80539 Access to Finance Forum Reports by CGAP and Its Partners No. 5, May 2012 Financial Access 2011 An Overview of the Supply-Side Data Landscape Oya Pinar Ardic, Gregory Chen, and Alexia Latortue This report was written by a team from IFC and CGAP. The authors would like to thank Kathryn Imboden, CGAP consultant, who conducted the interviews for Section V and led the drafting of the chapter. Our thanks also to Scott Gaul of MIX who provided a box on data measurement in sub-Saharan Africa. We are also grateful for the helpful comments and review provided by Nina Bilandzic, Tilman Ehrbeck, Kate McKee, Bikki Randhawa, Rich Rosenberg, Peer Stein, and Jeanette Thomas. We thank especially the policy makers and data experts who agreed to share their views on the importance of data, progress achieved, and the path ahead for data to help accelerate financial inclusion. They are Diane Jocelyn Bizimana, Raúl Hernández-Coss, Marten Leijon, David Porteus, and Hassan Zaman. Finally, the support of our partners is indispensable. This report was produced with the financial support from the Netherlands’ Ministry of Foreign Affairs to IFC and from AusAid to CGAP. © CGAP and International Finance Corporation, 2012. All rights reserved. CGAP 1818 H St., NW Washington, DC 20433 USA Internet: www.cgap.org e-mail: cgap@worldbank.org Telephone: +1 202 473 9594 This volume is the product of the staff of the Consultative Group to Assist the Poor (CGAP) and staff of IFC (International Finance Corporation), a member of the World Bank Group. The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. CGAP and IFC does not guarantee the accuracy, reliability or completeness of the content included in this work, or for the conclusions or judgments described herein, and accepts no responsibility or liability for any omissions or errors (including, without limitation, typographical errors and technical errors) in the content whatsoever or for reliance thereon. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of CGAP or IFC concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to CGAP at the address in the copyright notice above. CGAP and IFC encourage dissemination of their work and will normally give permission promptly and, when reproduction is for noncommercial purposes, without asking a fee. Permission to photocopy portions for classroom use is granted through the Copyright Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, MA 01923 USA. Contents I. Introduction 1 II. Overview: Landscape of Financial Inclusion Data 3 III. Supply-Side Data 7 IV. Improving Financial Inclusion data 15 V. Conversations on Data: Five Experts Share their Perspectives 19 Annex I. The AFI Core Set 26 i Abbreviations A2F Access to finance AFI Alliance for Financial Inclusion ATM Automatic teller machine BIS Bank of International Settlements CGAP Consultative Group to Assist the Poor CNBV Comisión Nacional Bancaria y de Valores ECB European Central Bank ECB MFI ECB Monetary and Financial Institutions database ECB BLS ECB Bank Lending Survey ECB HFCS ECB Household Finance and Consumption Surveys G-20 Group of Twenty G-8 Group of Eight GDP Gross domestic product GIZ Deutsche Gesellschaft für Internationale Zusammenarbeit GPFI Global Partnership for Financial Inclusion IFC International Finance Corporation IMF International Monetary Fund IMF FAS IMF Financial Access Survey IMF FSI IMF Financial Soundness Indicators IMF IFS IMF International Financial Statistics IMF WEO IMF World Economic Outlook MECOVI Measurement of Living Conditions in Latin America and Caribbean MSME Micro, small, and medium enterprise OECD Organisation for Economic Co-operation and Development SME Small and medium enterprise UN United Nations WB World Bank WB SFS WB Survey of Financial Services WB LSMS WB Living Standards Measurement Study WB CP/FL WB Consumer Protection and Financial Literacy surveys WB WDI WB World Development Indicators database WBG World Bank Group WOCCU World Council of Credit Unions WSBI World Savings Banks Institute ii Foreword T here is real momentum behind the belief that better, more con- sistent, and increasingly comprehensive data are key for better decision-making and tracking progress in advancing access to financial services for the poor. A growing number of countries are pav- ing the way nationally and informing global data efforts. Global data initiatives are learning from these country experiences and, in turn, provide guidance and examples for others. While different countries may be in different places in terms of measuring financial inclusion, the foundation for country-owned efforts, tailored to national priorities, is being laid in many markets. These efforts are informed by important progress on both demand- side and supply-side survey tools, and a greater recognition of how the two work together. On the demand side, the World Bank released in 2012 the Global Financial Inclusion (Global Findex) Database, a comprehen- sive, comparable, cross-country dataset that measures how women, men, and youth save, borrow, make payments, and manage risks. The project, which covers 148 economies, is funded by the Bill & Melinda Gates Foun- dation and is implemented in partnership with Gallup. On the supply side, the International Monetary Fund has enhanced its globally compa- rable Financial Access Survey database to distinguish between small and medium enterprises and households as well as different types of financial institutions that serve the poor. As a result of these efforts, a robust inter- national financial inclusion data architecture is emerging. The G-20 has also embraced financial inclusion data as a priority. One of the three subgroups of the Global Partnership for Financial Inclu- sion (GPFI) is focusing on advancing the data and measurement agenda as a strong base for informed policy action and market knowledge. As Implementing Partners of GPFI, we are pleased to contribute this report on the supply-side landscape of financial inclusion data. Although this year’s Financial Access does not contain new data, it provides an overview of data sources and discusses select methodolog- ical supply-side data issues. It also includes a discussion with leading policy makers and market actors that give voice from the field as to why and how data can advance access to finance. They comment on what’s been achieved and point to continued work ahead. iii We are pleased to share this report as GPFI is proposing the G-20 Basic Set of Financial Inclusion Indicators for the leaders’ consider- ation at the 2012 G-20 Summit in Mexico. G-20 recognition is a sign of how far financial inclusion has come. Among the many champions along the road, we would like to specially thank H.R.H. Princess Máxi- ma of the Netherlands, UN Secretary-General’s Special Advocate for Inclusive Finance for Development and Honorary Patron of the GPFI, who has tirelessly and passionately spoken out for investing in financial inclusion data. Tilman Ehrbeck Peer Stein CEO and Director Global Business Line CGAP Leader, Access to Finance Advisory IFC iv 1 Pa r t Introduction T he lack of data has long been recognized as a major barrier to extending access to financial services to low-income households and small businesses. Considerable progress has been made in recent years. In June 2004 at a meeting of heads of state at Sea Island, Georgia, United States, the Group of Eight (G-8) endorsed the “Key Principles of Microfinance” developed by CGAP. In September 2009, the G-20 leaders made important commitments to financial services for the poor at the Pittsburgh Summit, and their commitment to finan- cial inclusion has been reaffirmed at each subsequent Summit. At the Seoul Leaders’ Summit in November 2010, the Global Partnership for Financial Inclusion (GPFI) was established to institutionalize and im- plement the G-20 Financial Action Plan. A central theme of GPFI is data and measurement, with one of the three GPFI subgroups tasked with identifying the existing financial inclusion data landscape, assess- ing data gaps, and developing key performance indicators. National governments have also taken action, commissioning demand-side data surveys, setting national financial inclusion targets, and establishing cross-governmental agencies to tackle the issue. We are now at a tipping point. Interest in financial inclusion is at an all-time high. Policy makers and standard-setters, ranging from local central banks to global standard-setting bodies, increasingly view stability and inclusion as complementary, mutually reinforcing goals. Innovations in technology and business models offer new pos- sibilities for reaching low-income households and small businesses more cost-effectively. Global, regional, and national social investors are seeking opportunities at the base of the pyramid that can provide returns while fulfilling environment, social, and governance stan- dards. And importantly, there is increasing focus on clients and deliv- ering a range of quality services that respond to their needs and enhance their well-being and performance. A number of new data initiatives have also emerged to offer better and more meaningful demand- and supply-side data. This report comes at a time when the results of some new or improved surveys are being published—the World Bank’s Global Financial Inclusion (Global Find- ex) Database funded by the Bill & Melinda Gates Foundation and the enhanced Financial Access Survey (FAS) of the International Mone- tary Fund (IMF). Together, these will offer a far more comprehensive picture of the state of financial inclusion. 