, in an effort to increase awareness on the matter. Moreover, the Institute of Finance plans to set up focus groups/ in-depth scale, in order to design and develop a interviews with key stakeholders to discuss and validate the main �ndings from national strategy, aimed at increasing the the data and the relevant implications. level of awareness and education of all Leb- anese Citizens which is expected to further improve the behavior of individuals. Beyond Financial Education: Social Protection Financial literacy and education emerges as 12 an important topic relevant to several prior- ity policy areas for the Government of Leba- non, which also the World Bank is engaged in. One of such areas is a long engagement in intensive policy dialogue and technical assistance with the Government on social security reform. Currently, Lebanon has only an end of service indemnity scheme and no pension system. A draft law to introduce a pension scheme is expected to be passed in the future. Furthermore, a voluntary health insurance scheme was recently discontin- ued, and the Bank is providing advice on reforming this scheme. The approach of this work and the interventions under discus- sion rely, to a large extent, on behavioral response to stimulus, and the impact is maximized when workers’ actions are based on a higher level of �nancial capability. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT MALAWI Financial Capability Measurement Project Objectives This joint project for Malawi, Namibia and Zambia aimed to develop a set of �nancial capability indicators that are internationally comparable and that take account of the needs of low-income people Project Name in developing countries. These indicators will result in better national Financial Capability Project policy formulation for �nancial literacy programs, which are critical for increasing levels of access to �nance in Malawi, Namibia and Zambia. Partner Institution(s) The purpose of the focus groups and in-depth interviews was to test FinMark Trust and establish participant views on the dimensions of �nancial capability in order to develop a de�nition of �nancial capability that is appropriate Timeline for low-income countries. The purpose of the in-depth interviews was September 2010 – March 2012 to obtain a better understanding of how to measure the dimensions 13 of �nancial capability, in order to assist with developing questions for a national survey. Focus Groups Eight focus group discussions (FGDs) were held in Malawi: three (3) in Blantyre, (2) in Lilongwe and (3) in Mzuzu. A total of 69 persons participated in the FGDs. The FGDs were split according to income, gender and income. The emphasis was on day-to-day money management and on saving. Planning for the future was considered important and involved investing in children’s education, investing in agriculture, rental housing or another type of business. Having a house to live in and land to farm was seen as long-term goals and retirement planning. There was not much emphasis on choosing and using �nancial products. Managing money in the context of a relationship was mentioned a number of times. Communicating and creating consensus within a household about money and priorities appears to be an important skill. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. In-depth Interviews Interview respondents were be selected on the basis of criteria set by the Trust Policy Impact Fund. An interview guide was provided by FinMark Trust/World Bank. In Malawi, 20 and 15 in-depth interviews were conducted in the �rst and second round Survey �ndings will assist policy respectively. After the interviews were completed, a debrie�ng was organized makers to understand the current with the Trust Fund team. The results and feedback from in-depth interviews levels of �nancial literacy and to put informed the development of core questionnaire modules to be used for surveys that will be implemented in the second stage of the project. appropriate policies in place to pro- vide protection and recourse for con- Main Outputs and Dissemination sumers of �nancial products and ser- The project produced the following outputs as deliverables: vices. Repeated surveys will enable policy makers to track �nancial lit- • Transcripts and feedback from focus groups and in-depth interviews; eracy levels of consumers across the • Detailed report on the results from the focus groups; region to see improvement of policy • Detailed report on the results of the in-depth interviews. changes/initiatives implemented. Financial education will encourage Results will be disseminated by the Trust Fund in regional events and online appropriate use of �nancial products through the Trust Fund website. Further dissemination plans will be de�ned in cooperation with FinMark Trust. and services and provide opportu- nities for poor men and women to make use of �nancial products and services and bring economic recov- ery in the region. 14 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT MEXICO Financial Capability Measurement Project Objectives The CNBV and CONDUSEF applied to participate in this project with the World Bank with the objectives of: (i) establishing where the most important gaps in knowledge exist in the population, and for whom, Project Name with the goal of designing and targeting better �nancial education Mexico Financial Capability programs, (ii) benchmarking the �nancial capabilities of the Mexican Assessment population relative to other countries and (iii) supporting the design of public policies to enhance both the knowledge about and quality of Partner Institution(s) �nancial services internationally. National Commission for the Protection and Defense of Users Focus Groups of Financial Services (CONDUSEF); A total of 13 focus group sessions were conducted that included a mix 15 National Bank and Securities of participants in: (i) low and middle income groups (ii) formal/informal Commissioner (CNBV); labor market participants and (iii) urban and rural areas near Guadalajara, IPSOS Mexico Hunucmá (Yucatan), Mexico City, Monterrey, and Oaxaca. Timeline In-Depth Interviews July 2010 – September 2012 Individual interviews were conducted in two waves of 24 participants each to probe elements of their �nancial practices in greater detail and test question structure and wordings. The �rst wave of these interviewees had taken part in focus groups, while the second wave of interviewees were new to the topics addressed in the questionnaire. Survey Implementation Approximately 2000 individuals will be interviewed in a nationally representative survey. This survey will cover key areas such as daily money management, �nancial planning, and knowledge of basic �nancial concepts and is projected to be repeated three years after the initial baseline is conducted. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Main Outputs and Dissemination The project will produce several outputs, including: Policy Impact • Transcripts and synthesis of focus groups and in-depth interviews; The survey is being designed to sup- port new and existing regulations, • Report on Stage 1 �ndings (conclusion of qualitative research); more re�ned targeting of policies, • Quantitative baseline dataset on �nancial capabilities in Mexico; product innovation and potentially, to • Analysis of key �ndings from quantitative data. provide a baseline for future impact evaluations related to consumer protection, �nancial inclusion and �nancial education. Demographic data collected as part of this survey will facilitate more interesting analy- sis of �nancial capabilities including interactions with other variables. Ultimately, improved knowledge of �nancial issues in different segments of the population should permit con- sumers to achieve higher levels of individual and family welfare. 16 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT NAMIBIA Financial Capability Measurement Project Objectives This joint project for Malawi, Namibia and Zambia aimed to develop a set of �nancial capability indicators that are internationally comparable and that take account of the needs of low-income people Project Name in developing countries. These indicators will result in better national Financial Capability Project policy formulation for �nancial literacy programs, which are critical for increasing levels of access to �nance in Malawi, Namibia and Zambia. Partner Institution(s) The purpose of the focus groups and in-depth interviews was to test FinMark Trust and establish participant views on the dimensions of �nancial capability in order to develop a de�nition of �nancial capability that is appropriate Timeline for low-income countries. The purpose of the in-depth interviews was September 2010 – March 2012 to obtain a better understanding of how to measure the dimensions 17 of �nancial capability, in order to assist with developing questions for a national survey. Focus Groups Eight FGDs were conducted in Namibia in locations spread throughout the country, one (1) in Windhoek, one (1) in Gibeon, two (2) in Oshakati, one (1) in Ondangwa, one (1) in Eenhana, one (1) in Walvisbay and one (1) in Swakopmund. A total of 63 persons participated in the FGDs. In Namibia, as there is a greater level of formal �nancial product penetration, the market appears quite different from Malawi and Zambia and issues such as choosing products and the information required to make informed choices are more relevant in the Namibian context. There was a heavy emphasis on day-to-day money management issues, with alcohol, impulse shopping and credit being areas of concern that affect money management. In-depth Interviews Interview respondents were selected from the focus group participants on the basis of criteria set by the Trust Fund. In Namibia, the interview guide was not translated and the English version was used, although the interviewers did translate for the respondents, where English provided The Russia Trust Fund for Financial Literacy and Education a problem. The researchers struggled to �nd some of the respondents. was established in October 2008 to support the advancement This may be as a result of their initial selection, but people also appear of �nancial literacy and capability, through the development to be very mobile in Namibia. Respondents were selected by FinMark of methods and best practices for the assessment of �nancial on the same basis as for Zambia to provide for a spread across gender, capability and for the evaluation of �nancial education programs. income and the level of capability. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. A total of 14 and 15 interviews were conducted in the �rst and second round respectively. After each set of interviews, a debrie�ng was held with the Trust Fund team and results and feedback informed the development of core Policy Impact questionnaire modules to be used for surveys that will be implemented in the second stage of the project. Survey �ndings will assist policy makers to understand the current Main Outputs and Dissemination levels of �nancial literacy and to put The project produced the following outputs as deliverables: appropriate policies in place to pro- vide protection and recourse for con- • Transcripts from focus groups and in-depth interviews; sumers of �nancial products and ser- • Detailed report on the results from the focus groups; vices. Repeated surveys will enable • Detailed report on the results of the in-depth interviews. policy makers to track �nancial lit- eracy levels of consumers across the Results will be disseminated by the Trust Fund in regional events and online through the Trust Fund website. Further dissemination plans will be de�ned in region to see improvement of policy cooperation with FinMark Trust. changes/initiatives implemented. Financial education will encourage appropriate use of �nancial products and services and provide opportu- nities for poor men and women to make use of �nancial products and services and bring economic recov- ery in the region. 18 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT PAPUA NEW GUINEA Financial Capability Measurement Project Objectives The objective of the project is to assist PNG Central Bank (BPNG) and related government agencies to develop an understanding of �nancial capability in Papua New Guinea (PNG).This national survey is particularly Project Name relevant to �nancial sector policy development, especially in the case Financial Competency Study of micro and SME �nance, as it will result in a baseline database with information on behaviors in the use of money, decision making around Partner Institution(s) money, �nancial planning habits, and access to different types of formal and informal �nancial services. Bank of Papua New Guinea (BPNG); UNDP Paci�c Financial Inclusion Undertaking a comprehensive study of �nancial capability in PNG Program (UNPFIP) increases the ability of the national government and of international organizations to support local communities as they monetize. It also 19 Timeline enables the further development of methodologies to identify and October 2010 – June 2012 measure �nancial competencies. The primary goal of the survey and related efforts is to provide useful information on �nancial capacity of PNG population so as to facilitate the government to develop appropriate policy, regulation and program in promoting greater �nancial inclusion in PNG. Focus Groups A total of 8 focus group sessions (4 female and 4 male) were conducted for the three sample Communities: Talai Settlement, in Port Moresby, the nation’s capital (2 male and 2 female); Cloudy Bay Community where the PNG Sustainable Development Program, Ltd. operates (PNGSDP) (1 male and 1 female); and, The Rubber Plantation whose employees come from all over PNG, but predominantly from the Highlands provinces (1 male and 1 female). The interviews were conducted in two local languages: Tok Pisin and Hiri Motu. Each interview started with explanation of the objectives and the procedures. Consent for interviews was sought from all respondents of the focus groups, including permission to record the interviews. The respondents were informed that they had the right to tell the moderator and facilitators if there was anything they did not want recorded at any time during the interviews. In addition, permission The Russia Trust Fund for Financial Literacy and Education was sought from the respondents to allow information disclosed to be was established in October 2008 to support the advancement used for the purpose of the project. of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial The interviews were conducted in either community halls or conference capability and for the evaluation of �nancial education programs. rooms at interviewees premise or the interviewees were provided The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. transportation to the meeting venue outside their residence. A second run of the Talai Settlement group was found necessary to get the relevant information. About 4 hours were used to go through the questions on the focus group interview guide. Two professional transcribers were hired to transcribe Policy Impact verbatim the focus group sessions. There is emerging recognition of the In-Depth Interviews need for government programs to tackle the widespread and weak lev- The �rst round of in-depth interviews were conducted in the Talai Settlement and Galley Reach Estate communities among respondents who previously els of �nancial capability among the participated in the focus group sessions. A total of 16 in-depth interviews were low-income population of PNG. The completed: 7 female and 9 male respondents; 4 from Galley Reach Holdings results of the survey will feed directly and 12 from Talai Settlement. into the formulation of a national pol- Survey Implementation icy and strategy for �nancial inclu- sion and education to be led by the The national survey plans to use the same sampling as the PNG HIES (some output of tables yet to be completed) with some re-weighting of census districts Bank of Papua New Guinea (BPNG). due to migration in the inter-census period. The survey will contain additional It is anticipated that the work will PNG-speci�c questions to the set of core questions developed by the multi- be coordinated and undertaken by a country study so as to focus more on �nancial capacity of households in PNG. national task group to be convened Due to the complexity of local logistics and the upcoming election in June and chaired by BPNG. This group 2012, the national survey is postponed and will be conducted after the election. A team of 15 local facilitators/interviewers (both male and female) has been will be comprised of key stakehold- trained to carry out the pilot interviews in Port Moresby areas. Questionnaires ers including government agencies have been translated into local languages (Tok Pisin and Motu) and the Pisin (�nance, national planning, educa- version has been tested. Pilot testing of questionnaire has been completed and data will be collected and analyzed to validate the instrument and constitute a tion, provincial government, national �rst step towards understanding the challenges for �nancial education in PNG. statistics of�ce), the bankers associa- tion, micro�nance associations and Main Outputs and Dissemination the Institute of National Affairs. The The key output of the PNG survey as of June 2012 will be the PNG instrument World Bank and PFIP teams will func- 20 validation and the analysis report on the pilot conducted among randomly tion as technical assistance advisors selected communities in Port Moresby areas. Data analysis of the pilot in Port Moresby will reveal speci�c areas of competencies requiring strengthening to this task group. among the sampled households, which are representative of major urban and peri-urban population in PNG. BPNG plans to carry out the national survey once the HIES data is made available for national sampling and after the election. BPNG expects that that national survey will provide the needed information to facilitate the design of targeted and cost-effective interventions. This valuable dataset will also enable the measurement of progress in programs aimed at improving �nancial competency. Over the long term, improvements in the levels of �nancial literacy and competency of the PNG population are expected to have a positive impact on the level of �nancial inclusion in the country. The survey data will be made available through BPNG to the public and the key �ndings from the PNG survey will also be shared with relevant stakeholders in PNG through the task group on �nancial literacy chaired by BPNG (detailed below) through a workshop and a World Bank/BPNG joint publication. