## Financial Inclusion in Asia‑Pacific ## Source details **Canonical URL:** [Financial Inclusion in Asia‑Pacific](https://www.imf.org/-/media/files/publications/dp/2018/46115-dp1817-financial-inclusion-in-asia-pacific.pdf) ## Other formats - [Markdown version](/-/media/files/publications/dp/2018/46115-dp1817-financial-inclusion-in-asia-pacific.pdf.md) - [Structured JSON version](/-/media/files/publications/dp/2018/46115-dp1817-financial-inclusion-in-asia-pacific.pdf.json) --- ### Executive summary and overview - Definition: financial inclusion as access to and use of formal financial services by households and firms, measured across three dimensions: (1) access to financial services, (2) usage of financial services, and (3) quality of financial products and service delivery. - Regional gaps: - "47 percent" of the adult population in low- and middle-income countries does not have a bank account. - Only "9 percent" of the population borrowed from formal financial institutions in 2016. - Estimated impacts of closing inclusion gaps: - Closing the gap between the least inclusive country and the regional median could raise GDP growth by "1 percentage point" over a five-year period. - Moving less inclusive countries to the median could reduce the number of people in poverty in the Asia‑Pacific region by "about 4 percent". - A similar improvement could lower the Gini coefficient by "about 10 to 20 percentage points" from current levels of "30 to 43 percentage points". ### Key empirical findings and quantitative impacts - Growth and inclusion: - A "1 percent" increase in the FI index in low-income developing countries is associated with a cumulative "0.14 percentage point" increase in per capita income growth over a five-year period. - The same "1 percent" improvement for high-income countries is associated with a gain of about "0.09 percentage point" over five years. - Increasing financial inclusion from the bottom quartile to the median could be accompanied by an increase of "0.2 percent" in per capita income growth over five years. - Poverty and inequality effects: - An increase in the FI index level of "1 percent" (moving from fourth to third quartile) is estimated to lower poverty by "1.4 percent" of the total population in Asia‑Pacific over five years. - Convergence to Thailand’s level could reduce the region’s population in poverty by "about 4 percent" of the region’s 2015 poverty count (equivalent to "20 million people"). - Example of inequality reduction: Lao P.D.R. improving to Thailand’s level would lower its Gini from "0.38" to "0.33". - Distributional and household impacts: - Women and young adults (age "15–24") often face greater challenges in accessing financial services. - MSMEs and farmers face limited access to credit and often rely on informal markets at high costs. ### Stylized regional facts and fintech adoption - Progress and disparities: - Asia‑Pacific has made rapid progress, but cross-country and intracountry disparities are among the highest globally. - Income is a predominant determinant; geography, financial sector structure, and policies also matter. - Pacific island countries face geographic dispersion, remoteness, small market size, and correspondingly higher delivery costs. - Fintech uptake and gaps: - China accounted for more than half of total mobile payments in the region as of end-2015. - Mobile banking uptake in most Asia‑Pacific countries lags behind sub‑Saharan Africa. - Fintech usage is uneven with large rural–urban, gender, and income gaps; fintech use among the poorest "40 percent" is significantly lower than in other regions (except sub‑Saharan Africa). - Technology access elasticities (Asia‑Pacific): - A "1 percentage point" increase in mobile phone access → "4-unit" increase in deposit accounts per 1,000 inhabitants. - A "1 percentage point" increase in internet access → about "19 out of 1,000 inhabitants" increase in deposit accounts. - A "1 percentage point" increase in internet user share → "9 additional loan accounts per 1,000 people", on average. - Leapfrogging: - Leapfrogging (mobile/fixed) tends to be negatively associated with traditional access indicators generally, but positively associated with traditional banking in Pacific island countries (example: a "25 percentage point" increase in leapfrogging associated with "+150 bank loans per 1,000 inhabitants" and "+325 deposit accounts per 1,000 inhabitants" in PICs). ### Macroeconomic interactions — monetary policy - Interest rate transmission: - Output elasticity with respect to real interest rate: - High financial inclusion group: "−0.43" (mean estimate). - Low financial inclusion group: "−0.03" (mean estimate). - Monetary shocks and inequality: - A "100-basis-point" unanticipated policy rate increase can lead to a medium-term (four-year) increase in inequality of "3 percent" (Gini) on average. - Reestimation with