1 This year’s Financial Access presents an overview of the landscape of financial inclusion data, with a focus on supply-side data. It is markedly different from the two previous reports, published by CGAP and the World Bank Group, which provided data on the state of financial inclusion. The next Financial Access will include new fi- nancial access data. The overview that follows discusses the landscape of financial in- clusion data, with a presentation of key demand- and supply-side data sources and a brief look at the findings from Financial Access 2010. Part 3 provides a discussion of supply-side data, with information on country-level data and how global-level data build on it. Part 4 focuses on the gaps in financial inclusion data and recommends ways these can be addressed by different stakeholders. The final section offers the perspectives of leading experts on financial inclusion data. Their first-hand experiences and reflections provide insights on why data are important and how the creators and users of data can make prog- ress, both in data collection and in the use of data to further financial inclusion. 2 2 Pa r t Overview: Landscape of Financial Inclusion Data F inancial inclusion is increasingly a policy pri- variety of channels. Access points are often mea- ority for governments and a goal of the finan- sured in proportion to population and are also as- cial system. Financial Access 2010 showed that sessed by the reach and spread of different access almost half of the reporting countries had financial points, increasingly including nonbranch loca- inclusion strategies, and a majority of these strate- tions, such as ATMs and mobile phone networks. gies were created in 2004 or later. Data play a cru- 2. Usage of services. The purest measure of inclu- cial role in establishing a common understanding of sion is the extent to which clients use different the current state of financial inclusion, informing services. In its simplest form this would include action needed from various stakeholders, and as- the number of savings or loan accounts in pro- sessing progress. Financial institutions can use data portion to the population. However, more so- to better understand market opportunities. Regula- phisticated data can also provide further insight tors can use it to understand trends, identify risks, into which market segments use different ser- and make evidence-based policies. Policy makers vices. Market segments may be broken down by can look for gaps, establish priorities, and monitor income, gender, age, location, occupation/liveli- change over time. All of these users have a need for hood, and other demographic variables. When the data, and they can also play a role in increasing such data are available, they can help guide plan- the availability and quality of data. ning and targeting to improve financial inclusion However, financial inclusion is neither a simple by showing where there are segments or services concept nor easy to measure. Financial inclusion with the greatest opportunities. The level and refers to a state in which all working-age adults frequency of activity or usage is also pertinent. have effective access to credit, savings, payments, Setting an optimal level of usage across different and insurance from formal service providers.1 By market segments and the range of financial ser- this high standard, financial exclusion would in- vices is a complex and much discussed issue, and clude those underserved in addition to those not there is a need for more research. served at all. Moreover, inclusion does not mean the mere availability of services but rather wheth- 3. Quality of products and service delivery. The er various dimensions of the financial system are gap in access to finance for the unbanked and un- working effectively to extend demand-driven ser- derserved has been so large that, for a long time, vices to clients. In addition to access, there are at the focus was simply on closing the gap. As recent least two more dimensions to inclusion that, over microcredit crises have shown, the poor match of time, should be part of measurement: credit products to customer capacities can have deleterious effects on inclusion. More attention is 1. Access to financial services and reach of finan- needed to deliver the portfolio of services that cial infrastructure. Access reflects the depth of will meet low-income people’s underlying finan- outreach. The physical reach of branches, auto- cial needs. Beyond product diversification and matic teller machines (ATMs), and agent loca- suitability to clients, quality involves features tions is often a necessary (though not sufficient) such as transparency, safety, fair pricing, client condition for inclusion. This enables the formal value, and other basic core tenets of consumer system’s infrastructure to reach clients across a protection and financial capability. Price and  ee CGAP (2011). 1. S 3 nonprice barriers to access, such as fees or mini- number of indicators, usually on one or more spe- mum balance requirements, are an important cific dimensions of financial inclusion. component of product design. In addition, better Demand-side data often offer rich information financial infrastructure, for example, credit re- on how services are used and which customers are porting systems or secured transactions frame- being reached. Demand-side surveys, however, works, provide a sound foundation for the tend to be quite costly, take time, and are not al- high-quality delivery of financial services. ways comparable over time. Supply-side data are quite different in that they often require gathering Capturing the various dimensions of financial in- data from providers and are generally collected at clusion through the collection of comprehensive regular intervals. Supply-side data, however, usu- indicators can help inform the policy dialogue and ally offer aggregate-level numerical data—except accelerate progress toward responsible financial in- for financial institution surveys—and most of the clusion. To have the power to persuade and influ- time capture only organizations that report to the ence policy making and the business decisions of financial regulator, thus leaving out important financial institutions, data need to be credible and sources of financial services, especially informal consistent. Finally, data also ought to converge to- finance, upon which large numbers of poor and ward standard definitions so that comparisons can low-income people count. be made over time and across countries. CGAP/World Bank Group Financial Data Sources—Supply and Demand Access Series Financial inclusion data are derived from two Financial Access 2010 was the second in the series main sources. Demand-side data are collected of annual reports by CGAP and the World Bank from the users of financial services, such as indi- Group to monitor statistics for financial access in viduals, households, and firms. Typically data col- the world and inform policy debate. The 2010 sur- lection is done through surveys or focus groups, vey included specific questions on survey initia- including qualitative research. The second source tives at the national level to monitor access to is supply-side data that are collected from finan- financial services. Survey respondents were the cial service providers, such as banks, cooperatives, primary financial regulators—central banks or microfinance institutions, and other financial in- bank superintendents in most cases. The survey in- stitutions. In some contexts, other businesses, cluded questions on whether countries used such as mobile operators, may also be big players household, firm, and/or financial institution sur- in financial inclusion. Typically, supply-side data veys to monitor the state of financial access. The are collected at the national level by the financial survey also asked whether access to finance by regulator for regulated institutions via regular re- small and medium enterprises (SMEs) was moni- porting. Globally, associations of different types of tored specifically. This information, gathered by providers often collect data, as does the IMF and the CGAP/WBG Financial Access survey, is as of other international organizations or specialized end-2009 and is summarized in annexes available data analysts. online.2 Figures 2 and 3 highlight responses re- Demand- and supply-side data are complemen- ceived to questions on financial inclusion surveys, tary. Figure 1 provides a look at global and multi- monitoring, and strategies from participating country financial inclusion data initiatives, countries. organized by data source (demand- or supply- side) and depth of coverage. Broad coverage initia- http://www.cgap.org/financialindicators and http://www.ifc. 2.  tives provide data on a basic set of indicators, org/accesstofinance. The annexes include country-by-country while deeper coverage initiatives include a larger lists of (i) household, firm, and financial institution surveys, along with the frequency of data collection and (ii) specific ef- forts to monitor access to finance by SMEs. 4 figure 1 Existing global/multi-country demand- and supply-side data sets Broader coverage IMF FAS IMF IFS Global Findex IMF FSI Opinion Polls Supply Demand side side WBG Payment Systems WB LSMS BIS Payment Systems WB Enterprise Surveys WB SFS ECB HCFS WBG Remittance Prices ECB A2F of SMEs ECB MFI MECOVI ECB BLS FinScope Bankscope OECD Financial Education WSBI WB CP/FL WOCCU WB Migration & Remittances MIX Financial Diaries Microcredit Summit Deep coverage Source: Matrix representation is adapted from Bill & Melinda Gates Foundation (2010). “The Measurement Challenge,” Note prepared for the Global Savings Forum. See page ii for list of abbreviations and page 11 for brief descriptions of supply-side data initiatives. 5 figure 2 Country-level monitoring and data collection efforts How widespread is the use of surveys to monitor Do financial regulators monitor the level of lending access to financial services? to SMEs by regulated financial institutions? 