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT TANZANIA Financial Capability Measurement Project Objectives In 2009–10, a National Financial Literacy Framework was developed for Tanzania. The development of the framework was an initiative the Bank of Tanzania, supported by the Financial Sector Deepening Trust (FSDT). Project Name The framework was recently approved by the Government of Tanzania Tanzania Financial Capability and it identi�ed the need for de�ning the dimensions of �nancial Baseline capability in the country. It also identi�ed the need for a baseline survey, to enable policymakers and practitioners to con�rm priority Partner Institution(s) population segments for planned �nancial education initiatives and tracking �nancial capability progress over time within these segments Financial Sector Deepening Trust, and on a national level. Tanzania (FSDT); Bank of Tanzania, Marketworx Africa; National Bureau of Statistics Focus Groups 21 Eight focus groups were conducted in Dec 2010–Jan 2011 in three Timeline different regions (including Zanzibar). The groups were segmented October 2010 – September 2012 in terms of gender, income (low, middle and high) and age. Each group consisted of approximately eight participants. The focus group discussion guide was modi�ed for the local context and translated to Swahili. The moderator was Swahili speaking and assisted by a note taker. The group discussions were taped, transcribed, translated and the �ndings of each summarized in a table. In-Depth Interviews Seventeen in-depth interviews (IDIs) were conducted in March 2011. The purpose of the IDIs was to test and re�ne the �rst draft questionnaire. Respondents were selected from each of the focus groups conducted in December/January and reflected those which were perceived to be most and least �nancially capable based on the group discussions. The draft questionnaire was localized and translated to Swahili. The interviewers were Swahili speaking and assisted by a note taker. The interviews were taped, transcribed (with direct translation) and the �ndings summarized on a data. Findings and recommendations on the questionnaire were also captured onto a log frame. The Russia Trust Fund for Financial Literacy and Education A further six IDIs were conducted in August of 2011 to further re�ne the was established in October 2008 to support the advancement questionnaire. The same methodology was followed as with the �rst of �nancial literacy and capability, through the development set of IDIs, but this time respondents were included which were not of methods and best practices for the assessment of �nancial part of the focus group process. capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Survey Implementation A nationally representative sample (including Zanzibar) will be undertaken Policy Impact through strati�ed random sampling with approximately 3200 adults. The questionnaire will �rst be translated, piloted, adjusted, and back translated The �ndings of the survey will support before �nalization and full roll-out. The National Bureau of Statistics will assist other �nancial education and con- with the design and drawing of the sample, as well as �eldwork oversight, sumer protection initiatives launched whereas Synovate will conduct the actual �eldwork. Data will be captured, cleaned and weighted before data modeling, segmentation and analyses. by the Government of Tanzania. This is part of a broader set of goals to Main Outputs and Dissemination facilitate improved access to �nancial The main outputs from the focus groups (FG) include: services and strengthen consumer • Translated transcripts of each FG; protection.The results will assist poli- cymakers and practitioners in identi- • Summary analyses of each FG discussion; fying priority market segments and • Summary report of all FGs; the types of issues to be addressed • Presentation to the Trust Fund. through �nancial education and con- The main outputs from the in-depth interviews include: sumer recourse interventions. Finally, • Translated transcripts of each IDI; the survey will allow interested stake- • Summary data sheet for all IDIs; holders to track changes in consumer • Summary log frame with comments and suggestions on draft questionnaire; �nancial behavior and capability over • After second round IDIs: draft questionnaire for testing in pilot survey. time and thereby re�ne and improve Finally, upon completion of the survey �eldwork, the project will produce: future interventions. • Final questionnaire; • Data set; • Final analytical report. 22 A dissemination strategy for the survey will be developed and implemented within Tanzania. When results are available, the project team plans to hold a workshop with key stakeholders to launch the �ndings of the survey. The media will be invited as a matter of course. Smaller workshops will then also be held with interested parties such as the �nancial sector, educational sector, media, etc. The survey will be posted on the FSDT website for public access. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT TURKEY Financial Capability Measurement Project Objectives A nationwide survey of �nancial capability in Turkey is expected to inform the government’s current efforts to introduce a national strategy for �nancial education. The recent Turkey Country Economic Project Name Memorandum on Domestic Savings identi�ed this as a key policy Turkey Financial Capability Survey area in the country. The Financial Capability Survey is expected to meet two main objectives: (1) identify policy gaps and inform policy Partner Institution(s) options to mobilize domestic saving in Turkey (i.e. including �nancial education in school curriculums) and (2) establish the basis for Capital Markets Board of Turkey; regularly monitoring the level of �nancial literacy of households. Turkish Statistics Agency; Central The results of the survey will be shared widely to increase public Bank of Turkey awareness on the subject. 23 Timeline Survey Implementation October 2011 – June 2012 The sample size is 3000. Respondents will be selected randomly from the population at or above 18 years of age and it will exclude the population in institutions, in line with guidelines provided by the Turkish Statistics Agency (Turkstat). The survey will be nationally representative and will cover all 12 NUTS-de�ned regions of Turkey (NUTS Level 1 Statistical Regions). Except for Istanbul (that constitutes a region by itself), provinces from each region will be determined by the random selection of primary sampling units (i.e. districts). The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Main Outputs and Dissemination The main outputs are the micro-level data collected through the survey and a Policy Impact report containing the summary statistics of the survey. A joint dissemination event will be planned with the main counterpart agencies following completion Financial literacy and education is of the survey. on the agenda of the Turkish govern- ment. In March 2011, the Central Bank of Turkey and the Capital Markets Board organized an international con- ference on �nancial education with the main goal of raising awareness and laying the ground for the prepa- ration of a national strategy for �nan- cial education. This is the �rst such comprehensive attempt at improving the level of �nancial literacy of Turk- ish citizens. The World Bank and Turkey’s Min- istry of Development (with inputs from the Central Bank, Treasury and Capital Markets Board) have recently completed a Country Economic 24 Memorandum (CEM) on the role of domestic saving in Turkey’s eco- nomic growth. The CEM identi�es the low level of �nancial literacy as a key constraint for mobilizing domes- tic saving. A nationwide survey of �nancial capability will be a critical tool to inform the government’s formulation of a national �nancial education strat- egy and this is identi�ed in theTurkey CEM as a key policy area. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT URUGUAY Financial Capability Measurement Project Objectives Despite being one of the smallest countries in Latin America with approximately 3,300,000 inhabitants, Uruguay has a history of being a pioneer in technological and social innovation in the region. In keeping Project Name with this tradition of developing creative solutions to meet various social Uruguay Financial Capability and economic challenges, Uruguay is taking a similar approach to the Assessment problem of �nancial illiteracy. The Country sought to bene�t from the expert guidance of the Russia Trust Fund �nancial capability experts. In Partner Institution(s) order to make use of the latest learning in this area, the Central Bank of Uruguay made a commitment to engage in this groundbreaking Central Bank of Uruguay (BCU); international study. The Financial Capability Assessment in Uruguay Equipos Mori aims to establish a baseline measure of �nancial capability for the Uruguayan population. It also endeavors to understand, conceptually, 25 Timeline what constitutes �nancial capability in the Uruguayan context, especially October 2010 – June 2012 for low-income, �nancially excluded populations. The data provided by the assessment will inform the agenda for effective �nancial capability and economic education programming development by the Central Bank for all sectors of the population. Focus Groups During November and December 2010, nine focus groups were undertaken both in Montevideo and in the interior of the country in the city of Tacuarembó. Based upon the Trust Fund approach of intensive, qualitative study of �nancial capability in low-income groups during the focus group stage, six groups consisted of low-income participants and three were drawn from middle or upper-middle income participants. The low-income participants were selected from those living below the poverty line based on a socio-economic measure and at least half with basic needs unsatis�ed as de�ned by the Basic Needs Index. Participants included 22 men and 44 women between the ages of 15 and 80. The group discussions provided the conceptual inputs for the development and adaptation of the national survey instrument. The groups were conducted over approximately 2 hours each and according to a facilitators guide developed in conjunction with the Trust Fund experts. Among other questions, the The Russia Trust Fund for Financial Literacy and Education groups explored: how a �nancially capable and incapable person is was established in October 2008 to support the advancement de�ned; how participants understand the components of �nancial of �nancial literacy and capability, through the development capability; what terminology is used to discuss these concepts; of methods and best practices for the assessment of �nancial what perspectives exist on the subjects and how they are grouped capability and for the evaluation of �nancial education programs. naturally by the participants. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. In-Depth Interviews A �rst round of 15 in-depth interviews were conducted between February and Policy Impact March based on the initial dialogues initiated in the focus groups and looked to deepen this conversation with participants and test comprehension of the The Central Bank of Uruguay (BCU) questions to be used in the national survey. The interviews were conducted in has developed and instituted a per- Montevideo and Tacuarembó with �ve men and ten women between the ages manent program in Economic and of 15 and 70. Another set of 15 interviews were conducted to test the modi�ed version of the survey questionnaire. Financial Education: BCU Educa which looks to improve the everyday Survey Implementation money management skills of Uru- Following a pilot undertaken in December 2011 to re�ne the survey instrument guayan citizens. The Trust Fund sup- and �eld strategy, the national survey was conducted in May and early June ported Financial Capability Assess- 2012. It was a face-to-face household survey on the total population of Uruguay, ment will provide useful inputs for rural and urban, over 18 years of age. To be eligible, subjects had to be at least in charge of their own personal �nances. The survey was based on a 1400 developing and re�ning the BCU respondent sample. A survey team with experience in complex surveys carried Educa program. It will offer a core out the interviews. Interviewers underwent rigorous training and the �eld work diagnostic that will be used in the was closely supervised. The maximum margin of error calculated for this sample size is +/– 2.6%, with a 95% con�dence interval. The results are presently being design of activities to be carried analyzed by the Country team and the international expert group. out in 2012 and beyond. In addition, the results of the Financial Capabil- Main Outputs and Dissemination ity Assessment will be shared with A �nal report is presently being produced pro�ling the lessons from the Uruguay government bodies and other rel- survey and background research. In terms of dissemination of the survey evant groups supporting inter-insti- methodology and �ndings, the project will look to share the results nationally, regionally and internationally. Nationally, the results will be shared in relevant tutional policy development aimed government and educational realms, as well as with specialized audiences. at increasing levels of �nancial inclu- Presently the Central Bank and the World Bank are working together to develop sion in Uruguay. a communications dissemination strategy for the results. 26 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT ZAMBIA Financial Capability Measurement Project Objectives This joint project for Malawi, Namibia and Zambia aimed to develop a set of �nancial capability indicators that are internationally comparable and that take account of the needs of low-income people Project Name in developing countries. These indicators will result in better national Financial Capability Project policy formulation for �nancial literacy programs, which are critical for increasing levels of access to �nance in Malawi, Namibia and Zambia. Partner Institution(s) The purpose of the focus groups and in-depth interviews was to test FinMark Trust and establish participant views on the dimensions of �nancial capability in order to develop a de�nition of �nancial capability that is appropriate Timeline for low-income countries. The purpose of the in-depth interviews was September 2010 – March 2012 to obtain a better understanding of how to measure the dimensions 27 of �nancial capability, in order to assist with developing questions for a national survey. Focus Groups In Zambia, eight focus group discussions (FGDs) were held with a total of 72 FG participants. The FGDs took place in and around Lusaka in the central region of Zambia, and in and around Ndola, in the Copper belt region. These locations provided both urban and rural areas, and represent different types of economic activity in Zambia. As per the guidance from the World Bank expert group, the FGDs were spilt according to gender, income and rural/urban location. Two of the focus groups in each of the regions were held in urban areas, and two in rural or peri-urban areas around the urban centres. The primary emphasis was on day-to-day money management and on planning for the future. Budgeting and prioritization of expenditure, as well as �nancial discipline and self-control are closely associated with �nancial capability. There was only limited mention of the need to keep track of money. The importance of managing money as a family or a couple is mentioned repeatedly. In particular, the concern was that in cases where there is no joint planning and no joint vision for the family, it becomes impossible to make ends meet or even to plan for the future. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. In-Depth Interviews An interview guide was provided by the World Bank. In Zambia, the guide was Policy Impact translated into Bemba and Nyanja and the translations were independently back translated. The translations were reviewed after the initial back-translation The project will contribute to the to address all concerns. A total of 20 respondents were interviewed in each development of survey tools to assist of two rounds. The respondents were selected on the basis of gender and policy makers to understand the cur- income, but also on the basis of whether they were considered more or less �nancially capable, based on the information available from the pre-screening rent levels of �nancial literacy and questionnaire. After each set of interviews, a debrie�ng was conducted. The to put appropriate policies in place results and feedback from in-depth interviews informed the development of to provide protection and recourse core questionnaire modules to be used for surveys that will be implemented in the second stage of the project. for consumers of �nancial products and services. Financial education Main Outputs and Dissemination can encourage appropriate use of The project produced the following outputs as deliverables: �nancial products and services and provide opportunities for poor men • Transcripts from focus groups and in-depth interviews; and women to make use of �nancial • Detailed report on the results from the focus groups; products and services and bring eco- • Detailed report on the results of the in-depth interviews. nomic recovery in the region. Results will be disseminated by the Trust Fund in regional events and online through the Trust Fund website. Further dissemination plans will be de�ned in cooperation with FinMark Trust. 28 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Monitoring and Evaluation Monitoring and Evaluation When resources are scarce and social safety nets are weak, households’ ability to manage income and assets wisely may be an important determinant of economic security. However, many open questions remain about how households in developing countries gain and exercise �nancial capability, and the best ways for governments and the private/nonpro�t sector to help increase it. With the exception of a small but important number of studies that have recently been completed or are currently underway, robust evidence regarding the ef�cacy of �nancial capability interventions is relatively sparse compared to the level of interest and programmatic activity. One 31 reason for this is a lack of systematic evaluation. Although there is some evidence that �nancial education can enhance �nancial knowledge, there is limited empirical evidence that �nancial knowledge has an impact on behavior and outcomes, and on the ultimate goal of improving �nancial well-being, and little evidence exists of the impact of interventions intended to enhance �nancial capability. The objective of this element of the program under the Trust Fund is to develop and document methods for the impact evaluation of �nancial education and capability enhancement programs while contributing to the evidence base on the effectiveness of these types of programs. Through development and testing of methods and funding the evaluation of a wide range of projects in diverse settings, the Trust Fund aims to contribute to the knowledge base on the ef�cacy of �nancial education interventions, the impact of different ways of delivering these programs, and support capacity-building in client countries. To achieve these objectives, the Trust Fund is undertaking two complementary initiatives: • Development of an operational toolkit on methods for evaluating �nancial capability programs. The development of the toolkit has been contracted to RAND Corporation. • A set of individual pilot projects to test interventions, and a smaller set of cross-country comparative projects. These projects are intended to develop methods for evaluation and analysis and to inform the preparation of the evaluation toolkit. The knowledge derived from these efforts will be used to inform policy makers and others about the impact of �nancial capability on behavior related to �nancial-decision making, the types of programs that help enhance these levels of �nancial capability, and most importantly, improving the capacity of countries to develop effective �nancial education strategies over the longer term. 32 Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT AUSTRALIA and NEW ZEALAND The Impact of Financial Literacy Training for Migrants at Destination Purpose of Research Remittance costs are high in the Paci�c, which reduces the development potential of these flows. Remittance-speci�c �nancial literacy training is being used in Australia and New Zealand to determine whether this Thematic Area will serve to induce migrants to choose lower-cost methods. Remittances, credit and �nancial product selection Background and Objectives The World Bank and the United Nations Global Forum on Migration Target Segment have both made reducing the cost of remittances a core element of Migrants from Tonga, Samoa, Sri their global migration policy agendas. However, the ef�cacy of policies geared toward reducing the cost of remitting and spurring competition 33 Lanka, Hong Kong, Korea, China by increasing disclosure of costs relies heavily on the abilities of migrants to understand how to use the different methods available for Delivery Channel remitting and the costs incurred for each method. While systematic Group-based �nancial literacy evidence on the �nancial literacy of migrants is scarce, the data seminar available suggest migrants often lack knowledge of the components of a remittance cost, available methods, or how to compare such Evaluation Team methods. Thus there seems to be a natural and valuable opportunity David McKenzie (World Bank), in the Paci�c for �nancial literacy training to change remitting behavior. John Gibson (World Bank), The main goal of the present research is, therefore, to estimate the Bilal Zia (World Bank) causal impact of �nancial literacy training for migrants on their remitting behavior. In particular, the research will assess whether (a) �nancial Partners literacy training leads migrants to adopt new, cheaper, products such Ministry of Paci�c Island Affairs, as debit cards for sending remittances; (b) whether �nancial literacy New Zealand; University of Waikato training changes the amount and frequency of remitting, and therefore the amount received by migrants in the home country; and (c) how Timeline these effects differ by country of origin. March 2011 – April 2012 Additionally, many migrants do not use credit cards and rely on more expensive forms of credit such as pay-day loans and hire purchase agreements. A second component of the �nancial literacy training will be to provide information on the costs of these alternative forms of credit, and information on how to apply to get a credit card, with the goal of seeing whether this reduces the use of more expensive forms of credit. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement The training is delivered in a two-hour interactive meeting (or “fono�) of �nancial literacy and capability, through the development to groups of 20–30 migrants. It explores reasons people might opt to of methods and best practices for the assessment of �nancial remit. These include factors such as cost, convenience, speed, trust, capability and for the evaluation of �nancial education programs. and security driving remittance channel choice. It also covers the new The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. remittance debit card available in New Zealand, the different components of a remittance fee. Policy Impact Evaluation Methodology The training was fairly cheap to New Zealand and Australia were chosen as the migrant destinations for a number deliver, with the main costs aver- of reasons: (a) the World Bank’s strong policy ties to the remittance cost issue in these countries; (b) the high immigrant population (over one-�fth of the aging approximately $15-25 per total population) in these two countries, as well as the the common origins of attendee. The training did have some immigrants to these two countries which allows comparisons across two settings impacts on knowledge and behavior, with similar migrant groups; and (c) the large variation in remitting patterns found in the immigrants of these two countries, as well as the competitiveness of their but did not lead to �nancial bene�ts remittance market. This variation is an important point as it presents a range of over the period we measured. migrants from Paci�c countries who remit frequently and face high costs of doing so, to migrants from countries like Hong Kong who remit less frequently and face When asked participants why they relatively lower costs when they do. were not using the cheapest prod- Participants were recruited through community gatherings such as churches, and a uct, convenience was often the short baseline survey was then conducted. The �nal sample consists of three groups of migrants: 349 Paci�c Island migrants (mostly from Tonga and Samoa) interviewed reason given. One reason conve- in New Zealand; 352 East Asian migrants (from Korea, Hong Kong and China) nience may win out could be that interviewed in New Zealand; and 209 Sri Lankan migrants interviewed in Australia. the amounts saved through better In each group, migrants were matched into pairs on the basis of country of origin, city of residence, remitting patterns, and �nancial literacy, and then one individual in �nancial literacy may be too trivial each pair was randomly selected to be invited to attend the training sessions. to warrant action – the annual sav- Short-term follow-up surveys were then conducted at one, two, and three ing from switching to the cheapest months after training to track short-term impacts of the training on knowledge product may at most be $23-46. The and behavior; with a follow-up after nine months scheduled. Experimental impacts scope for changes in outcomes may of the program will then be obtained through comparison of the treatment and control groups. therefore be much larger if the focus is more on savings and budgeting Results behavior and getting migrants and 34 • The training led to large increases in speci�c �nancial knowledge that was their families to use the money taught during the course – migrants were 12-16 percentage points more likely to know it is cheaper to send one large transfer than two smaller ones; and 10- they have more effectively, rather 52 percentage points more likely to know the cheapest method of remitting. than just savings on transactions • Knowledge gains were highest for the Paci�c Island sample, who had the costs. Stay tuned for an impact note lowest initial levels of �nancial literacy. which looks at this type of training • Migrants changed some �nancial behaviors in response to this new in Indonesia. knowledge: they were more likely to use different sources of information to compare the costs of remitting; and the training made them less likely to Meanwhile, governments trying switch to new remittance methods that provide little bene�t from switching. to lower the cost of remittances • However, the training had no effect on either the frequency of remitting or the amount remitted, or on take-up of the new cheapest products. appear to need to do more than • There are some modest changes in use of credit among the Paci�c Island just supply better information and sample, who set up ROSCAs to avoid the high costs of payday loans. improve �nancial literacy – target- ing barriers to the entry of new providers and barriers to access of �nancial services on the receiving country side may also be needed. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT BRAZIL Impact Evaluation of Brazil’s School-Based Financial Education Program Purpose of Research This impact evaluation uses a randomized control trial (RCT) to assess whether high school-based �nancial education in Brazil improves Thematic Area students’ and parents’ �nancial knowledge, attitudes and behavior. Budgeting, savings and general Access to �nancial services has expanded rapidly in Brazil and the high �nancial education topics school-based �nancial education initiative aims to provide youth and their parents with tools that allow them to make sound �nancial decisions. Target Segment High school students and their parents Background and Objectives Improving the �nancial capability of the population so that citizens 35 Delivery Channel are able to make effective decisions around personal �nances and money management is increasingly more prominent on the policy Textbook with �nancial education agenda of governments of many low and middle income countries. case studies incorporated into the This is especially pertinent in Brazil, where the rapid evolution of the existing high school curriculum; �nancial markets has resulted in many inexperienced and vulnerable parent workshops in schools with consumers accessing different �nancial products and services, often �nancial education videos with very negative consequences. A 2008 survey found that 82% of Brazilian consumers were unaware of the interest rate when borrowing Evaluation Team money, that overdue installments were mostly caused by poor �nancial Rogelio Marchetti (World Bank), management, and that the saving rate of Brazilians is low, even among Miriam Bruhn (World Bank), affluent families. The survey also showed that 87% of families do not save for the future, and that 40% do not make any sort of investments Bilal Zia (World Bank) with excess income. In 2003, Brazilian Institute of Geography and Statistics found that family saving rates ranged from 0.06% of income Partners in the low income class, to 3.12%, among the wealthy. Comissão de Valores Mobiliários To empower the population to access �nancial services and make (CVM), Bovespa Brasil-Sao Paulo sound �nancial decisions, the Brazilian government assembled a Stock Exchange working group in November 2007 for the development of a National Strategy for Financial Education. The working group was composed of Timeline representatives from the Committee for Regulation and Supervision of August 2010 – June 2012 Financial Markets, Capital, Insurance, Pension Plans and Capitalization (COREMEC), in partnership with the Ministries of Education, Finance, Justice and Social Security, and representatives from several private sector institutions. As part of this strategy, the “Financial Education in The Russia Trust Fund for Financial Literacy and Education Schools� program was created with the objective of reaching students was established in October 2008 to support the advancement of �nancial literacy and capability, through the development in all public schools throughout the country. The program is coordinated of methods and best practices for the assessment of �nancial by the Securities and Exchange Commission of Brazil (CVM) and capability and for the evaluation of �nancial education programs. developed in partnership with the Ministry of Education (MEC), state The Trust Fund is managed by the World Bank implemented in part and municipal education bodies, and private sector institutions. A by the Organization for Economic Co-operation and Development. pilot project was introduced in 2010 with approximately 26 thousand students in 900 public high schools, comprising treatment and control groups, in the states of Ceará, Minas Gerais, Rio de Janeiro, São Paulo, Tocantins, and Distrito Federal. Policy Impact The program focuses on teaching �nancial literacy in high schools to educate students early in their life about �nancial decision-making. Although high school students are This study is the �rst large-scale, generally faced with fewer �nancial decisions than older adults, it is expected that they will be able to apply and practice some of the concepts taught in the courses, rigorous, impact evaluation to such as budget management, savings and responsible spending. For example, an measure whether a high school- increasing number of youth have the opportunity to use cell-phones, credit cards and make purchases in installments. Other concepts will become more relevant later in life, based �nancial education pro- but the �nancial literacy education is also meant to provide students with a knowledge gram can successfully change base that will allow them to make more informed �nancial decisions in the future. �nancial knowledge, attitudes, To assess whether the desired outcomes and changes in behavior are achieved through and behavior among students such intervention, and before scaling-up the program throughout Brazil, the government engaged with the World Bank to evaluate the pilot intervention. This study, supported and their parents. The results of by the Russia Trust Fund for Financial Literacy and Education, will produce a rigorous this study will inform the roll-out impact evaluation and analysis of the program employing a randomized methodology to of the program to other public study the impact on students and the household �nancial decision making in Brazil, as well as identify the mechanisms through which �nancial literacy helps change behavior. schools in Brazil. They will also The �nancial literacy program for students focuses on introducing a high school be useful for other governments curriculum of �nancial education integrated within several subjects. The student in developing countries that are textbook, with accompanying teacher manual, consists of 72 case studies and considering the possibility of exercises that can be introduced in any subject, mostly used in the Mathematics, Portuguese, Science, and History curricula. The material is interactive and includes introducing �nancial education exercises for students to be completed with participation of their household members programs in schools. (e.g. household budgeting and planning). The idea is that young people are the carriers of information and innovation in the household and would be the main vehicle for �nancial cultural change in the household, especially, inter-temporal decision-making on consumption and investment. Evaluation Methodology The impact of the program is evaluated through a randomized experiment at the school 36 level. Some schools have been randomly selected to receive the �nancial education textbooks and teacher training explaining the purpose, content, and application of the material. Other schools will not receive the text books or teacher training during the pilot phase. The pilot will run throughout three school semesters, from August 2010 to December 2011. By comparing treatment schools to control schools, the methodology will allow for measuring impact along a causal chain of (i) how �nancial literacy affects knowledge, (ii) how that improved knowledge leads to better �nancial decisions, and �nally (iii) how those better �nancial decisions improve household welfare. The project is divided into two parts: (i) measuring the impact of �nancial literacy programs for high school students on their �nancial knowledge acquisition and changes in �nancial decision making and behavior of their households; and (ii) measuring the impact of a �nancial literacy program for parents of high school children. The behavior changes to be measured include changes in household �nancial attitudes and decisions, and subsequently changes in household’s consumption, income, health and education expenditures. Information on �nancial knowledge, attitudes, and behaviors will be collected through surveys at the student and household levels. The Brazilian �nancial education project differs from other currently existing impact evaluation of �nancial education in that the �nancial education is administered through schools. It thus features a broad coverage of the target population and a comprehensive and longer-term coverage of the �nancial education material. This design will inform the rollout of the �nancial literacy program to more public schools in the country after the completion of the pilot study. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT BRAZIL From Shop to Stock: Understanding and Improving Household Investment Behavior Purpose of Research Drawing from the current literature on behavioral �nance, this study will work with the Brazilian stock market (BM&FBOVESPA) to test a range Thematic Area of interventions to better understand why people make the decisions Risk management, savings and they do and identify how to reduce common investor biases observed investment in the stock market through targeted �nancial education. Target Segment Individual investors Background and Objectives Delivery Channel In the next 20 years an estimated two billion people will enter the formal 37 Online stock market simulator �nancial system. Access to a broader range of �nancial instruments offers households an opportunity to diversify their assets but as �nancial Evaluation Team access improves, it becomes increasingly important to understand the mechanisms that drive household investment choices. This is important Arianna Legovini (World Bank), both to improve consumer protection (ensuring that investors make Vincenzo Di Maro (World Bank), fully informed decisions that do not expose them to unintended risk) Martin Kanz (World Bank), and to ful�ll a broader economic development objective of improving Aidan Coville (World Bank), savings rates to drive investment and strengthen capital markets to Deniz Anginer (World Bank) ensure ef�cient distribution of capital. Astrid Zwager (World Bank) Barriers to stock market participation include high transaction costs, Partners lack of commitment products, and – importantly – low levels of �nancial education. A substantial amount of evidence also supports the view BM&FBOVESPA Brazilian Stock that investors select investments for reasons that cannot be explained Exchange by standard models based on full rational agents, and that they often exhibit biases that result in sub-optimal investment decisions. Timeline For example, even when aware of the bene�ts of diversifying their January 2012 – December 2014 portfolios, investors might decide not to do so for a variety of behavioral reasons such as procrastination, information and menu costs, loss and ambiguity aversion. Common investment mistakes can be explained by a series of biases that affect investors: among others, investors commonly show a disposition effect (sell winners too quickly and hang on to losing stocks to avoid realizing losses), confuse good companies for good stocks, tend to be overcon�dent and hence overestimate the The Russia Trust Fund for Financial Literacy and Education accuracy of their forecasts, and show a preference for stocks with was established in October 2008 to support the advancement lottery features. of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. This evaluation has been designed to understand why: 1. Investors select investments for reasons that cannot be explained by Policy Impact standard models based on full rational agents; and The study offers a unique oppor- 2. Inexperienced investors often exhibit biases that result in suboptimal investment decisions. tunity to study the individual and interaction effects of �nancial liter- Understanding why people make the investment choices they do and what influence acy campaigns together with mul- �nancial literacy can have in this decision process can help identify interventions that can both improve consumer protection and support the development of tiple other interventions to better capital markets by improving ef�ciency through better investment decisions at the understand household investment individual level. behavior. While being applied BM&FBOVESPA has developed an online stock market simulator that will be used as in the Brazilian stock market, the an environment to test barriers to entry for stock market participation while the “home lessons learned will help illumi- broker� software provided by stock brokers in Brazil (which closely resemble the online simulator, offering investors the opportunity to invest online) will be used to selectively nate household �nancial decision provide interventions to stock market participants to reduce investor biases. making in general