FI index shows inequality can rise by as much as "5 percent" in response to a "100-basis-point" policy rate increase for higher-inclusion countries. - Financial inclusion also amplifies the impact of expansionary monetary policy in reducing inequality. - Monetary frameworks: - Higher financial inclusion correlates with central banks placing greater weight on inflation stabilization versus output stabilization (more use of a policy interest rate and formal inflation objectives). - Policy trade-offs and structural reforms recommended for monetary authorities: - Strengthen asset registries and expand credit bureau coverage. - Enhance payment systems and consumer/SME/farmer financial literacy and reporting. - Reinforce bank regulation and supervision and increase transparency. - Remove financial repression to allow interest rates to allocate resources more effectively. - Recognize that higher inclusion accentuates distributional impacts when policy tightens. ### Macroeconomic interactions — fiscal policy and fiscal multipliers - Channels by which inclusion affects fiscal effectiveness: - Payments: improve targeting and reduce leakage; reduce tax avoidance. - Savings and insurance: reduce pressure on social safety nets and encourage business activity. - Credit: unlock private complementary investments and reduce credit constraints for productive investment. - Empirical associations: - A country in the highest quartile of financial inclusion (access: ATMs and branches) has on average "2.4 percent of GDP" higher revenue compared with a country in the bottom quartile, controlling for income. - Financial inclusion correlates with higher VAT efficiency and higher social tax/expenditure efficiency in Asia‑Pacific based on usage measures. - Fiscal multiplier amplification: - Fiscal multiplier estimates vary by country: example shows Malaysia (51 ATMs or 81 cash‑out points per 100,000 adults in 2015) has an estimated fiscal multiplier almost "twice" that of Nepal (9 ATMs), with cumulative effect nearly "three times" in following two periods. - Regression estimates (selected coefficients — fiscal outcomes): - ATMs per 100,000 People → Revenue (% of GDP): "0.0748***" (se "0.0118"); Expenditure (% of GDP): "0.129***" (se "0.0112"); Tax Efficiency (VAT, %): "0.405***" (se "0.0388"). - Branches per 100,000 People → Revenue (% of GDP): "0.0781***" (se "0.0163"); Expenditure: "0.115***" (se "0.0158"); Tax Efficiency: "0.496***" (se "0.0627"). - Depositors per 1,000 People → Revenue (% of GDP): "0.00272***" (se "0.000485"); Expenditure: "0.00419***" (se "0.000478"); Tax Efficiency: "0.0162***" (se "0.001980"). - Borrowers per 1,000 People → Revenue (% of GDP): "0.0106***" (se "0.00245"); Expenditure: "0.0188***" (se "0.00223"); Tax Efficiency: "0.0705***" (se "0.00749"). - Cautions: results are correlations; endogeneity and data coverage limitations mean causal claims are limited. ### Policy recommendations and elements of a holistic approach - Core priorities: - Make financial inclusion part of national inclusive growth strategies, targeting vulnerable groups: rural populations, low-income households, the elderly, women, and geographically remote communities. - National financial inclusion strategies should identify gaps, improve monitoring, strengthen interagency coordination, and prioritize reforms and resource allocation. - Structural and regulatory reforms: - Establish or expand credit bureaus and asset registries. - Improve payment systems and micro-finance regulation. - Strengthen bank regulation and supervision to reduce risks from rapid credit growth. - Consider liberalization and competition in telecom and internet sectors to lower connectivity costs. - Fintech and technology policy: - Integrate fintech into national financial inclusion strategies and link fintech to broader development initiatives. - Ensure fintech entrants operate under risk‑based regulation; empower relevant regulators and address jurisdictional challenges. - Maintain AML/CFT standards; adapt KYC for digital channels while preserving safeguards. - Combine consumer protection with financial and technology literacy programs. - Collect and use timely granular fintech data (mobile payments volumes, digital firm types) to supplement conventional inclusion indicators. - Fiscal and social policy complementarities: - Use fiscal policy to amplify inclusion benefits (e.g., channel social transfers through accounts to improve targeting and generate savings). - Revisit subsidized or directed credit programs carefully; consider gender budgeting, microfinance support, and targeted subsidies like crop insurance where justified. - Capacity building and experimentation: - Allocate resources to financial and technology literacy for consumers and providers. - Encourage social experimentation and public–private partnerships to pilot grassroots solutions and scale promising