80 70 73 59 Number of countries 70 66 60 62 Among these 46: 60 57 Other responsible agency exists in 12 50 46 50 No agency is responsible in 24 50 41 40 40 30 30 23 20 20 20 13 10 10 0 0 HH survey Firm survey Financial institution Regular Financial Credit Irregular No survey reporting institution registry monitoring monitoring surveys estimates Use Do not use Source: CGAP/WBG Financial Access database. Left panel is based on responses by 120 countries. Right panel is based on responses by 120 countries, 23 of which use more than one method to monitor SME lending. figure 3 National strategy documents and data collection and monitoring efforts Strategy documents and data collection Monitoring SME lending by regulated institutions No strategy No strategy 27 40 document 26 41 document Strategy Strategy 45 19 document 46 18 document Number of countries Number of countries At least one survey No survey Monitor Do not monitor Source: CGAP/WBG Financial Access database. Based on responses from 131 countries. Financial Access 2010 highlights four important surveys are the least used. Third, SME finance indi- findings about data collection as illustrated in fig- cators are collected by the majority of countries, ures 2 and 3. First, about half of the respondents though usually by ministries promoting business monitor demand-side access to financial services development rather than by financial regulators. through some form of household survey, firm sur- Fourth, countries that have national financial inclu- vey, or financial institution survey. Second, house- sion strategy documents also tend to prioritize hold surveys are the most widely used, and firm financial inclusion data. 6 3 Pa r t Supply-Side Data P roviders of financial services, the supply- In some countries apex funders for many small un- side, track the services they deliver as a basic regulated institutions centralize data on a large function of their business. Provider data of- number of institutions, often with significant finan- ten include total numbers on loan or savings ac- cial inclusion implications. counts (including volume of loans and deposits); Often the data reported to regulators and apexes they may also include more detailed data on the are a matter of standard and regular periodic re- types of products as well as on the points of service porting. In such cases, formats for reports can be (number of branches, mobile banking penetration, developed and improved over time. The cost of data etc.). In some cases, there may even be client collection also decreases as providers become more data—number of individuals, number of firms, lo- accustomed to regularly collecting and reporting cation, gender, income levels, and other data that data. Formats and definitions are built into standard identify different market segments. information collection systems. Regular, standard reporting also allows for comparisons and for trends to emerge over time. Country-Level Data Are Ad hoc data collection efforts may also be used to examine a specific issue relevant to a particular Fundamental market that would not come to light in a global or The link between data analysis and policy design is regional survey. strengthened when using country-level data. That is both because there is often greater ownership and understanding of data collected at the country-level, Global-Level Data Build on and because surveys are tailored to specific ques- Country-Level Data tions or market issues identified by national actors. The level of disaggregation needed with regard to Supply-side data sets on a global scale are useful for ethnicity-based exclusion or urban–rural divide, for making comparisons across countries and over time, example, is often country and context specific. as well as for assessing trends in financial access Country-level data collection is increasingly around the world. Policy makers use globally com- common as policy makers recognize the impor- parable data sets to benchmark financial inclusion. tance of tracking levels of financial inclusion. Most Multi-country supply-side data sets are most of- often supply-side data are collected by regulators, ten based on country-level data. The IMF’s FAS and typically the central bank, and are often included as World Bank Group’s Payment Systems Survey col- part of regular reporting required of financial insti- lect data that central banks have already collected. tutions. Providers understand that reporting on Currently, FAS is the only supply-side data source data is an obligation of a licensed and regulated in- on a global scale that produces basic access and us- stitution; though there are also cases where lightly age indicators, enabling comparisons across coun- regulated or unregulated providers also provide tries and over time. Box 1 provides a snapshot of the data to the regulator. Beyond the regulator, apex in- global trends in access to finance based on FAS. stitutions, associations, or networks of smaller or- For global-level data to be useful, and to enable ganizations also consolidate data at the country cross-country comparisons, it is helpful to harmo- level. Examples include national microfinance as- nize definitions and standardize data collection sociations or networks/federations of credit unions. methodologies. This includes convergence toward 7 Box 1 The State of Financial Inclusion through the Lens of the IMF’s FAS Database Access to Financial Services Continues to continued to expand throughout the crisis. With the Grow, Albeit at a Slower Rate trend data available from FAS for 2004–2010, it is Over the past years, the IMF’s FAS has been col- now possible to construct Figure B1A, which plots lecting comparable time series data on the geo- the number of new deposit accounts and new loans graphical and demographic outreach of financial in the world by commercial banks, each year for services provided by a range of regulated financial which data are available (left scale), and contrasts institutions that report to their countries’ central these with the growth rate of world GDP (right bank. FAS data show trends in financial inclusion scale). With the significant drop in world GDP in and reveal that the number of savings and loan ac- 2009, we see the expansion of financial service use counts has continuously increased from 2005 to slowing down as well. 2010 (See Figure B1 A). ATM networks expanded during 2005–2010 at a The number of new deposit accounts created in faster rate than branches of commercial banks and commercial banks globally has increased every other deposit-taking institutions (see Figure B1B). year from 2005 to 2010, with the exception of Over the period, an average of 3.5 new ATMs per 2009, which coincided with the global financial cri- 100,000 adults was added per year. sis. However, after slowing down in 2009, the However, commercial banks and other deposit- growth rate of both new deposit accounts and new taking institutions also continued to build branches loans rebounded in 2010. Interestingly, the num- during this period. While branch networks of other ber of new deposit accounts has consistently out- deposit-taking institutions expanded more rapidly paced that of new loan accounts since 2006. than those of commercial banks in 2005, from 2006 Other deposit-taking institutions, such as credit to 2008, commercial bank branches expanded unions, financial cooperatives, postal savings banks, more rapidly. In 2009, the trend reversed once again and deposit-taking microfinance institutions, have as other deposit-taking institution branch networks experienced a decrease in the new accounts creat- started expanding faster. ed and new loans made per 1,000 adults at the ag- The positive story of growth in financial access, gregate since 2008. Other financial intermediaries, even with the financial crisis, however, masks large such as nondeposit taking microfinance institutions, regional and local variations. In 2010 developing did not expand the number of their customers (per countries, on average, had 539 deposit accounts in 1,000 adults) overall during 2004–2010 and experi- commercial banks per 1,000 adults, while high- enced declines in the number of new borrowers ev- income Organisation for Economic Co-operation ery year during the same period, except in 2007. and Development (OECD) countries had 1,560 de- Between 2008 and 2009, the total number of insur- posit accounts per 1,000 adults. Similarly, commer- ance policies globally declined by 110 million. cial bank loans per 1,000 adults average 149 and These findings are not entirely new. Financial Ac- 478 in developing and high-income OECD coun- cess 2010 reported that access to financial services tries, respectively. Note: The source for all the financial access data used here is IMF’s FAS. IMF started collecting data on financial access indicators in 2010, going back to 2004. The data are available at http://fas.imf.org. 8 FIGURE B1A  Expansion of deposit and credit services in commercial banks 50 5 4 # of new accounts per 1,000 adults 40 3 2 30 1 0% 20 –1 –2 10 –3 –4 0 –5 2005 2006 2007 2008 2009 2010 # of new deposit accounts per 1,000 adults # of new loans accounts per 1,000 adults world GDP growth (right scale) Note: The figure plots the number of new deposit accounts and new loans by commercial banks in the world each year, normalized by the number of adults (left scale), and the growth rate of world GDP (right scale). Data sources: IMF FAS for number of deposit accounts and number of loans; World Bank WDI for growth rate of world GDP. FIGURE B1B  Expansion of the physical outreach of the financial system # of new branches and ATMs per 100,000 adults 5 4 3 2 1 0 2005 2006 2007 2008 2009 2010 # of new commercial bank branches per 100,000 adults # of new branches by other deposit-taking institutions per 100,000 adults # of new ATMs per 100,000 adults Note: Data source—IMF FAS 9 the use of the same definitions and more common banks to formally contract companies to act as bank- data collection and indicator computation methods. ing agents (80 out of 135 countries did not allow for Convergence toward common use of terms ensures agents as of the end of 2009). comparability across countries and over time, helps To inventory the full range of global and multi- devise development strategies, and