and offer sug- The simulator offers people the opportunity to practice trading stocks in an gestions on how to effectively environment that replicates real-time stock data (with a short delay) and simulates use �nancial literacy campaigns . the experience of participating on an online “home broker� The simulator allows to support consumer protection a wide range of investment products and offers investors the opportunity to build portfolios from current stock options. Participants receive a startup account with and promote a new generation of �ctional Brazilian Reais in order to participate. The motivation for developing this people accessing �nancial instru- simulator tool was to overcome informational barriers to participation by exposing potential investors to the activities related to investing in the stock market without ments in emerging markets. exposing them to any real risk, and in so doing, educating potential investors through a learning-by-doing experience on how to participate in the stock market. Graduating from the simulator, the “home broker� system links investors directly to the stock market where trades can be made in real time. Targeted messages with �nancial information and suggested �nancial products will be provided to investors 38 exhibiting common trading biases. Evaluation Methodology Administrative data from BM&FBOVESPA will be used to test how �nancial literacy, in relation to costs and administrative barriers influence (i) participation in new markets (treasury bonds and stocks) and (ii) investor biases in the market by running a series of randomized controlled trials aimed at disentangling hypothesized constraints identi�ed in the literature. The online stock market simulator consisting of 120 000 potential investors will be used as a platform to measure barriers to entry by randomly (i) providing �nancial education (information constraint), (ii) varying entry and transaction costs (�nancial constraint) and (iii) offering start-up investment packages versus a menu of investment options (menu costs) to online participants interested in investing in the stock market. To measure investment biases, historic data on stock market dropouts will be analyzed to identify the most prevalent investor mistakes that cause investors to discontinue investing. This will inform the design of information interventions coupled with �nancial products aimed at mitigating biases exhibited by current stock market investors. Using real-time trade data of approximately 600 000 retail investors, we will then randomly provide �nancial education messages through the “home broker� system and suggest products that can reduce potential biases such as overcon�dence, under-diversi�cation and the disposition effect and measure how this influences investor performance. For example, to mitigate the risks associated with under-diversi�ed portfolios, evidence on the risk-return ratios of various portfolios will be presented and advice on how to invest in an Exchange Traded Fund (ETF) to ensure automatic diversi�cation. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT DOMINICAN REPUBLIC Increasing the Impact of CCTs through Financial Literacy in the Dominican Republic Purpose of Research Thematic Area This research project will evaluate the impact of �nancial literacy training Household and business �nancial offered to bene�ciaries of the Solidaridad conditional cash transfer management (CCT) program in the Dominican Republic, in conjunction with the Social Cabinet of the Government of the Dominican Republic (GCPS). Target Segment Solidaridad CCT bene�ciaries Background and Objectives Delivery Channel Solidaridad is a CCT program in the Dominican Republic started in 2004 39 Professional and peer trainings focused on promoting investments in human capital among low-income households in the Dominican Republic (DR). Bene�ciaries receive monthly Evaluation Team transfers to debit cards that can be used to purchase basic goods in qualifying stores (colmados). The program currently serves approximately Xavier Giné (World Bank), Dean 25% of the national population. Households de�ned as poor type Karlan (Innovations for Poverty I (extreme poverty) and poor type II (moderate poverty) are eligible to Action), Greg Fischer (Innovations participate in Solidaridad. Poverty levels are assessed by a countrywide for Poverty Action) census that uses seventeen indicators, including living conditions, employment status, and education levels for household members. Partners Evaluations of CCT programs such as Oportunidades in Mexico and The Social Cabinet of the Familias en Accion in Colombia have shown that such programs are Government of the Dominican successful in increasing usage of healthcare and education services. Republic (GCPS),The World Bank, However, even with improved health and education outcomes, it Innovation for Poverty Action (IPA) may remain dif�cult for CCT bene�ciaries to manage their household �nances, �nd stable employment, or start pro�table businesses. All of Timeline these problems affect bene�ciaries’ ability to graduate from the CCT May 2012 – December 2012 program and achieve a suf�cient level of economic stability on their own. Previous research conducted by the DR government has shown that Solidaridad bene�ciaries tend to have low levels of �nancial literacy and little access to �nancial services. In this context, Solidaridad is piloting a series of projects to increase �nancial literacy and access to credit and savings to improve the income The Russia Trust Fund for Financial Literacy and Education generating opportunities of current bene�ciaries. This �nancial literacy was established in October 2008 to support the advancement program targets the heads of households in Solidaridad with a moderate of �nancial literacy and capability, through the development poverty level, of which 40% are employed at least part time and over of methods and best practices for the assessment of �nancial 85% of which have at least primary school education (SIUBEN 2005). capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part IPA plans to evaluate the impact of a package of �nancial training by the Organization for Economic Co-operation and Development. services upon bene�ciaries’ economic wellbeing. The package includes training in household and business �nancial management, job skills, and access Cluster Level Interventions to �nancial products. We expect the training program to have short-term effects on bene�ciaries’ behavior, and long-term effects on their overall welfare. Layer 1: Professional vs. peer trainers. Of the 120 nucleos in the treatment group, 60 will receive In the short term, we expect the training program to impact bene�ciaries’ �nancial literacy training from professional knowledge of household �nancial management, job skills, and business trainers, and 60 from peer trainers. This will management. This change in knowledge should also drive changes in behavior. allow us to test whether being trained by peers We expect bene�ciaries will improve in the management of household and is more or less effective than being trained by business �nances, increase budgeting and record keeping, reduce consumption professional trainers. of temptation purchase, and increase savings. We also expect the training program to increase bene�ciaries’ use of �nancial services, such as savings and Layer 2: Business training vs. job skills training. lending institutions. Additionally, we expect the job skills training to increase Half of the nucleos in both the professional and knowledge on how to look for employment, improve bene�ciaries’ ability to peer training groups will receive an additional �nd a job, demonstrate how to be a better employee and ultimately impact training session on �nancial management for employment status. In the long term, we expect that improved household businesses. The other half of the nucleos in each �nancial management, ability to successfully apply for paid employment, and group will receive additional training on job skills capacity to manage their own businesses will lead to increased income levels and the job search process. This allows us to test for bene�ciaries. whether receiving microbusiness training leads to improved business outcomes and whether receiving soft job skills training decreases Evaluation Methodology unemployment of bene�ciaries – and to compare the economic outcomes of bene�ciaries who IPA will use a randomized controlled trial (RCT) design to evaluate the impact received each type of training. of the training programs. In an RCT, one group of bene�ciaries is randomly selected to participate in the training (the “treatment� group), while another Individual Level Interventions is randomly selected not to receive the training at the time of the evaluation (the “control� group). Randomly assigning bene�ciaries to treatment or control Layer 3: Budgeting notebooks. Within each allows us to ensure that any differences observed between the two groups nucleo in the treatment group, a random subset after the training program is over can be directly attributed to the training. of bene�ciaries will be selected to receive notebooks that can be used to keep household Treatment Assignment (See Figure 1) or business budgets. Of the sample of 3,600 individuals across all nucleos, a subset to be Solidaridad bene�ciaries are assigned to a group (nucleo) of 45 to 60 members. determined will receive notebooks. This will The study will include 240 nucleos in the greater Santo Domingo area in the allow us to test whether receiving a notebook to 40 Dominican Republic with a total of 3,600 individuals. Nucleos will be selected aid with the implementation of �nancial literacy from administrative data and randomly assigned to either the treatment or training increases the impact of the training. control group. The treatment group will include 120 nucleos and the control will have 120. Layer 4: Credit access. Within each nucleo in the treatment and the control group, bene�ciaries All members of the treatment group will receive �nancial literacy training who own businesses or are interested in starting intended to improve household �nancial management skills. Within this group, businesses will be evaluated to determine their however, the treatment design is broken up into two layers of sub-treatments suitability for receiving credit from micro�nance which will be assigned at the nucleo level (training methodology and training banks. Of the suitable bene�ciaries, a random content) and two which will be assigned at the individual level (budget subset will be selected to have their names notebooks and credit/savings access). passed on to banks, so that the banks might be able to offer them credit. The number of Figure 1 Total Sample Size bene�ciaries who may be eligible for credit 240 clusters will be determined after the baseline survey is complete. This will allow us to test whether increased access to credit for self-employed Financial Literacy Control bene�ciaries increases business outcomes. 120 120 However, we are not certain there will be a large enough number of bene�ciaries that are both Peers Professional interested and eligible for credit. Because of 60 60 this, we will decide if we have a large enough sample after the baseline is completed. Business Job Skills Business Job Skills 30 30 30 30 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT INDIA Impact Evaluation of the FINO Financial Education Program Purpose of Research This study is aimed at measuring the impact of the FINO �nancial Thematic Area education program on the participants’ knowledge, attitudes and behavior Day-to-day personal �nancial related to personal �nancial management and overall �nancial well-being. management Target Segment Low-income households in Uttar Background and Objectives Pradesh primarily employed in the In partnership with the World Bank evaluation team, FINO, a doorstep informal economy banking �rm, has developed and implemented a pilot �nancial education program to support the increased use of FINO’s smart 41 Delivery Channel card as a mechanism to encourage and facilitate saving. The �nancial Formal �nancial education (one-to- education program focuses on teaching the knowledge and skills one and classroom based) required to adopt good money management practices for budgeting, spending, and saving. Participants in the �nancial education program Evaluation Team are expect to be equipped with the information and tools necessary to Leopold Sarr (World Bank), make better �nancial choices, work towards their �nancial goals, and Dr. Nathan Fiala (German Institute enhance their economic well-being. for Economic Research), The pilot program, which targets low-income consumers working in Dr. Santadarshan Sadhu the informal economy, seeks to promote saving, increase active use (Institute for Financial Management of the FINO card to access saving products, and strengthen general Research – Centre for Micro�nance) awareness of formal �nancial services and products. Over a period Partners of time, gradual positive changes in �nancial knowledge, skills and behaviors are expected to lead to changes in the �nancial well-being of Financial Information Network & program bene�ciaries. Operations, Ltd (FINO), Institute for Financial Management and One of the innovations of this �nancial education program is that it will Research-Centre for Micro Finance examine how people interact with the technology that facilitates their (IFMR, CMF) access to bank services, in this case, the FINO smart card and the Point of Transaction (POT) of the Business Correspondents (BCs). Research Timeline shows that one of the barriers to the uptake and use of branchless November 2010 – June 2012 banking in many developing countries is the low levels of familiarity and trust with the technology behind electronic cards and mobile phone banking among the poor. The FINO �nancial education program will The Russia Trust Fund for Financial Literacy and Education therefore seek to add to individual’s �nancial literacy as well as address was established in October 2008 to support the advancement the gaps in the operational knowledge and skills in conducting smart of �nancial literacy and capability, through the development card transactions, as well as strengthen trust. of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. A 5-day Master Training of Trainers (TOT) workshop and consultant The Trust Fund is managed by the World Bank implemented in part trainers were trained initially on �nancial education. The Master by the Organization for Economic Co-operation and Development. Trainers oversaw the entire training program while the consultant trainers delivered the training to the BCs for two days. This was intended to strengthen the BCs’ capacity to promote saving and to educate bene�ciaries on the value Policy Impact proposition of the FINO card as a tool to help them to more effectively manage their money. Then, a two-day �nancial education workshop was held by the The impact evaluation of the FINO BCs for the bene�ciaries. program is aimed at developing a more precise understanding of the �nancial literacy levels of the low- Evaluation Methodology income populations of India and the This evaluation utilizes a randomized treatment and control identi�cation extent to which that is a barrier and/ strategy. For the purposes of this study, the treatment group is de�ned as those individuals living in villages where the training program will be implemented or facilitator of their personal �nan- in the coming months, while the control group is villages which will receive cial management in conjunction training during a later phase of the program, after the last follow-up survey has with door step banking, which is an been conducted. If the treatment and control are balanced at baseline, then differences at follow-up for key indicators can be attributed to the intervention(s), extended component of the interven- rather than to some pre-existing difference between the two groups. A lottery tion to provide access to banking ser- system was used to select which trained BCs provided �nancial literacy training vices to a population that otherwise in phase one and which will provide training in phase two. would be unbanked. The results are As it is not feasible to interview all of the individuals in a village, a selection expected to guide researchers and of 15 individuals in each village that have a smart card were randomly made in both the treatment and control villages. These individuals were then given policy makers on whether providing a baseline survey of approximately 30 minutes each to understand their education in conjunction with provid- current knowledge of �nancial tools and their current �nancial behaviors. The ing access to banking services, all baseline was completed and the follow-up survey will be conducted in early March 2012. developed in an environment that is familiar and comfortable for the participants, is an effective approach in improving savings and inducing sound �nancial management among 42 the unbanked poor. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT INDIA The Role of Financial Access, Knowledge and Service Delivery in Savings Behavior and Welfare Purpose of Research Thematic Area This evaluation will focus on three primary areas of interest. The �rst Savings and budgeting; day-to-day involves how the expansion of the doorstep banking program offered money management; selection of by EKO, a micro�nance institution in India, will impact levels of �nancial �nancial products inclusion for low-income families in Bihar. The second will explore the impact of �nancial literacy training on the �nancial management of Target Segment these families. The third will evaluate different ways of delivering this Low-income households in Bihar training and look for the optimal method to deliver �nancial education. State in India 43 Delivery Channel Traditional �nancial capability training Background and Objectives sessions; special training with An evaluation of a �nancial education and inclusion program will be household �nancial decision makers; conducted jointly with EKO India Financial Services, a micro�nance reminders and follow-up visits institution in India. The objective of this research is to assess the extent to which the intervention can positively impact individuals’ knowledge Evaluation Team skills and attitudes, related to �nancial management. Most importantly, the research considers whether it, in fact, leads to a behavioral change. Leopold Sarr (World Bank), It will also consider how various delivery mechanisms for �nancial Dr. Nathan Fiala (German Institute for education impact people’s learning. Finally, it aims to explore the link Economic Research), between �nancial capability enhancement and �nancial inclusion. This Dr. Santadarshan Sadhu (Institute for is all based on the expectation that more �nancially capable consumers Financial Management Research) should, in theory, be able to increase the number of interactions they have with formal �nancial services and products. Partners EKO India Financial Services EKO is one of the few institutions in the state of Bihar to offer “doorstep� banking and �nancial services. Most of the villages in Timeline which EKO operates have little or no access to banks. The problem of access to banking is further exacerbated by the fact that the people March 2011 – March 2012 with whom EKO works come from very low-income families and thus face numerous dif�culties in opening savings accounts. To solve the access problem, EKO provides a savings program in these villages, which allows people to conduct �nancial transactions within their The Russia Trust Fund for Financial Literacy and Education village and close to their homes, instead of having to traveling long was established in October 2008 to support the advancement distances to access a bank. EKO’s delivery mechanism is unique in the of �nancial literacy and capability, through the development sense that the �rm’s agents, known as the Customer Service Providers of methods and best practices for the assessment of �nancial (CSPs), carry out banking transactions on the premises of their primary capability and for the evaluation of �nancial education programs. occupation. Most of the CSPs facilitate the banking transactions of the The Trust Fund is managed by the World Bank implemented in part clients along with their regular business. Thus, the revenues generated by the Organization for Economic Co-operation and Development. by the CSPs are basically supplementary income, rather than being primary income, which makes it a viable business option for them. Policy Impact The �nancial literacy material to be used in the training of the treatment group focuses mainly on the importance and bene�ts of savings. The savings training From the �nancial inclusion aspect module, designed by Freedom from Hunger has been adapted to suit the of this project, which promotes conditions of rural Bihar. The training includes ten sessions that run over 12 weeks (including a two week buffer period). The sessions will cover topics on increased access to doorstep banking, savings ranging from “what are savings?