models. - International cooperation and IMF technical support: IMF to provide policy advice and capacity development and expand tools for financial inclusion analysis. ### Country case highlights (selected) - Bangladesh: - Top five MFIs outstanding disbursements in 2016 equivalent to "10 percent" of private sector credit extended by banking system. - Agent banking agents increased from "51,000" in 2012 to "772,109" in 2017; active mobile money accounts from "0.5 million" to "28 million" between 2012 and 2017. - Cambodia: - MFI-assets-to-GDP rose from "4 percent" to "20 percent" between 2011 and 2016. - MFIs provided loans to over "2 million borrowers (20 percent of the adult population)" in 2016. - Share of adults borrowing from a financial institution around "28 percent"; adults saving at a financial institution only "4 percent". - Monetary system highly dollarized: "95 percent of deposits are denominated in U.S. dollars." - India: - PMJDY added "more than 280 million" bank accounts by previously unbanked individuals since August 2014; nearly half opened by females and "60 percent" in rural areas. - Access metrics (end-2015): ATMs per 100,000 adults: "2024"; Branches per 100,000 adults: "1446"; Deposit accounts per 1,000 adults: "1,542"; Loan accounts per 1,000 adults: "154". - MUDRA and PMMY initiatives target MSME credit; Stand-Up India targets loans to women, SCs and STs. - Myanmar: - Regulatory steps: 2013 Directive on Mobile Banking; 2016 Regulation on Mobile Financial Services with interoperability, agent networks, KYC, and customer protection; Wave Money licensed in 2016. - Mobile expansion presents leapfrogging opportunities but requires supervisory capacity, branch network, and financial literacy. - Samoa: - National Financial Inclusion Strategy 2017–20 launched January 2017 with priorities on access, appropriate product design, and public–private partnerships. - Lower 40th percentile: only about "one-third" have formal financial services access; youth exclusion estimated at "55 percent" excluded or relying on informal sources; youth unemployment around "16 percent". ### Agricultural insurance and fiscal cost considerations - Scale and subsidies: - China’s Agricultural Insurance Program covers "$330 billion" of agricultural production and receives "$6.3 billion" in annual subsidies. - India’s Prime Minister’s Crop Insurance Program public subsidies capped at "50 percent" of premium value; budgeted "about $1.4 billion" in 2017–18. - Actuarial performance: - Producer loss ratios often "closer to or larger than 1.0". - Example: U.S. average loss ratio of crop insurance about "2.2" (1995–20015). - Operating costs often exceed "25 percent" of premium; loss ratios above "75 percent" imply government subsidies needed. - Alternatives and recommendations: - Consider self-insurance and savings-matching schemes (farm management deposit schemes) that are less prone to moral hazard and adverse selection. - Target subsidies toward small farm holdings and pair insurance with financial literacy and account access. - Use technology to improve monitoring and reduce risk premiums. ### Data, indices, and measurement - Main data sources used: - Financial Access Survey (FAS), Global Findex, International Financial Statistics (IFS), World Economic Outlook (WEO), World Development Indicators (WDI), SWIID, GFDD, GSMA Mobile Money Metrics, and others listed in the chapter. - Financial inclusion indices: - Three indices constructed (Index 1: access/outreach; Indices 2 and 3: incorporate usage) using FAS indicators and factor analysis. - Indicators include ATMs per km2, bank branches per km2, ATMs per 100,000 adults, bank branches per 100,000 adults, bank depositors per 1,000 adults, bank borrowers per 1,000 adults. - Total countries covered vary by index (examples: Index 1 total countries (2015): "163"; Index 2 total countries (2015): "88"; Index 3 total countries (2015): "70"). - Measurement cautions: - Traditional access indicators may understate fintech-driven inclusion; fintech-specific measures and administrative data are recommended to supplement surveys. ### Limitations and research caveats - Many empirical results are associative and subject to endogeneity and data coverage limitations. - Financial inclusion indicators vary in sample coverage; more granular and timely fintech data are needed for causal inference and policy design. - Policy trade-offs exist between rapid inclusion and financial stability (e.g., rapid MFI credit growth, NPLs, foreign funding and dollarization risks). _Italic: Source: IMF staff analysis — "Financial Inclusion in Asia‑Pacific" (selected sections of DP/46115-dp1817)._ ### Executive Summary ������������������������������������������������������������������������������������������������������ ### Executive Summary ### Overview - Much progress has been made in financial inclusion globally and