can be used to country data sources available, IFC and CGAP con- adapt or design informed policies. Some countries ducted a financial inclusion data stocktaking prefer an approach that is customized to their own exercise. This exercise inventoried resources and unique circumstances, even though the more cus- helped to identify key gaps in the global data re- tomized the approach, the less comparable such sources (IFC and CGAP 2011). A key global resource data are. For example, Brazil tracks data on banking is the IMF FAS supply-side initiative, which collects agents because this delivery channel is key to reach- the majority of the core indicators of financial access ing the underserved and unserved market in Brazil. that enable annual comparisons across countries.3 Yet, banking agent data may have less meaning in It collects data from financial regulators of more another context. Indeed, Financial Access 2009 indi- than 150 countries in a two-stage process: financial cated that regulation in many countries do not allow regulators collect data from financial institutions Table 1  Selected global and multi-country supply-side data collection efforts: A comparison IMF WBG Payment WB Survey on FAS Systems Survey Fin. Services MIX IMF IFS IMF FSI Publicly Available Yes No No Yes No Yes BASICS Frequency Annual Bi-annual Irregular Annual Varies Varies Developing Coverage Global Global Global Countries Global Global Basic Usage Indicators S C P S C S C S C S C DATA COLLECTION P I P I Access/Infrastructure S C P S C S C P I P Barriers to Access Yes Regulatory/Enabling Environment P Aggregated Yes Yes Yes Yes Yes USER Firm Household/Individual Yes Commercial Banks Yes Yes Yes Yes Yes Co-ops & Credit Unions Yes * Yes Yes * Specialized State Fin. Inst. Yes * Yes Yes * PROVIDER Microfinance Institutions Yes * Yes Yes * Insurance Providers Yes Yes Finance Companies Yes Yes Yes Informal providers** * IMF data sets categorize deposit-taking institutions as “commercial banks” and “others.” ** Informal providers include informal NGOs and savings groups. S – Savings, C – Credit, I – Insurance, P – Payments    Problem areas   Major data gaps Core indicators are suggested by Beck, Demirgüç-Kunt, and 3.  Martinez Peria (2007). 10 and aggregate at the country level, and FAS collects credit unions, financial cooperatives, and microfi- data from regulators and compiles these at the glob- nance institutions and (ii) the separate identifica- al level. This underscores that the key building block tion of SMEs, households, life insurance, and is country-level data collection that is sufficiently nonlife insurance companies. http://fas.imf.org consistent to be consolidated into a global database. The IMF FAS is not the only supply-side survey Microfinance Information Exchange (MIX). available. There are numerous databases on pay- MIX includes data on a significant majority of ments, financial services, microfinance, and other organizations globally that self-identify as microfi- categories that complement FAS. Table 1 provides nance institutions. These include a range of finan- a summary of selected current global or multi- cial institutions that primarily provide services to country supply-side data collection efforts, with a low-income market segments. Some are regulated focus on those that have broad coverage in terms as banks, cooperatives, or nonbank finance com- of products, countries, have dimensions of finan- panies, while others are nonprofits. The data in- cial inclusion, and are most relevant for low-access clude raw outreach numbers but also cost and financial markets. Table 1 illustrates that basic financial performance indicators. As of 2012, 2,000 usage and access indicators are reasonably well- institutions have reported to MIX. http://www. developed in the form of country-level aggregates, mixmarket.org especially for commercial banks. A more complete list of supply-side data collection Microcredit Summit Campaign Report. This ini- efforts follows. The list (presented alphabetically) in- tiative collects data on microfinance institutions and cludes surveys or initiatives that collect data uniquely verifies these data against reports by practitioners from their members. Note that, in many cases, it is and network or umbrella institutions to avoid dou- necessary to use several data sources to get a more ble-counting. The data set includes the number of complete picture. (See Box 2 for a brief example). active clients of microfinance institutions based on their poverty levels and gender. This effort is updat- Access to Finance ed annually. http://www.microcreditsummit.org Financing SMEs and Entrepreneurs—An OECD Scoreboard. This initiative provides a framework World Council of Credit Unions (WOCCU). to monitor access to finance by SMEs at country Country-level aggregated indicators on the number and global levels, in addition to a tool to support of credit unions, the number of credit union mem- policy design and evaluation. The framework con- bers, penetration, and volume are compiled in the sists of 13 core indicators, the majority of which are WOCCU database, based on reporting by member supply-side, spanning multiple dimensions of ac- credit unions. This initiative is similar to those of cess to finance for SMEs. MIX and Bankscope, though only country-level ag- gregates are publically provided; institution-level IMF Financial Access Survey. FAS was launched data are not provided. http://www.woccu.org in October 2009; it aims to collect high-quality, cross-country, annual geographic and demograph- World Savings Banks Institute (WSBI). The WSBI ic data on access to basic financial services on a database consists of institution-level data on WSBI global scale for use by policy makers and research- member savings banks, including loan and deposit ers. The latest round of data went online in June volume information. http://www.wsbi.org 2011 and includes data on more than 150 countries for 2004–2010. FAS is the only source of supply- Financial Sector side data from regulators worldwide that contains European Central Bank (ECB) Monetary Finan- the majority of the basic access and usage indica- cial Institutions. This is a database that summariz- tors. The 2012 round of FAS data collection is be- es monthly information reported by monetary ing conducted in collaboration with CGAP and financial institutions to ECB. Monetary financial in- IFC. The 2012 questionnaire features the follow- stitutions are defined to include central banks, resi- ing changes: (i) the addition of time series for dent credit institutions, and other resident financial 11 institutions that take deposits, give credit, or invest WBG Payment Systems Survey. This bi-annual in securities. The database provides balance sheet survey of the World Bank Group collects data on information of reporting institutions and aggregates payment products; physical outreach of payment these data at a national level as well as for the entire systems, such as ATMs; legal and regulatory euro zone. http://www.ecb.int/stats/money/mfi framework regarding payment systems; and re- lated reforms. Data are collected from central IMF Financial Soundness Indicators (FSI). FSIs banks on a global scale. http://www.worldbank. aim to support macroprudential analysis and to as- org/paymentsystems sess strengths and vulnerabilities of financial sys- tems. The FSI database provides data reported on a WBG Remittance Prices Worldwide. This World regular basis by a number of IMF member coun- Bank Group database provides the cost of sending tries for 12 core and 28 optional indicators. Coun- small amounts of money internationally. Data are tries may report monthly, quarterly, semiannual, or collected through a mystery shopping approach annual FSIs. Measures such as deposit-to-loan and designed to be representative of global pricing. household debt-to-GDP ratios can be derived from The database is updated semiannually. these data, which add another dimension to finan- cial inclusion tracking. http://fsi.imf.org Banks Bankscope. This database by Bureau van Dijk in- IMF International Financial Statistics (IFS). IFS cludes detailed information on public and private is a database of regularly updated statistics on inter- banks worldwide, including the volume of depos- national and domestic finance on a global scale. For its and loans. most countries, IFS data are collected monthly, quarterly, semiannually, and annually. IFS provides ECB Bank Lending Survey (BLS). This is a survey global standardized data on money and banking of euro-area banks implemented four times a year aggregates that are helpful to indicate the overall by ECB to assess financing conditions, for which the size and trends in the financial sector, though they respondents are senior loan officers. Credit stan- do not necessarily provide detailed information on dards for loan approval, credit terms and conditions financial inclusion. for firms and individuals, and conditions affecting credit demand are among the topics covered by the Payment Systems and Remittances survey. http://www.ecb.int/stats/money/surveys/ Bank for International Settlements (BIS) Pay- lend ment Systems Data. The Committee on Payment and Settlement Systems (CPSS) of BIS publishes WBG Survey on Financial Services (SFS). This statistics on payment and settlement systems by survey, implemented by the Finance and Private member countries periodically. Data are collected Sector Development Research Group at the by central banks and include indicators of retail World Bank Group, is a direct survey of financial payment systems, payment instruments, and whole- institutions. Questionnaires are sent out to some sale systems used among banks, trading platforms, of the largest commercial banks around the clearing houses, and settlement systems for securi- world, and respondents are asked about the prod- ties as well as on the systems used to perform cross- ucts and services they offer as well as the associ- border transactions. http://www.bis.org/statistics/ ated fees and procedures to assess the barriers to payment_stats.htm access globally. This survey is not conducted reg- ularly. Data are aggregated at the country level and are available publicly. 