� to “how to build an emergency fund� . researchers expect to see an increase Trainers will continue to visit and monitor participants a month after the training in the number of �nancial transactions is completed. customers undertake. Meanwhile, the �nancial education training compo- nent will help shed light on the mecha- Evaluation Methodology nisms through which intra-household The research methodology for this evaluation is a randomized control trial (RCT) with process evaluation. All participating villages in the evaluation will �nancial decisions are made. Finally, be those in the state of Bihar that have not yet been exposed to the EKO researchers will explore the potential program. These villages will then be randomly assigned to either the treatment group (expansion program) or the control group. A baseline listing survey of all impact that intensive �nancial educa- people in both treatment and control villages will be conducted to collect basic tion training has on influencing �nan- information on the participants from both villages, as well as ensure the sample cial behaviors and wellbeing of the is randomized and well balanced. participating individuals by compar- EKO will then roll out its “doorstep� banking program in a group of randomly ing treatment with control groups. selected villages. Financial literacy training will then be organized also in a random selection of villages. Within the respective villages, a set of households will be randomly selected to receive the �nancial literacy program with phone and in-person follow-ups. Another group will receive the same training, however it will be administered to multiple members of the same household, with a particular focus on members who are influential in household decision-making. One year after the baseline survey, an endline survey will be conducted on a random selection of people in the villages. The speci�c indicators to be 44 measured are changes in �nancial literacy knowledge (i.e. knowledge about the value of savings); behaviors; and changes in welfare, as measured by consumption and productive assets. Along with the extensive quantitative research as explained above there will also be a process evaluation component which will involve qualitative surveys aiming to disentangle the mechanism through which attitudes and behaviors are shaped by the �nancial literacy intervention. The qualitative research will involve face-to-face interviews and focus groups. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT INDIA AND KENYA Social Networks, Financial Literacy and Index Insurance in India and Kenya Purpose of Research This research project examines the role of �nancial literacy on farmers’ decision to purchase index-based weather insurance. It also studies the Thematic Area social network spillovers of �nancial literacy provision in Kenya and India. Long-term planning and selection of �nancial products Target Segment Background and Objectives Rural, small-scale farmers In recent years, a range of initiatives have sought to develop insurance products to help rural households mitigate exposure to weather risk in Delivery Channel many developing countries such as Ethiopia, Malawi, Thailand, Kenya, 45 Comic Books India, Nicaragua, and Mongolia. Regarded as a promising alternative to traditional crop insurance, market participants, NGOs, donors, and Evaluation Team governments are all testing the applicability of index based weather risk management instruments in different contexts. Xavier Giné (World Bank), Dean Karlan (Innovations for A signi�cant advantage of this type of product is that payouts can be Poverty Action), Muthoni Ngatia calculated and disbursed quickly and automatically without the need for (Innovations for Poverty Action) households to formally �le a claim because they are based on measured rainfall. This, in turn, reduces transaction costs, which would otherwise Partners tend to drive up the price of the insurance. Fast payouts are also likely to be valued by policyholders in an environment where households are poor FSD-Kenya; Equity Bank; APA and often liquidity-constrained. A second advantage is that the product Insurance; Ruiri and Ntima Coffee is free of adverse selection and moral hazard problems that often plague Farmer Cooperative Societies; insurance markets. This is because payouts are based only on publicly- ICICI-Lombard observed data, rather than private information about the bene�ciary. Timeline Demand for index-insurance products has, however, been low. A variety March 2011 – March 2012 of factors may discourage participation, including household credit constraints and limited trust in the insurance provider. A potentially important reason for the low demand for index insurance products is limited understanding of the product amongst potential consumers. The event that triggers a payout of an index insurance products depends on complicated triggers and the calculation of the index itself that may prove particularly dif�cult to explain, compared to well known event The Russia Trust Fund for Financial Literacy and Education insured under standard insurance. was established in October 2008 to support the advancement of �nancial literacy and capability, through the development There has been limited empirical evidence to date on the importance and of methods and best practices for the assessment of �nancial impact of the �nancial literacy efforts on farmer understanding and take- capability and for the evaluation of �nancial education programs. up. When participation is voluntary, the presentation and explanation The Trust Fund is managed by the World Bank implemented in part of a product is the primary method of encouraging participation. As a by the Organization for Economic Co-operation and Development. result, the information provided and the means to convey it are critical for the success of the program. In addition, spillovers from information dissemination are likely to occur and should be measured when assessing the ef�cacy of Policy Impact �nancial literacy efforts. Initial evidence from Kenya suggests The goal of this study is to examine the role of �nancial literacy on farmers’ that social networks play an important decisions to purchase index-based weather insurance and to examine social network spillovers of �nancial literacy provision in Kenya and India. The role in farmers’ decisions to purchase researchers propose to benchmark these effects with the effect of providing drought insurance. The study pres- discount vouchers off the price of insurance. ents suggestive evidence that �nan- The study uses comics as the central �nancial literacy delivery mechanism. The cial literacy materials are ef�cacious comic used in this study describes a family which had faced a drought in the in encouraging take-up when farm- previous season. The drought had adverse effects on their savings and well- being. The comic carefully details the index-insurance product and shows how ers’ social contacts similarly receive it can help the family protect themselves from the risk of drought. The comic access to �nancial literacy materials. presents the insurance product in an accessible and relevant manner and has sustainability as an educational tool as it can further be shared with other farmers. Evaluation Methodology The study employs a randomized controlled trial (RCT) to determine which type of �nancial literacy is most effective. An initial phase of the study was carried out with coffee farmers in Ruiri and Ntima, two drought-prone areas in Eastern Kenya in early 2011. This study will be replicated in Rajasthan, India in 2012. In Kenya, researchers worked in collaboration with Ruiri and Ntima Coffee Farmer Cooperative Societies with whom the insurance company had partnered to offer drought insurance to farmers. The enumeration area covered 14 villages: 12 villages in Ruiri and two villages in Ntima. Interviewers �rst visited all coffee growing households within the enumeration area and collected some basic information about the 46 household: their GPS coordinates, farm size, number of coffee trees and contact information. Following the census, households were grouped together into clusters based on geographic proximity. Each household had a circle with a 60 meter radius drawn around their GPS marker. All overlapping circles became part of the same cluster. Households that were isolated were added to the nearest cluster. From 1117 households, 196 clusters were thus formed. Enumerators then revisited the households and administered a baseline survey. Two interventions were tested: a comic on index insurance and discount vouchers for the purchase of insurance. At the cluster level treatment intensity was randomized and orthogonal across treatments. This design allowed for an accurate assessment of spillover effects of �nancial literacy compared to spillover effects of higher participation from the discount vouchers. After completing the baseline survey, enumerators administered the interventions. Each household was randomly assigned to receive a comic or not depending on what comic intensity the household’s cluster was assigned. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT KENYA Comic FX: Can Cartoons Improve the Effectiveness of Financial Education? Purpose of Research This research project seeks to identify the effectiveness of alternative modes of delivering �nancial education to high school students. It Thematic Area compares the standard delivery of a structured course of materials with Savings, �nancial management a series of weekly comic book episodes that personalize, contextualize, and make pertinent to the target population the lessons of the course. Target Segment Kenyan high school students Background and Objectives Delivery Channel Kenyan youth face an uncertain and volatile �nancial landscape, with Classroom instruction vs high un- and under-employment, questionable long-term job prospects, 47 comic strips and little or no protection against the vagaries of ill health and injury. However on the other hand, young Kenyans have more opportunities to Evaluation Team invest in and plan for their futures than perhaps any earlier generation: market liberalization and a stable macro-economy have facilitated Nada Eissa, James Habyarimana, steady growth in recent years, educational and training options abound, William Jack, of Georgetown and access to �nancial services has expanded considerably.1 It is University essential to give this population the tools and �nancial capability to grasp these opportunities. Partners Well Told Story, Junior Achievement The project will evaluate a �nancial capability project that seeks not Kenya, Digital Divide Data only to inform young Kenyans about �nancial facts and options, but also to induce them to change the way they think about and make �nancial decisions. In a context in which it is often dif�cult to see beyond the Timeline short-term, the authors focus on encouraging long-term planning as it January 2012 – December 2012 relates to investment in education, training and small business creation, and precautionary �nancial behavior such as saving, asset allocation, and insurance against uncertain events. The study is being implemented by two Kenyan organizations that have sought to help young Kenyans prepare themselves for economic opportunities through innovative means. Well Told Story (WTS) is a for-pro�t company that publishes a monthly comic book called Shujaaz, read by upwards of 600,000 young people The Russia Trust Fund for Financial Literacy and Education across the country, as well as a radio program broadcast on 30 radio was established in October 2008 to support the advancement stations nationwide, featuring the lead character of the comic book. of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial Junior Achievement Kenya (JAK), is a local NGO engaged with more capability and for the evaluation of �nancial education programs. than 200 high-school based youth clubs and associations through which The Trust Fund is managed by the World Bank implemented in part a range of educational and behavioral interventions are conducted. by the Organization for Economic Co-operation and Development. 1 FSD, FinAccess (2010). Through JAK’s clubs, students will receive �nancial literacy materials both through standard classroom delivery mechanisms, and as part of the Shujaaz storyline. Regular Shujaaz characters – both heroes and villains – with whom Policy Impact readers are already familiar, will serve as role models or anti-role models, accordingly. The objective is thus to assess the extent to which delivering In order to take full advantage of a �nancial education materials through popular media has different effects – be relatively stable and growing mac- they different in nature and scope, larger in size, or more permanent – than more traditional methods of delivery. roeconomy, young Kenyans must be given the tools to exploit the new To augment this exercise, the authors will also investigate the role of broadcast opportunities they face. Financial media in augmenting the effects of the interventions, by developing radio clips with �nancial education material that will be played on CD players provided innovations in particular, including to schools. These will supplement the radio shows that are currently aired in the expansion of banking services, conjunction with the regular Shujaaz comic, featuring the lead characters. mobile money, and insurance offer great potential, but also risks. Equip- Evaluation Methodology ping the next generation with the The authors combine a randomized control trial with quasi-experimental skills to navigate this more sophisti- quantitative techniques to assess the impact of each of the interventions, cated �nancial landscape should be compared with no program, as well as vis-à-vis each other. The interventions are randomized across Junior Achievement Clubs: part of a general education strategy aimed at empowering youth. • about one quarter of the clubs receive a �nancial literacy program developed by JAK’s parent organization, Junior Achievement, which has been used in This study will also provide more other countries outside Kenya; widely applicable insights into how • about a quarter of the clubs will receive the Shujaaz comics on a weekly basis, with stories matching the content of the in-class material; best to reach and to teach young peo- • about a quarter of the clubs will receive the regular Shujaaz comic strip, ple, and to empower them to make which will act as a placebo intervention; and responsible long-term decisions. To • about a quarter of the clubs will receive no speci�c intervention, acting as the extent that sound investment, a control group. careful debt management, and pru- 48 Half of each of the two groups that receive the comics will also receive the dent saving and business decisions CD intervention. are part of a successful transition over Since one of the objectives of the regular comic is to induce responsible the longer term to middle income behavior change in young adults through the portrayal of both positive and country status, understanding the negative role models, it is imperative to include a placebo group in the design. This group is exposed to the same role models, and is exposed to information power of mass communication chan- that could be useful in everyday life, but is not exposed to the speci�c �nancial nels in inducing behavior change is education material. of interest to policy makers in East In order to measure the impact of the interventions, the authors will collect Africa and beyond. data at baseline, before the programs, and at endline after their completion. As well as administering a survey to about 5,000 students, they will engineer a situation in which a non-trivial �nancial decision must be made. First, both at baseline and endline all students are asked how they would allocate about $25 of unexpected earnings between cash, a bank account, and a mutual fund on the Nairobi stock exchange. Subsequently, authors deliver prizes of this amount to each of roughly 2,000 students, again at baseline and endline. The hope is to discern changes in both the stated and the actual allocations, between baseline and endline in the treatment groups. This will provide a unique outcome measure of students’ actual �nancial behavior, to complement the more qualitative data we collect in interviews. Finally, a series of behavioral games will be conducted to measure differences in attitudes and preferences, for example related to discounting, patience, and long-term perspectives across students exposed to the different treatments. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT MALAWI Financial Development and the Psychology of Savings: Field Experiments in Rural Malawi Purpose of Research Thematic Area This project investigates innovative ways to address low levels of formal Savings and �nancial knowledge savings in developing countries. The study explores to what extent psychological mechanisms can be leveraged to increase formal savings. Target Segment In particular, the study determines whether direct deposit of wages – Low-income agricultural wage as opposed to receiving cash– can help individuals to match desired savings and expenditure patterns with actual behavior. In addition, the earners and smallholder farmers study explores how a combination of formal �nancial products and training can help to activate mental accounting to facilitate savings. Delivery Channel 49 One-on-one sessions at participants’ homes (�nancial knowledge), direct transfers of Background and Objectives wages into bank accounts, offers of The study is motivated by market failures and behavioral phenomena new labeled accounts that limit the ability of poor households to save and invest in productive activities. Incomplete credits markets in developing countries force Evaluation Team many to rely heavily on accumulating personal savings to �nance Xavier Gine (World Bank) investments. However, behavioral phenomena, such as self-control problems, may lead to suboptimal savings and under-investment Partners relative to desired levels. Opportunity Bank of Malawi and The study has two parts: In the �rst part, the study evaluates the Lujeri Tea Estates introduction of a new direct deposit system at large agricultural �rm in Malawi that pays workers’ wages into individual accounts instead Timeline of paying in cash. Receiving wage or farm revenues in the form of February 2012 – March 2013 cash may exacerbate self-control problems, since the temptation to spend may be higher when cash is on hand. If wage or farm revenues are sent directly into a bank account, on the other hand, temptation problems may be reduced, raising savings levels. Drawing on surveys and administrative data from bank records and the tea estate the study seeks to measure changes in the consumption patterns –timing and allocation across expenditure categories– as well as formal and informal savings and borrowing behavior. The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement In the second part, the project looks to study mental accounting and of �nancial literacy and capability, through the development the role of labeled bank accounts to activate such mental accounting. of methods and best practices for the assessment of �nancial Mental accounting describes the phenomenon that money is not capability and for the evaluation of �nancial education programs. always fungible across expenditure categories. Savings that are The Trust Fund is managed by the World Bank implemented in part mentally assigned to a speci�c savings goal or set of expenditures, by the Organization for Economic Co-operation and Development. like inputs for farming or children’s school fees, may become less available for expenditures for another set of expenditures, like unplanned purchases. Offerings savings accounts that stress such mental separation by being labeled Policy Impact for speci�c expenditure categories such could therefore help individuals achieve their savings goals. The study seeks to measure impacts on �nancial behavior This project will study changes expen- and household wellbeing through combination of survey and administrative diture and savings patterns of agricul- records. Additional behavioral experiments will help to disentangle competing theories about effect channels and the nature of mental accounting. tural wage workers and smallholder farmers in Malawi in response to An important question is what role �nancial education can play in complementing offers of �nancial products designed or even as an alternative to dedicated �nancial products designed to leverage behavioral phenomena. As part of this study we will test whether targeted to leverage behavioral mechanisms. �nancial training can help to bolster the effects of the novel �nancial products By measuring the impacts of different introduced here, and to what extent such training can substitute for such products for example by highlighting directly the bene�ts of mentally separating novel �nancial products and offering expenditures and setting savings goals. complementary �nancial training, we hope to determine the potential of the products’ potential to increase Evaluation Methodology savings, and to understand the role The evaluation utilizes a randomized controlled trial approach. As described of �nancial literacy in both amplify- above, the project consists of two parts: the �rst evaluating the impact of ing impacts and as alternatives to the switching from cash to direct deposit, the second exploring the usefulness of labeled accounts to activate mental accounting In the �rst part, working with a same �nancial products. population of wage earners at a large agricultural �rm in Malawi, a new direct deposit system of wage payments will be introduced using a random phase- in. A subset of workers will be switched to the new system earlier while the remainder serves as a control group to be able to compare �nancial behavior and well-being between the two groups. A complementary, cross-randomized roll-out of training sessions with individual households will evaluate whether impacts can be ampli�ed through �nancial literacy training, and whether impacts of the introduction of direct deposit for wages can be emulated with such training. 50 The second part of the study works with a population of smallholder tea farmers with regular revenues from a single buyer which are directly deposited into farmers accounts. Participants will be randomly allocated in one of several treatment groups with varying offers of savings products. The offers include a plain, second savings account, a second account labeled for speci�c expenditure categories based on participants savings goals and additional commitment and default features that enable farmers to pre-specify automatic regular deductions from sales for savings in a separate account. In both parts of the study all participants will be interviewed before and after the interventions to measure any changes in savings and consumption behavior and other �nancial behavior amongst the different treatment groups. In addition, bank records will track changes in the use of �nancial products by study participants both pre and post intervention. Knowledge about impacts of the products introduced as part of this project will help inform policy in several ways. Signi�cant impacts of a direct deposit system for wages would lend further support for �nancial development in developing countries which would facilitate private �rms moving from cash payments to direct deposit. Furthermore, understanding behavioral phenomena and how they relate to �nancial decision-making can inform the design of products that leverage such psychological mechanisms to make it easier for poor households that need to rely heavily on personal savings for both their consumption smoothing and investments needs. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT MEXICO Evaluating the Effectiveness of Mexico’s Credit Disclosure Reforms on Consumer Understanding and Financial Decision-Making Thematic Area Savings and credit Purpose of Research This project explores the interaction between pricing transparency and Target Segment �nancial capability. It looks to measure whether �nancial capability and consumer decision-making can be improved through policy Low-income consumers of credit/ interventions more typically viewed as serving consumer protection basic savings products purposes. With the often high cost of traditional curriculum-based �nancial education programs, �nding regulatory approaches to improve Delivery Channel �nancial capability through market conduct could possibly introduce 51 Government-mandated product alternative policy options for resource-constrained governments disclosure formats; sms and interested in �nancial capability. telephone counseling Evaluation Team Xavier Gine (World Bank), Background and Objectives Rafe Mazer (World Bank), An important question in �nancial capability policy is whether Cristina Martinez Cuellar consumers learn �nancial lessons more easily through school-based traditional methods (such as in the classroom, in training seminars (World Bank) and in workshops, etc), or by accessing clear and easy to understand information in the marketplace. In Mexico, this question is of increasing Partners importance for a number of reasons: increased use of �nancial products, National commission for the particularly amongst low-income consumers; evidence of repayment defense and protection of �nancial problems and hidden charges in some credit and savings products; service users (CONDUSEF) and many types of regulated and non-regulated �nancial institutions, making oversight and market monitoring a constant challenge. Timeline This study explores the intersection between consumer protection December 2011 – January 2013 and �nancial capability by comparing how product-speci�c information and broader �nancial capability messages impact short and medium- term �nancial decision-making of the low-income consumers in Mexico for basic credit and savings products. It will also evaluate the impact of several different channels for communicating important The Russia Trust Fund for Financial Literacy and Education �nancial messages to consumers, including: point-of-sale disclosure; was established in October 2008 to support the advancement information to facilitate comparison shopping; �nancial counseling by of �nancial literacy and capability, through the development phone; and periodic �nancial capability SMS messages. of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. Modi�cations to loan disclosure forms and two other CONDUSEF- The Trust Fund is managed by the World Bank implemented in part developed �nancial literacy tools is the primary interventions against by the Organization for Economic Co-operation and Development. which consumer behavior and lender practices are measured, and against which the 2009 reforms to disclosure regulations are evaluated for effectiveness. The study will seek to measure how improved disclosure and pricing transparency impacts the immediate decision to accept or not accept Policy Impact the product offered, awareness of the true costs of products, understanding of key �nancial information and terms, and how such awareness could impact This study will compare changes in consumers’ thinking about the �nancial products they use, and in turn their low-income consumers’ �nancial deci- �nancial decision-making in the long-run. Several other components will be used to inform and complement the study: Mystery shopping to determine sion-making and capabilities through existing practices of disclosure and �nancial advice in the consumer credit both product-speci�c information and market; testing and modi�cation of draft disclosure formats through consumer broader �nancial capability messages. focus groups; and a lab experiment where a consumer shopping experience is simulated with similar products that use different methods of disclosure to By measuring the impacts that differ- determine which methods are most salient and informative to consumers. ent product disclosure methods and light-touch �nancial education inter- ventions have on �nancial capabil- Evaluation Methodology ity and decision-making, we hope to The evaluation utilizes a randomized controlled trial approach. As described determine the effectiveness of exist- above, the program to be evaluated consists of several different interventions ing transparency and �nancial educa- focused on product transparency and �nancial education targeting low-income and �nancially-active consumers in four cities within Estado de Mexico, a state tion policies in influencing �nancial bordering the capital, Mexico Distrito Federal. The interventions will compare decision-making, and offer alterna- existing policies with proposed alternatives so that the study can both inform, tives to resource-constrained govern- and facilitate, further policy improvements. These interventions include: ments looking to improve �nancial • A control group that will use information provided by �nancial institutions capability even if unable to implement under the 2009 Transparency Law to make their �nancial-decisions. curriculum-based approaches. • Provision of a modi�ed disclosure form that offers comparative information across several providers as an enhancement of the existing disclosure format • Varying incentives to encourage consumers to consult CONDUSEF’s existing �nancial consumer hotline for one-on-one counseling services • Monthly SMS communication from CONDUSEF to consumers with basic 52 �nancial capability messages Through the use of these four different interventions the study will not only explore the �nancial capability impact of improved disclosure, but also measure the relative impact of different channels in delivering �nancial capability messages and product information to low-income �nancial consumers. To measure changes in knowledge, �nancial behavior and decision-making amongst the different treatment groups, all participants will complete �nancial capability surveys prior to and one year after the intervention to measure any changes in �nancial capability and understanding of key product terms and features. In addition, the �nancial products used by consumers both pre and post-intervention will be recorded to measure changes in behavior. The primary theory being tested in this evaluation is that consumers not only improve �nancial capability through traditional educational approaches, but that they also can improve �nancial capability through better information in the marketplace. We know that consumers learn through experience, but in the case of �nancial products this leaves them vulnerable to learning from negative experiences that have real economic impacts on their lives. So is it possible to improve �nancial capability during the shopping experience, before the consumer has signed onto an inappropriate �nancial product? To accomplish this we will use both a �nancial capability test and review changes in actual consumer �nancial behaviors via credit reports that track post-intervention �nancial product usage to several years prior. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT NIGERIA Money and Popcorn: Can Entertainment Media be used to Improve Financial Capabilities in Nigeria? Purpose of Research Thematic Area Microenterprise owners often lack access to formal education. Using Savings and credit and general popular culture and entertainment as an alternative delivery mechanism �nancial planning to promote �nancial literacy is a potentially powerful means of facilitating access to �nancial information. This research project evaluates the Target Segment extent to which a feature �lm produced through the Nigerian Film Industry (Nollywood) can increase �nancial capabilities of households Microenterprise owners by promoting responsible borrowing and saving strategies. Delivery Channel 53 Feature Film Background Evaluation Team In Nigeria, households and entrepreneurs lack appropriate information about, and access to, credit and other �nancial services. Although Giuseppe Iarossi (World Bank), improvements have been registered in the last three years, 46% of Florentina Mulaj (World Bank), the population remains �nancially excluded, with no access to formal Vincenzo Di Maro (World Bank), or informal �nancial services. This compares unfavorably to countries Aidan Coville (World Bank), such as Kenya and Botswana (33%), while in South Africa only one Billy Jack (Georgetown University) quarter of the population is �nancially excluded. This lack of access is not the product of a lack of interest or demand. Almost 75% of the Partners unbanked population in Nigeria would like to have a bank account and over 80% of the population gets �nancial advice from family Credit Awareness and friends. Furthermore, the great majority of Nigerians would like to know how to save for their old age (63%), how to save regularly Timeline (60%), how to open a bank account (55%), and how interest rates September 2011 – February 2013 are calculated (52%).1 This evidence underscores the size of the �nancially underserved population and some of the factors that deter Nigerians from using formal �nancial institutions. It also points to their desire and need for �nancial education. In countries with diverse social and economic pro�les like Nigeria, �nancial literacy is particularly relevant for microenterprise owners as they are vulnerable to persistent �nancial pressures. Building �nancial capability in Nigeria represents a considerable step in helping consumers to acquire the skills and knowledge to be capable, con�dent, and self-reliant when The Russia Trust Fund for Financial Literacy and Education making �nancial decisions. There is, however, a lack of evidence on was established in October 2008 to support the advancement the best means of building these capabilities in consumers. This of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. 1 EFInA, Enhancing Financial Innovation & Access, 2010 National Survey The Trust Fund is managed by the World Bank implemented in part http://www.efina.org.ng/our-work/research/access-to-financial- by the Organization for Economic Co-operation and Development. services-in-nigeria-survey/ research evaluates the effectiveness of improving �nancial capabilities through entertainment (in this case a feature length �lm). It looks to provide inputs to the current debate in the �nancial capability �eld around the value and viability Policy Impact of non-classroom style, alternative delivery mechanisms. Leveraging the support of Nollywood, the initiative uses a full feature �lm “The Story of Gold� Movies have often been associated to enhance �nancial literacy in viewers. It then assesses the impact of the with negative behavioral impacts �lm on individuals’ �nancial capabilities (namely, literacy, skills, attitudes, and behavior). Essentially, it looks to test whether a popular media initiative such such as increasing violence or as this is able to: (1) increase people’s awareness and understanding of best encouraging smoking, but can they practices for day-to-day �nancial management, saving and borrowing money; also lead to positive behavioral and (2) lead to more responsible �nancial decision making. changes? Nolly-wood’s mass distri- “The Story of Gold� is a �lm produced by Nollywood and distributed by Credit bution model is an ideal platform for Awareness. Credit Awareness is a Nigerian NGO which has undertaken an active and phased national program of promoting �nancial literacy in cooperation with disseminating �nan-cial education a large number of supporters, including the Central Bank of Nigeria. The movie messages across Africa. However, it aims to teach low income entrepreneurs with limited formal education some is not clear a priori whether viewing of the core concepts around �nancial planning. The �lm revolves around the core values of “smart savings and responsible borrowing� and disseminates a movie is suf�cient to improve peo- this message through large screenings in public locations across Lagos State. ple’s �nancial capabilities. If proven Focusing on this simple behavioral goal and highlighting the repercussions of effective, �lms such as “The Story of poor �nancial decisions, “The Story of Gold� looks to leverage the popularity of Nollywood to build awareness and impact the �nancial decision-making Gold� may offer a unique opportunity processes of low-income households. to improve �nancial literacy among The movie aims to provide �nancial education that will elicit behavior changes by low-income microenterprise owners. setting these messages within emotive contexts. But emotions fade with time, and the window for action is limited. If given the opportunity to take action on their emotions in the moment, it is hypothesized that participants would make more �nancially astute decisions. If these decisions have potential long-term implications, such as opening a savings account, this action, driven by emotion linked to the movie could potentially influence participant behavior in the long run. The study will thus invite Micro�nance Institutions (MFIs) to participate in the screening event, providing the platform required for motivated movie-goers 54 to act on their motivations and open a savings account on the spot. Evaluation Methodology Markets will be selected in the outskirts of Lagos and a screening area will be chosen that will be accessible for all participants in order to maximize take up rates. An initial listing will be conducted that will entail a full census of stall owners within a 1 mile radius of the screening event. The listing will act as a mini baseline and will include a selection of key indicators of interest together with identi�cation indicators (including GPS coordinates). Once the listing has been completed, participants will be randomly selected into 4 different groups: (i) “Story of Gold� screening only, (ii) “Story of Gold� screening with MFIs opening savings accounts for viewers on the spot, (iii) A placebo movie screening (control) and (iv) the placebo screening with MFIs. Survey participants will receive color-coded wrist bands and personal invitations the day prior to the movie screening to minimize the chance of contamination and to improve participation rates. By randomly selecting who will receive which intervention, we will be able to measure the causal impact of both the movie screening and the movie screening reinforced with MFI representation to disentangle the long-term educational effects of the movie from the immediate emotive context on influencing savings and borrowing behavior. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT NIGERIA Using Media to Encourage Savings through Learning-By-Doing Purpose of Research This project explores how new media and “learning-by-doing� can encourage �nancially unsophisticated consumers to open and maintain Thematic Area savings accounts. We evaluate a promotion by a large Nigerian bank to Precautionary savings, learning-by- examine (1) how consumers respond to different types of new media doing, new media campaigns, e.g. internet and celebrity endorsements, and (2) how the experience of maintaining a savings account over a three month period Target Segment can improve �nancial literacy and change long-term precautionary Middle-income Nigerian consumers savings habits. Delivery Channel 55 Television, Radio, Billboards, Print Background and Objectives Media, YouTube, F acebook Formal savings accounts are not widely used in Nigeria, but the Evaluation Team consumer �nance sector is growing rapidly. As a large emerging market with low �nancial literacy, Nigeria can inform us about the interaction of Alex Kaufman (Federal Reserve) retail and household �nance in developing countries. Kelly Shue (University of Chicago Booth School of Business) In 2011, InterContinental Bank in Nigeria launched a nationwide Benedikt Wahler (Roland Berger) savings promotion called “I-Save I-Win� (ISIS). It featured a large Florentina Mulaj (World Bank) number of heavily publicized lottery prizes for those who opened or Martin Kanz (World Bank) maintained savings account and maintained savings balances above various threshold amounts -- N50,000 (US$320) for regional lotteries, N100,000 ($640) for the national prize -- for 90 days. Its message: Partners although the special promotion included lottery prizes, every saver is Access Bank . a “winner� ISIW was promoted with a media push including celebrity Roland Berger Strategy Consultants endorsements and media releases through Facebook and YouTube, which were staggered over a period of several months. Timeline This project will evaluate ISIW’s impact on customers’ savings habits and March 2012-February 2013 �nancial literacy. First, we explore how the experience of maintaining savings accounts during the 90-day promotional period affects savers’ long-term savings habits. Second, we investigate how different media promotions affect customer participation in the program and savings behavior: which media pushes led to more sign-ups, and to what extent The Russia Trust Fund for Financial Literacy and Education , did each lottery winner have a “demonstration effect� i.e., inspiring was established in October 2008 to support the advancement others to increase savings at her branch. of �nancial literacy and capability, through the development of methods and best practices for the assessment of �nancial capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. Evaluation Methodology We partner with Access Bank, which recently acquired InterContinental Bank. Policy Impact Our analysis uses detailed transaction-level individual savings account data. To study the intensive margin of how ISIW affected the savings activity of This research investigates the existing account-holders, we conduct a “differences-in-differences� analysis: potential for new media and sav- We compare savings account balances before and after each phase of the ISIW promotion. To control for general trends in savings activity over the same time ings promotions to increase �nan- period, which may not be due to the savings promotion, we use data from a cial literacy and to build a habit of second bank that did not have a savings promotion to create a control sample. precautionary savings. Learning by To study the extensive margin of how ISIW affected new account sign-ups doing, accompanied by new media and savings, we conduct a similar differences-in-differences analysis comparing promotions, may provide a valuable sign-ups at InterContinental Bank before and after each phase of the savings promotion, and again control for time trends using data from a second bank complement to other forms of �nan- that did not have a savings promotion. cial literacy training. In addition, we examine the effect of winning the lottery on lottery winners’ savings behavior and the savings behaviors of individuals associated with the same banking branch as the lottery winner. The lottery randomization offers a clean source of identi�cation through which we can establish a causal effect of lottery winnings on savings behavior. We will also conduct surveys to learn about changes in �nancial behaviors that are not directly observable from bank account data, such as �nancial literacy and the source of funds brought into InterContinental Bank (e.g. consumption, other banks, stored cash). 56 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT SOUTH AFRICA Evaluation of the Use of Entertainment Education to Improve Financial Capability in South Africa Purpose of Research Thematic Area This project aims to pilot entertainment education as an information delivery tool to influence sound �nancial management in South Africa. Credit and debt It will do so by assessing the usefulness of this instrument to increase �nancial capabilities of the population with a robust impact evaluation Target Segment framework. Low-income South Africans with and without existing consumer debt Background and Objectives Household debt has been growing at a startling rate in South Africa 57 Delivery Channel over the last decade, especially among low income segments of the Television Soap Opera population, posing a direct threat to the country’s economic stability. The ratio of household debt to disposable income amounted to 78.2% Evaluation Team in the third quarter of 2010.1 In September 2010, the South African Thomas Losse-Mueller (World Bank), National Credit Regulator (NCR) reported that of the 18.35 million Gunhild Berg (World Bank), credit-active consumers, 46.3% had impaired records. Over the last two years, this number has increased by an average of 6.9% per year. Bilal Zia (World Bank) The percentage of consumers who remain in arrears for three months or more stood at 17%. The South African National Debt Mediation Partners Association (NDMA), established in 2008 by the �nancial industry to National Debt Mediation provide an independent debt resolution process for the over-indebted, Association (NDMA) reports that 160,000 cases of over-indebtedness have already been registered with the National Credit Regulator (NCR). Approximately Timeline 8,000 additional cases are recorded each month. May 2011 – September 2012 The objective of this study is to assess the ability of entertainment education as a social marketing instrument to increase �nancial capabilities of the consumers in South Africa using a robust impact evaluation framework. The program aims at enhancing knowledge, attitudes and behavior regarding sound �nancial decision-making with a particular focus on managing debt. The potential gains of such a program in terms of increased awareness of debt-related problems and behavior change, such as avoiding debt and seeking help once over- indebted, are of particular relevance in South Africa given the high and The Russia Trust Fund for Financial Literacy and Education increasing levels of household debt. was established in October 2008 to support the advancement of �nancial literacy and capability, through the development The sharp increase in over-indebtedness and impaired consumer credits of methods and best practices for the assessment of �nancial in recent years in South Africa has been triggered by job losses and a capability and for the evaluation of �nancial education programs. The Trust Fund is managed by the World Bank implemented in part by the Organization for Economic Co-operation and Development. 1 South African Reserve Bank, Quarterly Bulletin, December 2010. reduction in household incomes following the recent economic downturn. It also points to inappropriate or self-prejudicial �nancial decisions taken by borrowers, as well as potentially unsound business practices by banks and other lenders. Going Policy Impact beyond literacy, over-indebtedness is also associated with attitudes and behaviors of consumers in �nancial decision making. Therefore, any measures to control the Addressing private indebtedness income spending and high debt of the South Africans must take into account the is a high priority for South Africa. need to change attitudes and behaviors, in addition to improving literacy. The National Treasury of South The project entails the content development, production, and impact evaluation of Africa is currently developing a �nancial capability storylines, which will be included in a popular South African soap national �nancial literacy strategy opera called “Scandal!� .The target individuals are low-income South Africans with and without existing consumer debt. The show is broadcasted on eTV, the second most and an action plan for its imple- popular station in South Africa. The �nancial education storyline will stretch over a mentation. The February 2011 period of three consecutive months, a time that is deemed necessary for the viewers to emotionally connect with the soap opera characters, for the events to unfold, and National Treasury Policy Docu- for the �nancial literacy messages to sink-in. The storyline for the soap opera has ment “A Safer Financial Sector to been developed by the production company of Scandal, Ochre Media, together with Serve South Africa Better� calls the NDMA, and a team of �nancial capability and social marketing experts. Financial education messages in the storyline were tested through focus groups. for strengthened market conduct regulation and better consumer education. Improving �nancial Evaluation Methodology capabilities, especially around An evaluation of the impact of a soap opera on behavior and attitudes needs to debt management, has increased respond to a number of substantial methodological challenges. First, the effect of the soap opera’s message must be separated from the potential impacts of other in importance in South Africa, a messages on similar topics from other sources to which the individuals may be result of the large and growing exposed. Second, it is necessary to identify a valid counterfactual, an estimate of possible outcome for participants in the absence of the program. Third, because the levels of household debt in the effect of the message may be confounded by the delivery channel, it is dif�cult to country. Therefore this project is establish causality between the �nancial literacy and any potential behavior change. of particular importance to the To overcome these concerns, the evaluation will deploy a random encouragement design methodology. South African authorities, who are also actively engaged with project 58 The key components of the research design include: development through an advi- • A �nancial incentive provided for half of the sample (the randomly selected sory committee. The outcomes of treatment group) to watch “Scandal!� the project in terms of increased • A �nancial incentive provided for the other half (the randomly selected control group) to watch a similar soap opera that is aired around the same time but awareness of debt-related prob- does not have a �nancial literacy component lems and behavior change, such • Award of the incentive if the participants are able to answer several questions as avoiding debt and seeking help about the non-�nancial content of the show correctly once over-indebted, will inform • A phone-based survey with transfer of the �nancial incentives coming in the the instrument choices that South form of airtime credit African policy makers will make • A �nal face-to-face survey looking at the longer-term effects of the soap opera 4 months after the �nancial literacy part of the show was broadcasted in designing �nancial capability building strategies. Beyond South The estimation of the impact of the soap opera on knowledge, attitudes, and behavior will be straightforward as the randomization allows for a direct Africa, the project is expected comparison between the treatment and the control group. The random selection to make a major contribution to and comparison to a control group allows for attribution of the effects found. The the policy discussion on the use �nancial incentive insures that a comparison sample is available for interviews and the incentive for both treatment and control groups eliminates any confounding of entertainment education as effects of the incentive itself. Cumulative and/or non-linear effects of the �nancial a behavior change tool and to literacy messages can be detected through several rounds of interviews. And inform policy choices globally. lastly, spillover effects can be measured by comparing those who did not watch but heard about the content with those who neither watched nor heard about it. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT SOUTH AFRICA South Africa Old Mutual Financial Education Study Purpose of Research Financial literacy levels are low in South Africa. Informal burial societies are the primary source of cross-insurance, yet such societies provide, at best, incomplete coverage. The purpose of this �nancial literacy Thematic Area program is to encourage individuals to seek more ef�cient �nancial vehicles, as well as to save and use credit wisely. Savings, credit, insurance, and �nancial product selection Target Segment Background and Objectives Members of burial societies and Governments, NGOs, and aid organizations are increasingly focusing on women’s development groups in �nancial literacy education as a tool for improving welfare. Yet to date, the Eastern Cape area there is little rigorous evidence that �nancial education is effective. This 59 , project evaluates Old Mutual’s “On the Money� a one-day �nancial Delivery Channel education program that provides training on saving, �nancial planning, Group-based, interactive �nancial budgeting, and debt management. The training program is very similar literacy seminar in content and delivery to a �nancial literacy evaluation being conducted by the same authors in India. Evaluation Team To rigorously measure the impact of the intervention, we use a Bilal Zia (World Bank); Shawn Cole randomized control trial. We identify and survey about 1,300 individuals: (Harvard Business School) 610 organized in 43 Burial Societies and 690 organized in 36 Women’s Development groups in the Eastern Cape and KwaZulu Natal. A Partners randomly-selected half of these groups will receive �nancial education. The other half will form the control group for the duration of the study. Old Mutual South Africa; J-PAL South Africa After the baseline survey and intervention, the follow-up survey will be conducted in approximately six months after groups receive training. Timeline Any differences we will observe between the treatment and comparison November 2011 – November 2012 group can be causally attributed to the training as randomization ensures that the training and control groups are on average identical at the outset of the study. The evaluation is based on Old Mutual’s “On the Money� �nancial education program. The program is administered during a full day seminar with scheduled groups, and covers �ve main topics: (1) Saving, (2) Financial Planning, (3) Budgeting, (4) Dealing with Debt, and The Russia Trust Fund for Financial Literacy and Education (5) Investing. The course is taught in the native language of the group was established in October 2008 to support the advancement using power point slides as well as educational videos, and includes of �nancial literacy and capability, through the development a workbook that helps to reinforce the concepts. In addition, the of methods and best practices for the assessment of �nancial workbook is used during interactive group activities and discussions, capability and for the evaluation of �nancial education programs. where attendants can practice creating budgets together and discuss The Trust Fund is managed by the World Bank implemented in part their �nancial plans for the future. by the Organization for Economic Co-operation and Development. Evaluation Methodology In order to establish a causal effect, our study incorporates random assignment Policy Impact of burial society groups (BSSP) and borrowing groups of the Women’s Development Bank (WDB) to receive Old Mutual’s “On the Money� �nancial In the last few decades, large amounts literacy training. This ensures that those who receive training are statistically of funding have gone into develop- indistinguishable from those who do not, and that any observable difference between the groups (in terms of saving, borrowing, etc.) after the intervention ment initiatives to improve access to was implemented can be attributed solely to the training program. �nance for low-income individuals Examining the scope and pathways by which �nancial education affects �nancial throughout the developing world. behaviors requires that we obtain a rich dataset on trained participants and the Yet, practitioners have recently control group. To that end, we will obtain outcome measures from a variety of become concerned that consumers sources, including individual surveys and administrative data collected by Old Mutual and WDB, which will allow us to understand how �nancial literacy evolves, lack the knowledge of �nancial con- the ways individuals change their �nancial behavior in response to training. cepts and are ill-prepared to make To estimate the true causal impact of �nancial education, we need to establish �nancial decisions that will ben- the correct counterfactual—i.e. what our trained participants would have done e�t their economic well-being. As a had they not received the training. Studies that simply compare individuals who result, �nancial literacy has received receive �nancial education to those who do not are susceptible to selection bias, meaning that people who choose to take �nancial literacy courses much attention from various interest may differ from those who choose not to take such courses. Our evaluation groups, organizations, and govern- methodology—the randomized controlled trial—eliminates that bias. By randomizing assignment to treatment or control, we ensure the offer to attend mental agencies. training is not correlated with any potentially confounding factors like level of education, income, or motivation. This study will add to the growing body of evidence analyzing the rela- tionship between levels of �nancial literacy and potentially harmful �nancial choices and poor �nancial situations. It will further rigorously 60 test whether the notion that �nancial literacy education is an important tool to influence individual behavior and improve the �nancial condition of individuals. Lastly, the compari- son with a similar study in India will provide valuable insight into how differences in context can explain dif- ferential impacts. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT UGANDA Financial Education in Rural Uganda: Insights from Financial Diaries and In-Depth Interviews Purpose of Research Thematic Area The purpose of the project is to examine the use of a new, innovative Savings; Budgeting; Debt survey methodology – Financial Diaries – in combination with in-depth Management; Bank Services; interviews, in understanding and measuring the �nancial capabilities of Financial Negotiations, and Risk low-income individuals. Management and Insurance. Target Segment Background and Objectives Low-income individuals and Financial Diaries are multi-period surveys of individuals that record 61 households in the Luwero District all their economic transactions over a period of a number of months. in Central Uganda They generate detailed information on the economic behavior of respondents, including their �nancial service use, and, as a result, have Delivery Channel the potential to be a useful tool in the analysis of behavioral change Traditional �nancial capability resulting from �nancial education and other initiatives. To examine that training sessions potential, the project addresses the following three questions: Evaluation Team 1. What can Diaries, in combination with in-depth interviews, tell us Dr. Guy Stuart (Consultant to about the �nancial capabilities of low-income people that we might Micro�nance Opportunities) not know otherwise? Dr. Craig Tower (Micro�nance 2. How can change in indicators of �nancial capability be tracked Opportunities) through Diaries, in combination with in-depth interviews, over time? Dr. Monique Cohen (Micro�nance 3. Under what circumstances is it appropriate to use �nancial Diaries Opportunities) to evaluate the impact of a �nancial education program? Partners The answers to these questions are grounded in the conceptual Habitat for Humanity, Uganda framework for understanding �nancial capability that we have developed for this project. The framework builds on work by others, Timeline notably Elaine Kempson and Monique Cohen, and also draws on the April 2011 – August 2012 work of behavioral economists. It identi�es three key components of �nancial capability and identi�es three key dimensions of these components: the cognitive, the behavioral and the environmental or structural dimensions. The components are: The Russia Trust Fund for Financial Literacy and Education was established in October 2008 to support the advancement 1. Day-to-day cash flow management, which includes: making ends of �nancial literacy and capability, through the development � meet for both ordinary expenses and “lumpy, large expenses of methods and best practices for the assessment of �nancial (behavioral), keeping track of your cash flow and budgeting capability and for the evaluation of �nancial education programs. (cognitive), understanding the difference between necessary and The Trust Fund is managed by the World Bank implemented in part unnecessary expenses (cognitive), and avoiding unnecessary by the Organization for Economic Co-operation and Development. expenses (behavioral); 2. Planning for the future, which includes: making contingencies, such as saving up or buying insurance, for future emergencies/risks, including a time when one can no longer work, or opportunities (behavioral), as well Policy Impact as the cognitive process of articulating to oneself what those might be and developing a strategy for making such contingencies; Preliminary results from baseline 3. Financial service use, which includes: choosing the right �nancial tool data collected from May to July 2011 (savings, loan, insurance, or grant/gift) for the right occasion (behavioral), suggest: based on an understanding of the costs and bene�ts of each (cognitive), and choosing the right �nancial service provider of each of these (behavioral) 1. The prevalent use of home sav- given the choices available (environmental). ings as a cash flow management From these components the project develops indicators of �nancial capability, tool, despite it being regarded which it then operationalizes through the analysis of Financial Diaries and in- depth interview data. For example, in examining whether individuals plan for by respondents as a second-best the future we can use Diaries data to examine how they behave in the face of �nancial tool. an unexpected expense such as a hospital bill. Do they borrow money from a �nancial service provider? Or do they reach out to friends and family for �nancial 2. Home savings and saving in SAC- support? The Diaries data can provide answers to these questions because they include information on all expenditures and also the source of the money COs or savings groups seemed to used to pay for those expenditures. In addition, because the data are gathered have been, to some extent, substi- weekly, they allow researchers to observe the coincidence of expenditures and tutes. But these two ways of sav- cash inflows in time. As a result, when an emergency expenditure is observed it is possible to identify the sources of cash used to pay for those expenditures. ing differed in a key dimension: the latter required a disciplined The project will not only demonstrate how to use Diaries to detect these sorts of transactional sequences (behaviors), but will also examine whether commitment to saving, while the Diaries can detect changes in transactional sequences or behaviors over time. former did not. To extend the previous example, do we see a change in behavior over time whereby unexpected expenditures are no longer paid for from the proceeds of 3. The limited use of debt as a way a loan, but paid for out of savings the respondent has accumulated? to manage �nancial challenges – it was uncommon for someone to Research Design borrow money to cover a de�cit, 62 The research is taking place in two communities where HFHU is offering pay for a lump sum expenditure, �nancial education and in two comparison communities outside of HFHU’s or cover an emergency expendi- service area. All respondents reside in the Luwero district and in economically and demographically similar communities to each other. Financial diaries data ture. They were far more likely to are collected at weekly visits with respondents before, during and after the withdraw cash from their home intervention. The period after the intervention consists of two time periods: one savings or get money from a friend immediately following the intervention and another twelve months after the start of the Diaries data collection. The sample consists of 103 respondents, 47 or relative. in the treatment group and 56 in the comparison group, in Phase 1, which covers the pre-intervention period, the intervention period, and the immediate post- 4. People are more likely to turn to intervention period. In Phase 2, which covers the long-term, post-intervention another individual for help during period there are 90 respondents remaining from Phase 1, 38 in the treatment and 52 in the comparison group. The attrition in the number of respondents is an emergency in comparison to due to a mix of changes in circumstances of respondents (e.g. death, moving when they need money to pay for away) and some refusals to rejoin the study. an ordinary lump sum expendi- This design allows us to observe changes in behavior (manifested in changes ture. This suggests that people are in the pattern of economic transactions) made in both the short- and the long- embedded in networks of mutual term in both the treatment and comparison groups. We will also be able to observe changes in behavior by comparing patterns of transactions during the support, which they can tap into in same time of year, one year apart and on either side of the intervention, to take an emergency. into account seasonality. In-depth interviews are conducted with respondents before, immediately after, and ten months after the intervention to identify any changes in knowledge among the treatment and comparison groups. June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Measuring Financial Capability and the Effectiveness of Financial Education PROJECT SNAPSHOT UGANDA Network Effects in SME Clusters Purpose of Research This impact evaluation uses a randomized control trial (RCT) to assess whether vocational and business training for small scale entrepreneurs can impact business knowledge and outcomes. Further, the research Thematic Area maps out business networks and seeks to identify to what extent such enhanced knowledge is shared among network members. Vocational and business training, and business network effects Target Segment Small scale industries Background and Objectives Identifying the determinants of entrepreneurship is an important Delivery Channel research and policy goal, especially in emerging market economies Practice based vocational training where lack of capital and supporting infrastructure often imposes 63 and classroom based business stringent constraints on business growth. training It is important to note that businesses do not operate in a vacuum. They Evaluation Team interact with neighboring businesses, with close associates operating in other areas, and with businesses above and below in the supply Markus Goldstein (World Bank) chain. Importantly, many businesses are part of larger networks, some Francisco Campos (World Bank) of which make joint sales decisions, share costs, revenues, working Obert Pimhizai (World Bank) capital, and production information. Existing research has identi�ed the Antoinette Schoar (MIT) breadth and scope of such networks as being quite large and widely Mattea Stein (Paris School of prevalent in developing countries. Economics) But how is information shared among networks, is it voluntary or Bilal Zia (World Bank) involuntary? There are two competing hypotheses of information Partners sharing within networks -- on the one hand, business networks may be formed to overcome failures in the credit, labor, or other input Katwe Small Scale Industries markets, and here network members operate as one large �rm, jointly Association in Kampala, maximizing pro�ts and welfare. In this model, information would Government of Uganda freely flow to all network members even if only a subset is trained. An alternative view, however, is that network membership is involuntary Timeline (inherited or otherwise enforced), and in a competitive environment, July 2011 – December 2012 network members have an incentive to withhold information from others in order to gain a competitive edge. In our proposed work, we have the opportunity to explicitly test these The Russia Trust Fund for Financial Literacy and Education two predictions in a competitive market of informal small-scale industrial was established in October 2008 to support the advancement producers. These are metal fabricators, shoe makers, caterers, and the of �nancial literacy and capability, through the development like, operating in industrial clusters in the outskirts of Kampala, Uganda. of methods and best practices for the assessment of �nancial We are working with the local industry association to implement an impact capability and for the evaluation of �nancial education programs. evaluation of a comprehensive technical and business training program The Trust Fund is managed by the World Bank implemented in part sponsored by a joint Ugandan Government and World Bank initiative. by the Organization for Economic Co-operation and Development. Evaluation Methodology Owing to the physical clustering of workshops, our study will focus on delivering Policy Impact trainings at the cluster rather than individual level. Moreover, the randomization will be at the cluster level. This will avoid spillovers to immediate neighbors Studying how business networks who might be in the control group, and importantly will be more acceptable to operate is important for policymak- workshop owners participating in the evaluation. ing as any identi�ed positive exter- Although neighborhood spillovers will be controlled by design, we hope to be nalities could justify scaling down able to capture spillovers across business networks. Our initial pilots and focus group discussions have shown that many workshops are linked to others in spending on business and �nancial business networks that are not necessarily dependent on immediate physical training programs while still being proximity. Hence, in the baseline survey, we will record the names and contact able to reap many if not all the ben- details of �ve such closest business associates of each �rm owner, followed by a series of questions on how information, knowledge, working capital (such as e�ts. For example, if we �nd network labor), and sales orders are shared with these businesses. This will enable us effects are signi�cantly positive, then to map out the network tree of each entrepreneur in our sample, separate and apart from the physical cluster. policy can be designed to provide training to only a subset of those After the trainings are implemented, we will conduct follow-up surveys not targeted, who can then spread the only on treatment and control clusters (in order to get �rst-order effects of our training program), but also on network members of both treatment and control knowledge to others. These potential groups. The comparison of information sharing and outcomes for these groups positive spillovers would constitute will then identify the value of business networks. ef�cient ways to scale the impact The difference in the nature of technical and �nancial skills will further allow of trainings and provide a natural us to test the two competing theories of information sharing discussed above. source of leverage for these pro- Speci�cally, technical skills are highly observable and cannot be hidden from others, who can learn these skills from direct observation. In contrast, �nancial grams. Furthermore, the resources management skills are not so observable, and accounting books and record- saved in doing so could be spent on keeping can easily be done behind closed doors. Hence, if networks are joint welfare maximizers, then we would not see any difference in the sharing of expanding training in other areas or technical and �nancial knowledge. However, the alternative hypothesis would on other development projects. imply that there is less sharing of �nancial knowledge. 64 June 2012 The �ndings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank and 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. For more information please visit www.�nlitedu.org. Evaluation Toolkit for Financial Capability Programs in Low and Middle Income Countries While there are many useful existing toolkits that address different aspects of program evaluation, the Toolkit �nanced by the World Bank Trust Fund for Financial Literacy and Education is focusing on the speci�c challenges of evaluating �nancial capability interventions in the low-income and middle-income country setting. This Toolkit and its related activities (including testing and dissemination) will be targeted to organizations, �rms, and individuals undertaking evaluations of 65 different types of �nancial capability interventions, both in the public and private sector, including program staff and managers who are contracting out such evaluations. The overarching goal is to facilitate understanding of and build support for evaluations among this potentially diverse set of stakeholders. To this end, the Toolkit team, led by the RAND Corporation, will address three interrelated tasks: 1) designing the Toolkit itself to serve as a guide for designing and conducting evaluations, bearing in mind the needs and capabilities of the intended users; 2) supporting the selection and ongoing interaction with the pilot programs already funded under the broader Trust Fund initiative in order to incorporate their experience into the Toolkit, and 3) conducting dissemination activities in collaboration with the World Bank. As part of the formative research for the Toolkit, the team conducted a stock take of existing �nancial capability interventions as well as a series of stakeholder interviews with various practitioners currently undertaking evaluations, to understand key research gaps and practice needs. Based on the goals of the Trust Fund and the initial formative research, the basic components of the Toolkit are currently under development. In the introductory chapters, the Toolkit will provide an overview of the need for �nancial capability, and introduces a model of �nancial capability (including knowledge attitudes, beliefs and latent skills, manifest skills and behavior, and ultimately �nancial well-being and security) in order to frame the role of �nancial capability interventions. It will outline notable examples of interventions of interest, including school-based programs, social media or edutainment based and programs that are built into government-to-people (G2P) interactions, as well as new innovations that draw on psychology and behavioral economics. Finally, it will discuss the broad rationale for evaluation, and the role that evaluation plays in reshaping programs as well as broader policy directions. The Toolkit will then move on to the use of a logical framework as a starting point for the evaluation of any intervention of interest: what is the constraint or problem that the intervention is intended to relieve, and how, in principle, does the intervention lead to the outcome (for instance, by changing decision inputs or processes, by introducing incentives or removing constraints). From the basic logical framework approach, the Toolkit will then go on to address basic empirical methods for performing 66 both monitoring and process evaluation and impact evaluation, as well as cost-bene�t analysis. What will set this Toolkit apart will be the focus on issues speci�c to �nancial capability programs, as well as the use of concrete examples drawn from the Trust Fund pilots and other relevant studies. In the �rst case, for instance, the Toolkit will describe practical and ethical challenges that may occur in such programs, such as applying innovative survey methods for eliciting sensitive �nancial information or measuring �nancial knowledge and numeracy, or understanding whether �nancial service provider incentives lead to conflicts of interest that could negatively affect consumers. In the second instance, the Toolkit will link these discussions to illustrations and resources from the �eld, such as a repository of well-validated sample survey instruments and protocols for informed consent and disclosure. By drawing on real-world examples and providing practitioners with actionable tools, the Toolkit’s goal is to make state- of-the-art evaluation tools relevant and accessible to the community of policymakers and practitioners in the �eld of �nancial capability. THE RUSSIA TRUST FUND FOR FINANCIAL LITERACY AND EDUCATION COUNTRY NAME | PROJECT NO. | FINANCIAL LITERACY & EDUCATION Contact Information Trust Fund Management Financial Capability Measurement Impact Evaluation Richard Hinz Valeria Perotti Florentina Mulaj HDNSP HDNSP HDNSP rhinz@worldbank.org vperotti@worldbank.org fmulaj@worldbank.org 202-458-2604 202-458-7488 202-458-8043 www.�nlitedu.org