in Asia and the Pacific, but gaps remain large across and within countries. - Close to half of the adult population in low- and middle-income Asia-Pacific economies does not have a bank account, and less than 10 percent has borrowed from formal financial institutions. - Lack of access to formal finance leaves poor households more vulnerable to adverse shocks and poverty traps. - The paper defines financial inclusion as access to and use of formal financial services by households and firms and measures it across three dimensions: (1) access to financial services, (2) usage of financial services, and (3) the quality of financial products and service delivery. ### Key findings - Financial inclusion matters for economic growth and is associated with a reduction in poverty and inequality. - Empirical results indicate potentially significant growth benefits from financial inclusion, with the largest gains for low-income and developing countries. - Financial inclusion can enhance central banks’ ability to stabilize economic activity and can affect fiscal policy effectiveness and public financial management. - Fintech is playing a growing role in improving financial inclusion and should be an important component of a national financial inclusion strategy. - Asia-Pacific’s use of fintech is uneven, with large gaps between the rich and poor, men and women, and rural and urban areas. ### Quantitative impacts and empirical results - Today, 47 percent of the adult population in low- and middle-income countries still does not have a bank account, and only 9 percent of the population borrowed from formal financial institutions in 2016. - Closing the gap between the least financially inclusive country and the median country in the region could raise GDP growth by 1 percentage point over a five-year period. - Moving less inclusive countries to the median could reduce the number of people in poverty in the Asia-Pacific region by about 4 percent. - A similar improvement in inclusion could lower the Gini coefficient by about 10 to 20 percentage points from current levels of 30 to 43 percentage points. - Women and young adults (age 15–24) often face greater challenges in accessing financial services. - Many micro, small, and medium enterprises (MSMEs) and farmers face limited access to credit and often rely on informal markets at high costs. ### Stylized regional facts - Asia-Pacific has made significant progress and the pace has been rapid, but cross-country and intracountry disparities are among the highest in the world. - Income is a predominant determinant of the level of financial inclusion; geography, financial sector structure, and policies also play important roles. - Pacific island countries face unique challenges because of geographic dispersion, remoteness, and small market size. - Some Asia-Pacific countries lead in fintech, but on average the region lags others in online (internet) purchases, electronic payments, mobile money, and mobile government transfers. - Mobile banking uptake in most Asia-Pacific countries lags behind sub-Saharan Africa, where mobile transactions are heavily used in countries like Kenya, Uganda, Tanzania, and Zimbabwe. ### Fintech and technology - Adoption of digital financial services has picked up in many countries (EFTPOS-supported branchless banking, mobile banking, and mobile money), with notable growth in Bangladesh, Indonesia, and Mongolia. - Mobile-based financial products have seen substantial uptake in some Pacific island countries, such as Samoa. - Fintech should be leveraged as part of national development strategies to boost resources and align with other development initiatives. - Policymakers and regulators must coordinate to strike the right balance between encouraging fintech innovations and ensuring financial stability. ### Policy recommendations and elements of a holistic approach - Financial inclusion should be part of a country’s inclusive growth strategy, targeting vulnerable groups: the rural population, low-income households, the elderly, women, and geographically remote communities. - Structural reforms and financial sector policies should remove obstacles to access: improvements in credit bureaus, asset registration, payment systems, and micro-finance institutions can reduce costs and increase transparency. - Invest in technology infrastructure (internet and mobile phone connectivity) and consider liberalization of telecommunications and internet industries to reduce costs and improve services. - Leverage fintech innovations while tilting policies toward disadvantaged groups to close digital divides; foster close collaboration with the private sector. - Allocate more resources to financial and technology literacy for both service suppliers and consumers to increase demand, close inclusion gaps, and mitigate risks from greater exposure to financial services and