12 Box 2 Piecing Together the Full Picture in Africa Financial diaries and demand-side surveys have re- to FAS, and even then, these data cover only regu- peatedly shown that the financial lives of the poor lated providers of credit. are complex. Poor people rely on a range of differ- ent providers for financial services, often many at Building better knowledge on Africa once. While commercial banks are part of that pic- Despite the challenges to building reliable and real- ture, regulated providers are only part of that story. istic estimates for financial inclusion in Africa, there Properly reflecting the true financial access of poor are bright spots. Many regulators provide public list- people in financial inclusion measurement efforts ings of regulated institutions and high-level statistics ought to include a wide range of data sources. on monetary indicators. Local and international net- Information gaps are especially prominent for works and industry associations have stepped for- sub-Saharan Africa. Africa has a diverse landscape ward to fill gaps for other financial service providers. of financial services providers—banks, credit unions, Using this information, MIX built an access-to- postal savings banks, village savings-and-loan asso- finance dataset for sub-Saharan Africa that compiles ciations, and specialized microfinance institutions. data from over 60 distinct sources. These data cover Data for these different initiatives are often held in some 23,000 providers holding 71 million accounts. different sources. Some databases on access to fi- Figure B2A shows the share of data provided by dif- nance, such as the IMF’s FAS, have only limited cov- ferent types of organizations; the main international erage in Africa. For example, only five African coun- data aggregators are grouped separately. tries—Comoros, Ethiopia, Madagascar, Mauritius, The important role of local networks and industry and Rwanda—report data on total credit outreach aggregators, such as WOCCU for credit unions and FIGURE B2a  Types of Data Sources for Sub-Saharan Africa Data source Networks GSMA WOCCU MIX WSBI SAVIX Microcredit Summit Regulators Research and news Other credit union networks Donors and funds Self-reported Technology Providers 0M 5M 10M 15M 20M 25M 0K 5K 10K 15K Number of clients Number of MFIs continued on next page 13 Box 2 continued the Savings Group Information Exchange reporting of obvious importance and also falls outside most system for community-managed microfinance, is existing surveys on access to finance. A more mean- clear from Figure B2B. Note that many of the provid- ingful picture of financial inclusion is possible only by ers covered by these sources are unregulated or in- accessing a range of data sources well beyond the formal providers. If data are grouped by the type of traditional regulated banking system. financial services provider, the same picture emerges: credit unions, savings groups, and specialized micro- This box is contributed by Scott Gaul, MIX. For further finance institutions all play a significant role in provid- information, see http://africa.mixmarket.org. ing financial services to the poor. Mobile banking is FIGURE B2B  Types of Data Sources for Sub-Saharan Africa Provider type Credit Union/Cooperative Mobile Network Operator NBFI/NGO Bank Savings Bank Savings Groups Postal Savings Banks 0M 5M 10M 15M 20M 0M 2M 4M 6M 0M 5M 10M 15M 20M Number of clients Number of borrowers Number of savers 14 4 Pa r t Improving Financial Inclusion Data F or financial inclusion data to effectively in- that will be discussed at the Los Cabos Summit in form decisions made by policy makers and fi- June 2012. In a second stage, the Data and Measure- nancial institutions, they must meet a range of ment Sub-group will develop a process for integrat- criteria. Not every country or every database will ing additional indicators as they become available meet all the criteria, but the more that can be met and standardized over time. It is preferred that each the more useful data can be. Data must be credible country takes responsibility for collecting and mon- and consistent. Inconsistencies or irregularities un- itoring its financial inclusion indicators, but the ta- dermine credibility. Data also ought to converge to- ble does list the appropriate data sources in case ward standards that apply nationally and interna- country-level data are not available. tionally. Full compliance may be illusive, but close alignment is critical so that comparisons can be Gaps in Data Collected made between countries and trends can emerge • Indicators on access and aggregate usage levels over time. Collecting data on a regular basis helps to are usually good but often leave out details on promote the standardization of report formats and customer segments, the full suite of financial ser- lowers costs over time as reporting becomes a mat- vices (e.g., insurance), and inactive (dormant) ter of routine.4 from active accounts. The analysis of global financial inclusion indica- tors reveals a wide range of different indicators, col- • There is little tracking of the quality or price of lected through various sources and often with services. slightly different definitions. Systematic gaps in the • Commercial banks are often the best document- data landscape persist. There is considerable varia- ed institutional type of provider since, as regu- tion as countries differ in their data collection efforts lated institutions, they must report to the central (IFC 2011). bank. Other kinds of organizations that include To help bring coherence and focus across coun- cooperatives, credit unions, or smaller, less for- tries and at the global level, the GPFI Data and Mea- mal organizations are typically less documented. surement Sub-group is proposing G-20 Basic Yet, in many countries, unregulated and informal Financial Inclusion Indicators that are built on the services provide the lion’s share of poor people’s AFI Core Set, a series of indicators developed joint- financial services. ly by developing country policy makers and focused on country-owned data sources.5 Although basic, • Data on access by households are more developed the indicators are selected from existing global sur- than data for firms or enterprises (see Table 1). veys that meet standards of quality, robustness, sus- tainability, and continuity. Table 2 presents the proposed G-20 Basic Financial Inclusion Indicators Challenges 4.  The UN Statistical Commission adopted the Fundamental • The people and resources to track financial in- Principles of Official Statistics in 1994, based on earlier work clusion indicators are usually limited at the by the Economic Commission for Europe, to guide the policy makers and implementing agencies. For further details see country level, leading to spotty collection and http://unstats.un.org/unsd/methods/statorg/FP-English.htm weaker quality. and http://www.imf.org/external/data.htm. 5.  For further details, see AFI Financial Inclusion Data Working • Some data sets are not publicly available (see Group (2011) and Annex I. Table 1). 15 Table 2  The Proposed G-20 Basic Financial Inclusion Indicators (as of April 2012) Existing Global / Dimension of Financial Categories Indicators Multi-country Source Inclusion Measured 1  Formally banked adults % of adults with an account at a formal financial institution Global Findex Usage Number of depositors per 1,000 adults OR number of IMF FAS deposit accounts per 1,000 adults 2  Adults with credit by % of adults with at least one loan outstanding from a Global Findex Usage regulated institutions regulated financial institution Number of borrowers per 1,000 adults OR number of IMF FAS outstanding loans per 1,000 adults 3  Formally banked enterprises % of SMEs with an account at a formal financial institution WBG Enterprise Surveys Usage Number of SMEs with deposit accounts/number of IMF FAS deposit accounts OR number of SME depositors/number of depositors 4  Enterprises with outstanding % of SMEs with an outstanding loan or line of credit WBG Enterprise Surveys Usage loan or line of credit by Number of SMEs with outstanding loans/number of IMF FAS regulated institutions outstanding loans OR number of outstanding loans to SMEs/number of outstanding loans 5 Points of service Number of branches per 100,000 adults IMF FAS Access • Lack of financial identity weakens the reliability For example, in an effort to improve data and of supply-side data on usage. As users cannot be measurement of financial inclusion, the Superinten- uniquely identified in forming country-level ag- dent of Banking, Insurance Companies, and Private gregates, supply-side indicators on usage are Pension Funds in Peru developed a set of financial prone to multiple counting. inclusion indicators for tracking the state of finan- cial inclusion in the country (see Box 3 for details). • Lack of harmonized definitions, standardized data collection, and indicator construction (for Recommendation 2 example, SMEs, active vs. dormant accounts) Use Harmonized Definitions and Standardized lead to challenges with comparability of indica- Methodologies tors over time and across countries. Harmonization of data definitions and standardiza- tion of methodologies and indicator computation are Recommendation 1 essential to ensure comparability across countries Build Country-Level Data Capacity and over time. These also enable consistency and Building or improving national capacity to meet na- transparency and help to avoid misinterpretation of tional, regional, and international data needs on data. Harmonization of definitions is especially im- financial inclusion is a critical step toward con- portant for those dimensions of financial inclusion structing a comprehensive data landscape. This is for which data and indicators are currently under especially important in countries where financial development or lacking, such as access to finance by inclusion is an explicit objective. It is also helpful to SMEs and women-owned SMEs, active versus dor- have the necessary capacity to standardize and har- mant accounts, and the quality of financial products monize data collection in line with international and services. Standardization is important for devel- norms. Investment in capacity often requires early oping common data collection methods and indica- effort and expense, but once data collection is stan- tor computation methods. Efforts for standardization dardized it can become a matter of routine, incur- may borrow existing standards/classifications from ring relatively little cost or effort, especially when similar fields. For example, FAS uses definitions and compared to the benefits of having credible data for standards consistent with the IMF’s Monetary and making decisions. Financial Statistics Manual. 