technology. - Encourage social experimentation and public–private partnerships to foster innovation and grassroots solutions for financial inclusion. - Macroeconomic and financial inclusion policies should consider their interlinkages: financial inclusion can strengthen the interest rate transmission mechanism, enhance monetary policy effectiveness, affect distributional impacts, boost revenue collection, improve spending efficiency, and raise the fiscal multiplier. *Executive Summary — Financial Inclusion in Asia-Pacific* ### 1. Rapid Growth of Financial Inclusion2. Distribution of Financial Inclusion Outcomes ### 1. Rapid Growth of Financial Inclusion2. Distribution of Financial Inclusion Outcomes ### Income level and determinants of financial inclusion - Income level is a key determinant of financial inclusion, but banking structure, geography, and policy also play a role. - Examples from the source: - Per capita income in India is higher than in Cambodia; a higher percentage of people in India have bank accounts than in Cambodia; Cambodia has greater mobile payments inclusion due to a strong public-private partnership. - Indonesia and Sri Lanka have similar per capita income levels, but Sri Lanka has used the formal banking system for inclusion while Indonesia has relied more on mobile banking because of greater geographic dispersion. ### Challenges for small states and Pacific island countries (PICs) - Small, widely dispersed island states face: - Expensive delivery of financial services due to sparse populations and geographic dispersion. - Severe infrastructure gaps, narrow production bases, high dependence on imports increasing transaction costs. - Heightened vulnerability to natural disasters raising the risk premium on financial services. - Difficulty maintaining correspondent banking relationships, increasing costs and complexities in transferring money and remittances. ### Within-country disparities in financial inclusion - Large gaps remain between rich and poor, urban and rural, and men and women: - Indonesia: about 10 percent of adults from the poorest quintile have a formal bank account, compared with about 60 percent from the richest quintile. - India: about 46 percent of male adults from the poorest quintile have a formal account, compared with 79 percent from the richest quintile. - India’s disparity is larger for use of mobile transactions (fourfold difference) and borrowings from a financial institution (about threefold difference). - South Asia: less than 40 percent of women have a bank account, compared with nearly 60 percent of men. - Concentration of financial exclusion among the young, uneducated, unemployed, and the rural poor is highlighted as a major concern. ### Stylized facts: financial inclusion index and country rankings (2015) - The FI index is based on 3 indicators: ATMs per 100,000 adults, Bank branches per 100,000 adults, and Bank branches per 1,000 km2; it mainly captures financial access. - Selected country quartile examples (Table 1, ranking 2015): - Fourth Quartile: Singapore 1, South Korea 2, Japan 3, Maldives 4, Thailand 5, Brunei 6, Indonesia 7, Tonga 8, India 9, China 10, Sri Lanka 11, Philippines 12, Bangladesh 13, Samoa 14, Malaysia 15. - Second Quartile: New Zealand 16, Marshall Islands 17, Fiji 18, Vietnam 19, Australia 20, Vanuatu 21, Nepal 22, Mongolia 23, Bhutan 24, Cambodia 25, Timor-Leste 26, Lao P.D.R. 27, Solomon Islands 28, Myanmar 29, Papua New Guinea 30. - Malaysia represents the median level of financial inclusion among countries in Asia. ### Empirical relationships: financial inclusion, growth, poverty, and inequality - Financial inclusion and growth: - A 1 percent increase in the FI index in low-income developing countries (LIDCs), equivalent to improving from the fourth quartile to the third quartile (example: from Cambodia’s level to Vietnam’s), is associated with a cumulative 0.14 percentage point increase in per capita income growth over a five-year period. - The same 1 percent improvement for high-income countries is associated with a gain of about 0.09 percentage point over five years. - Increasing financial inclusion from the bottom quartile to the median could be accompanied by an increase of 0.2 percent in per capita income growth over a five-year period. - Financial inclusion and poverty: - An increase in the FI index level of 1 percent (moving from fourth to third quartile) is estimated to lower the poverty level by 1.4 percent of the total population in Asia-Pacific over a five-year period. - If all Asia-Pacific countries with low financial inclusion improved to Thailand’s level (Asia’s emerging market frontier), the region’s population in poverty could fall by about 4 percent of the region’s 2015 poverty count (equivalent to 20 million people). - If less developed countries converged to Malaysia’s median level, this could be associated with a