16 Box 3 Financial inclusion data and measurement in Peru • Peru experienced an average annual per capita 2. Number of branches per 100,000 adults real GDP growth of 6.7 percent over 2005–2010. 3. Number of ATMs per 1,000 km2 This growth was accompanied with an expansion 4. Number of ATMs per 100,000 adults of commercial bank deposit volume and loan 5. Number of agents per 1,000 km2 volume (both as percentages of GDP) by an av- 6. Number of agents per 100,000 adults erage annual rate of 5.2 percent and 10.2 per- Usage indicators cent, respectively, over the same period. 7. Number of depositors per 1,000 adults • In spite of this progress, Peru lags behind the 8. Number of borrowers per 1,000 adults regional average in terms of financial penetra- Average deposit size as a ratio of GDP per 9.  tion, measured by deposit-to-GDP ratio. capita 10. Average loan size as a ratio of GDP per capita • To address the situation, Superintendencia de Banca, Seguros y AFP (SBS, the Superintendent Indicators on geographical inequality in terms of of Banking, Insurance Companies, and Private financial inclusion Pension Funds) in Peru developed and started Difference between participation of loans 11.  measuring a set of indicators of financial inclu- and participation of deposits originating out- sion in 2010. side of Lima (numbers) Index of total loans in provinces to total de- 12.  • The data and measurement effort aims to do the posits outside of Lima (values) following: Gini indexes for loans, deposits, and access 13.  –  Assess the depth of access and usage points – Track the trends in financial inclusion in the past decade • While the first and the second groups of indica- – Design policy measures to expand financial tors are commonly used and mostly standard- access ized indicators (based on Beck et al., 2007), the third group is developed based on country- • SBS uses 13 indicators: specific needs to assess the degree of inequality Access indicators in accessing and using financial services. 1. Number of branches per 1,000 km2 Sources: Data referenced are from the IMF FAS and World Bank Group WDI. More information on the Peruvian experience on financial inclusion data and measurement can be found in Reyes, Canote, and Mazer (2011) and Superintendencia de Banca, Segu- ros y AFP (2011). Background on indicators on access and usage can be found in Beck, Demirgüç-Kunt, and Martinez Peria (2007). Recommendation 3 provide data to a range of different authorities or Proactively Seek Data from a Range of apexes. More effort should go toward communi- Providers, Beyond Commercial Banks cating with alternate regulators, where they exist, Supply-side country-level aggregates on access to and/or data aggregators and networks, such as and usage of financial services draw heavily on MIX, SAVIX, and WOCCU, etc., that collect data data from commercial banks since these are the on certain types of institutions to complement primary providers of financial services that can be data from primary financial regulators. Additional easily tracked, often through one regulator. How- data from national associations or apexes can also ever, savings groups, financial cooperatives, and be helpful in aggregating data on certain catego- microfinance institutions are often equally impor- ries of institutions. In many markets, unregulated tant, if not more significant, sources of finance for or informal providers have substantial financial poor and low-income people. They often do not inclusion coverage. report to the main financial regulator, but instead 17 Recommendation 4 um enterprises (MSMEs), which also require a suite Use Unique Financial Identity More of financial services. However, currently, few inter- Systematically national or multi-country data collection and com- Financial identity can help supply-side data collec- pilation initiatives focus extensively on MSMEs. tion by serving as a unique identifier for counting Data on access to finance by microenterprises are the number of users of formal financial services. especially challenging as it is not easy to count such The primary functions of establishing financial enterprises. In many cases, microenterprises are un- identity are enabling access to financial services, registered businesses, and their use of formal finan- complying with know your customer (KYC) re- cial services is difficult to distinguish from personal quirements, screening, and monitoring financial ac- finance. The larger and more formal the firm, the tivities. Another important use of financial identity easier tracking data ought to be. A major challenge is enabling the aggregation of the number of users in collecting cross-country comparable data on ac- of financial services across different financial insti- cess to finance by SMEs is the lack of consensus tutions and products at the country level. In the across countries in how SMEs are defined. A vari- absence of such a unique identifier, supply-side ety of criteria is used by different countries or even data collection is prone to multiple counting, as by different agencies within one country, which households or enterprises with accounts in more are, in general, based on number of employees, as- than one bank would be counted more than once, sets, volume of sales, or loan sizes. Furthermore, leading to an over-estimation of access. An exam- within each criterion, different cutoffs are used by ple of unique identity systems includes the Aadhar different countries. For example, while the major- unique number in India, which meets KYC re- ity of countries use having less than 250 employ- quirements and could soon be linked to all individ- ees as the cutoff for an SME, some have 50 ual-level accounts. employees as the cutoff.6 Recommendation 5 Recommendation 7 Collect More Detailed Data on Customer Promote Open Access to Data Segments Ensuring open data access will lead to further Financial institutions collect a variety of informa- knowledge creation and an improved understand- tion on their clients. Mining existing data of finan- ing of problems and challenges, and as a result, bet- cial service providers can help to disaggregate ter solutions and policies. However, some existing customer segments to accompany supply-side us- financial inclusion data initiatives—both at country age data, such as gender, age, income level, occupa- level and on a global scale—are publicly unavailable tion/livelihood and combine this with usage in part if not fully. The benefits of open data access indicators across different financial services. include increased awareness and transparency by encouraging use and also greater integration of dif- Recommendation 6 ferent data sets to draw a more complete picture of Include More Firm Data, Especially That of financial inclusion. MSMEs Financial inclusion is not only about households or For a variety of SME definitions used within and across coun- 6.  individuals; it also includes micro, small, and medi- tries, see CGAP and the World Bank Group (2010). 18 5 Pa r t Conversations on Data: Five Experts Share Their Perspectives T he previous sections describe the data land- development finance sector of South Africa with scape, with a focus on supply-side data. In private and public financial institutions as well this section we learn the perspectives of five as FinMark Trust. experts who use data for practical purposes and as • Hassan Zaman is the senior economic adviser to a tool in decision making. We interviewed policy the governor at Bangladesh Bank. His responsi- makers from Asia, Africa, and Latin America and bilities include advising on financial inclusion is- two market analysts from leading international sues. Before joining Bangladesh Bank, he was groups in financial inclusion: lead economist at the World Bank. During his 13- • Diane Jocelyn Bizimana is a supervisor in the year career at the World Bank, his various re- Department of Bank and Microfinance Supervi- sponsibilities included working on microfinance sion at the Bank of the Republic of Burundi and projects in several countries. Before joining the is a member of the AFI Financial Inclusion Data World Bank, Zaman worked on microfinance is- Working Group. sues at BRAC in Bangladesh. • Raúl Hernández-Coss is director general for Access to Finance at the Mexican National Bank- ing and Securities Commission (Comisión Na- Data and transparency are vital for cional Bancaria y de Valores [CNBV]), where he promoting financial inclusion established a new area responsible for promot- ing financial inclusion. He is deputy executive secretariat for the National Council on Financial • Good data can help rally all stakeholders Inclusion of Mexico, co-chair of the Subgroup around a common goal or vision. on Data and Measurement for GPFI, and policy • Better data are vital for champion on data with AFI. – Understanding and meeting client • Marten Leijon is chief executive officer of MIX, needs which provides objective, qualified, and relevant – Building stronger business models microfinance performance data and analysis on and improving the quality of financial the institutions that provide financial services to services the world’s poor. He has many years of experi- – Developing effective markets ence in leading advisory, information, and re- – Informing evidence-based policy de- search businesses, with a primary focus on velopment financial services. – Measuring progress on financial in- clusion • David Porteous is managing director of Bank- able Frontier Associates, a consultancy firm • Data offer a factual basis for productive based in Boston. He has undertaken consultancy discussion and dialogue, setting the assignments in the areas of financial strategy and stage for analysis, consensus building, policy for a wide range of public and private sec- and informed decision making. tor clients. Before relocating to Boston in 2004, he was active in executive leadership roles in the 19 Bizimana: Unless you know what’s going on from not, and there’s no way a policy maker can make both the demand and supply sides, you can’t know policy. Transparency is important for everyone, as a policy maker, what to do. It is through data that from the regulator to the consumer. Bangladesh policy makers can comprehend how customers Bank has launched an Open Data Initiative, with perceive financial services and products offered to online access and downloadable files. There is a them, to what extent those services and products range of data (economic, exchange rates, national meet clients’ needs, and how providers can be more income data), and we will next bring in scheduled transparent to end-users. Data are “a light” to see bank statistics. Beyond this, we need to have a good where we are and where we have to go. Data collec- mapping of how the various initiatives Bangladesh tion is an issue that no country can avoid if one Bank has in place for financial inclusion is making a wants to make financial inclusion a reality. difference in access indicators plus understand why there are variations in access. We need to have wide Porteous: Getting the big picture on financial in- availability of data: having an open data initiative clusion is a bit like getting the view of the Earth requires a mindset shift. from the moon landing in 1969, which led to a whole new appreciation of the Earth as a small planet, forming the basis for the growth of the ecological Effective data collection requires movement. We can’t have a sense of proportion and starting with what you already needs without an overall view of financial inclusion and financial exclusion. have and building over time Hernández-Coss: The “why” of data collection has several angles, but there are three that are key: the • Getting started may be the hardest part; important influence of data on policy to improve fi- build on whatever you already have and nancial inclusion, support for financial institutions take an incremental approach to make in developing business models that address finan- progress. cial inclusion, and the necessary information to • Data collection costs money. Not every measure progress on the actions implemented by government will prioritize funding for authorities. financial inclusion data so, for some countries, external (donor) funding may Leijon: Collecting data is critical to enabling more be required. effective markets for funding and delivering ser- vices that meet clients’ needs and the sectors’ aspi- • Data collection requires capacity, tools, rations for access and quality. From a more and a systematic approach. practical perspective, it comes down to rallying • There is not one approach to build use- stakeholders around a goal—providing a fact base ful data—the sources, institutional part- for productive discussion, a common language, an nerships, methodologies, and choices in understanding of gaps and tools to track progress terms of breadth and depth of data differ. made. • Rapid data feedback mechanisms to test Zaman: Without the right data, you can’t know the outcomes of policy changes and make who’s included in the financial system and who’s adjustments as needed are critical. 20 Bizimana: Until now, all that the Bank of the Re- Country ownership of data public of Burundi has collected are regular data on collection, across concerned providers’ performance in compliance with the le- gal and regulatory frameworks in use. A project is agencies, is fundamental underway, funded by GIZ/AFI, that will be the first national financial inclusion survey to be completed. It will be demanding and costly, and in a country • Country ownership of data collection like Burundi, it would not be possible without ex- processes and analysis is indispensable. ternal funding. • Leadership ought to convene all national actors that can both source—and use— Hernández-Coss: What’s important is to trigger financial inclusion data. the first step. In Mexico, the creation of the Nation- al Council on Financial Inclusion puts a greater em- • The right national champion can pro- phasis on measurement, because the discussion vide leadership while fostering commit- within Mexico requires data to inform policy. We ment to data collection efforts among wanted to know where we were because we cannot the range of financial inclusion stake- advance a financial inclusion agenda without know- holders. ing more about where we are coming from. The Na- • Data can help break silos across country tional Survey for Financial Inclusion will gather structures and set the stage for open, information to create a baseline for measuring fi- fact-based conversations and consensus nancial inclusion in Mexico. Whatever the country building on policy. context, identify the sources of data that a country already has. Porteous: National surveys are not for every coun- Hernández-Coss: Financial Access 2009 (CGAP try—they are often expensive, and there are risks of and World Bank 2009) helped a lot to build aware- not doing it right. What may be more important and ness. Now countries need to take ownership of the needed is to set up rapid data feedback mechanisms process to link progress on their domestic agendas to test the outcomes of policy changes and make ad- on financial inclusion with measurement. The actu- justments as needed. To be effective, these feedback al institution collecting data doesn’t matter as long mechanisms need to be designed in conjunction as it champions the idea among other authorities. In with the policy change, not left until it is too late. our case, the president of CNBV was very support- ive. One important institutional player is the Nation- Zaman: The Institute of Microfinance conducts a al Institute of Statistics. You need to understand and demand-side survey, and the microfinance data determine the institutional arrangements in the module from the 2010 national household income country, whatever the name and position, to pro- and expenditure survey is now being exploited. We mote financial inclusion policies, which are often want to embed this data collection in the national done in silos, without overall planning. Data could statistical office. It is vital to build data collection be a means to put the cards on the table; data are into existing structures and initiatives. Ideally, part less controversial than policies and can get policy of the national data collection effort would be funded makers and regulators to start talking. by the public exchequer. The ideal would be for fi- nancial inclusion data to be given the importance of Leijon: Although there are clear benefits to fully prices, national income, money supply, but financial scaled technology and global coordination to keep inclusion will never get to this level of importance. costs down and cross-market exchange strong, ul- Donor funding is therefore necessary, for periodic timately, local ownership of issues and possible updates of the state of financial inclusion. solutions helps ensure that data inform decisions. 21 There are a wide variety of answers to who Hernández-Coss: Different dimensions of finan- should “own” related processes. Regulators have cial inclusion require different tools to measure a strong mandate in the regulated part of their progress. Access requires supply-side data. Usage market, but this can lead to an incomplete view of requires demand-side data. For financial literacy the full market. There is also a very important and consumer protection issues, demand-side data role played by local networks and associations, are not enough, focus groups or in-depth interviews across a diverse landscape. with actual users will yield far richer insights. One critical point we need to address is timing. We may Zaman: Ownership should lie with national stake- want to move faster, but we may not have the right holders. In our case, for financial inclusion broad- tools to measure. We need to enhance the data. We ly, there is a balance to be struck among the need to cross data sets, for example, branch data national statistical office, Bangladesh Bank, and with population data, identifying indicators that the Microcredit Regulatory Authority. This is a correlate poverty reduction with financial inclusion. shared responsibility. We need to rack our brains on how to do a better job of incorporating financial Leijon: We are still learning about the range of data inclusion into data collection across agencies. sets emerging across financial inclusion: their qual- ity, scope, and freshness. We need to make sure that the information is meaningful for decision-making. As data collection efforts Local data are stronger today than five years ago, by progress, the integration and far. However, improvements can be made in con- necting demand-side and supply-side data, for ex- consolidation of data sets offer ample, for addressing the goal of having a maximum greater usefulness of data distance to a bank, or the question of striking a bal- ance between mobile banking versus branches. There is an opportunity for increasing granularity • The greatest value comes from putting and integrating data with geographic overlay, such different data sets together to tell a more as linking branch-level data with physical infra- complete and coherent narrative of finan- structure details. There is movement underway to cial inclusion. have data sets speak to each other, and it needs to be accelerated. Integrating data sets is a focus of MIX. • Getting to integration and consolida- The issue is how to build bridges between islands tion is a process—you cannot move too of data in a meaningful way. It is the connection of quickly, but need to know where you data sets that will inform the broader debate. want to get. • Technology is playing an increasingly Porteous: In more and more places, my wish is not greater role and is opening the door to for more data but for better integrated, conform- new opportunities. able data that can be pieced together to form a co- herent view of what’s going on in the country. Some of the financial inclusion reports being published by central banks start this process of combining and testing various sources of data into a coherent, credible narrative of inclusion. 22 Moving from data to better policy Porteous: Governments are becoming more inter- ested in evidence-based policy making. But what is and business decision-making evidence-based policy making? Progress has been made on raw data collection; the current challenge is integrating the appropriate use of data into the • Data alone are of nominal use. The goal is policy making cycle and into product and channel to inform better policy-making and busi- design. ness decisions and drive change. One challenge is designing feedback loops so • Understanding how to integrate data that the right indicators to assess the outcomes of analysis into decision-making cycles and policy are identified upfront and then collected processes is key. It does not happen auto- and reported in a timely manner. It is no good rely- matically. ing on a triennial national survey to judge the out- come of new agent regulations, for example. These • The “providers” of data and the “users” of indicators then have to be reviewed in a disciplined data have to be in close contact. fashion to draw conclusions and make adjustments where needed. Another is building measurement into the policy process. I heard recently of how Re- serve Bank of India watched the implementation Bizimana: Policy makers will depend on results of of the business correspondent model and realized our survey to frame policy and strategies. For ex- that it wasn’t working. This allowed it to adjust the ample (1) data on the geographic coverage of finan- model. cial institutions may show that one area has a big concentration of financial institutions—this could Zaman: Bangladesh Bank has implemented finan- lead to giving no new licenses where there is too cial inclusion initiatives via the banks and is using much concentration of financial services already; data to look at who’s doing more and who’s doing (2) if it is found that a population is not using finan- less. For example, the data used to monitor the fi- cial services, this information could lead to adding nancial institutions include data with regard to “10 financial education programs to school curricula. Taka accounts.” It’s not about a target, but rather Financial service providers can see what customers having Bangladesh Bank look at the banks’ perfor- really need and what the barriers to access are (e.g., mance, and for those who are lagging, call them up the conditions to open an account). to gently encourage them to increase the number of accounts. Hernández-Coss: There is value in the data in themselves (plain data, with angles on geographic access and usage, that everyone can access online I nternational efforts play an and exploit) and in an analytical report, which be- comes a tool to disseminate information on finan- important role in promoting cial inclusion. It’s a great vehicle to keep financial data collection and use at the inclusion on the map. country level Leijon: Data matter if they inform decisions and help drive change. At some level, you have to look at • Global efforts and national efforts are mu- the locus of decisions and the locus of where data tually reinforcing. collection is owned. If there is too much distance between the two, data may be ineffective to drive • GPFI’s Data and Measurement Sub- change. Group and AFI’s data measurement ini- tiatives are paving the way to consensus on core indicators. 23 Hernández-Coss: We should move into an interna- there is a clear benefit of learning from others. tional agreement of the core set of financial inclu- Multilateral organizations shape policy. They sion indicators, and GPFI is planning to do just this, need to work from a consistent database, recog- building on the AFI Financial Inclusion Data Work- nizing patterns. International entities can ensure ing Group’s core set of indicators (see Annex I). that there is international coordination that Sharing knowledge horizontally with other coun- doesn’t hamper local initiatives, can promote fas- tries is powerful. Pulling together countries that cinating exchanges on infrastructure for aggrega- share similar challenges is effective. One example is tions, can promote the increased use of technology, the establishment of a financial inclusion report for and can develop standards. Brazil (Banco Central do Brasil 2010) following the establishment of the Mexican financial inclusion Porteous: The international institutions can cre- report (Comisión Nacional Bancaria y de Valores ate greater coincidence of interests. Countries are 2009). increasingly seeing the value of better data, so that collection is no longer forced on them by interna- Leijon: Regional and global information helps to tional bodies but also helps them directly. Interna- identify patterns, thereby driving new insights, tional organizations help ensure that surveys are helping to avoid mistakes, and identifying leverage repeated consistently to give trend data. points in multiple markets. For most countries, 24 References AFI Financial Inclusion Data Working Group. 2011. “Measuring Financial Inclusion: A Core Set of Indicators.” AFI Financial Inclusion Data Working Group Paper. http://www.afi-global.org Banco Central do Brasil. 2010. Report on Financial Inclusion. Number 1. Brasilia, Brazil: Banco Central do Brasil. Beck, T., A. Demirgüç-Kunt, and M. S. Martinez Peria. 2007. “Reaching out: Access to and Use of Banking Services across Countries.” Journal of Financial Economics 85: 234–66. CGAP. 2011. “Global Standard-Setting Bodies and the Poor: Toward Proportionate Standards and Guidance.” Washington, D.C.: CGAP. A white paper prepared by CGAP on behalf of the Global Partnership for Financial Inclusion. CGAP and the World Bank Group. 2010. Financial Access 2010: The State of Financial Inclusion through the Crisis. Washington, D.C.: CGAP and the World Bank. Comisión Nacional Bancaria y de Valores. 2009. Financial Inclusion Report. Mexico City, Mexico: Comisión Nacional Bancaria y de Valores, December. IFC. 2011. Financial Inclusion Data: Accessing the Landscape and Country-Level Target Approaches. Washington, D.C.: IFC. A discussion paper prepared by IFC on behalf of the Global Partnership for Financial Inclusion. IFC and CGAP. 2011. “Financial Inclusion Data Stocktaking and Gap Analysis.” Unpublished. Priale Reyes, G., L. Allain Canote, and R. Mazer. 2011. “Financial Inclusion Indicators for Developing Countries: The Peruvian Case.” http://www. microfinancegateway.org Superintendencia de Banca, Seguros y AFP. 2011. “The Peruvian Experience in Financial Inclusion: The Financial Supervisor’s Perspective.” Presentation at the AFI Global Policy Forum, Riviera Maya, September. http://www.afi- global.org 25 Annex I. The Afi Core Set As a first step in establishing a common and basic set of financial inclusion indi- cators, AFI Financial Inclusion Data Working Group compiled the core set of indicators in 2011. This set includes a limited number of quantitative indicators that aim to measure and track the state of access and usage dimensions of finan- cial inclusion for households. The Core Set is, by design, a limited set constructed to guide countries in col- lecting a minimum number of indicators using a common framework for in- formed policy action, but it is not a comprehensive set of financial inclusion indicators. Table A lists the core set of indicators along with definitions. Table A  AFI Core Set of Indicators Dimension Definition of dimension Core indicator Proxy indicator Definitional comments Access Ability to use formal financial Number of access points 1. Regulated access points services, i.e., minimal barriers per 10,000 adults at a national where cash-in (including to opening an account level and segmented by type deposits) and cash-out • Physical proximity and relevant administrative transactions can be • Affordability units performed. Demand Percent of administrative 2.1. side indicators of units with at least one access point distance may help here, 2.2. Percent of total population living but would be nationally in administrative points with at determined. least one access point Usage Actual usage of financial 3.1. Percent of adults with at least 3.a. Number of deposit Adult is 15 or older, or services/ products one type of regulated deposit accounts per 10,000 an age defined by • Regularity account adults country. Define active • Frequency 3.2. Percent of adults with at least 3.b. Number of loan accounts and seek to • Length of time used one type of regulated credit accounts per 10,000 measure in the future. account adults Note: Table reproduced courtesy of AFI (2011). Measuring Financial Inclusion: Core Set of Financial Inclusion Indicators. AFI Financial Inclusion Data Working Group Report, p. 3. http://www.afi-global.org 26