decline in the region’s population in poverty by 1 million. - Financial inclusion and inequality: - Closing gaps relative to Singapore’s level could narrow intracountry inequality, with the Gini coefficient falling by as much as 10–20 percentage points (from current levels of 30–43 points) depending on starting points. - Example: If Lao P.D.R. improved financial inclusion to Thailand’s level, its Gini coefficient would be restored from the current level of 0.38 to 0.33. - The inequality-reducing effects are mainly driven by improvements at lower income percentiles. ### Macroeconomic interactions: monetary policy and financial inclusion - Financial inclusion affects monetary policy transmission: - Greater inclusion increases the role of interest rates; more economic agents save and borrow formally, making behavior more sensitive to interest rates. - Estimated output elasticity with respect to the real interest rate: - High financial inclusion group: −0.43 (mean estimate). - Low financial inclusion group: −0.03 (mean estimate). - Results are robust to controls for economic development, openness, and financial depth. - Distributional impact and monetary policy: - The distributional impact of monetary policy is ambiguous through different channels (payments, savings, credit, insurance). - Model simulations show that a 100-basis-point unanticipated policy rate increase can lead to a medium-term (four-year) increase in inequality of 3 percent as measured by the Gini coefficient. - Expansionary shocks are found to be more powerful in reducing inequality while contractionary shocks are less powerful in increasing inequality. ### Policy implications and priorities - Financial inclusion should be part of the inclusive growth agenda for Asia-Pacific countries. - National financial inclusion strategies can: - Identify inclusion gaps, improve monitoring, strengthen national focus, and facilitate interagency coordination. - Help prioritize reforms and resource allocation and exploit policy synergies, including interactions between macro and financial inclusion policies. - Strategies should target vulnerable groups such as the rural population, low-income households, women, and remote communities and will be country-specific. *Source: IMF staff analysis in "Financial Inclusion in Asia-Pacific" (selected sections).* ### 2.5 percent in the United States. These estimates are, however, larger in mag- ### Financial Inclusion in Asia-Pacific — Interactions between Macroeconomic Policies and Financial Inclusion ### Financial inclusion, monetary policy shocks, and inequality - Estimates of the impact of monetary policy shocks on inequality: 2.5 percent in the United States; prior estimates for the United States by Coibion and others (2012) of 1.1 to 1.5 percent are smaller. - The impact of monetary policy shocks on inequality increases with the level of financial inclusion: - Reestimation with the FI index shows inequality can rise by as much as 5 percent, compared with 3 percent on average, in response to a 100-basis-point increase in the policy rate. - For high financial inclusion countries, the average decline in the Gini coefficient in Asia over the past three years was about 5 percentage points (to put the 5 percent impact in perspective). - Financial inclusion also amplifies the impact of expansionary monetary policy in reducing inequality. - Financial deepening (measured as credit-to-GDP ratio) similarly amplifies the impact of monetary policy on inequality: - A 100-basis-point increase in the policy rate increases inequality by about 5 percent when financial depth is high. - With low financial depth, monetary policy has an insignificant impact on inequality. - The largest impact on inequality occurs when both financial depth and financial inclusion are high, but effects are not additive, suggesting overlapping mechanisms. ### Link between financial inclusion and monetary policy frameworks - Theoretical and empirical findings: - Mehrotra and Yetman (2014) show that higher financial inclusion shifts welfare-maximizing monetary policy toward preferring inflation stabilization over output stabilization; financially included households can smooth consumption, making inflation stabilization welfare-maximizing even at the expense of higher output variability. - More robust regressions confirm: - (1) Higher financial inclusion has a statistically significant and positive relationship with minimizing inflation volatility over output volatility (stronger central bank preference for price stability). - (2) Higher financial depth (credit-to-GDP ratio) is also associated with greater price stability relative to output stability. - Despite the increasing variance ratio, both inflation variability and output variability decline as financial inclusion increases; declines are also related to the overall stage of economic development. - Asia-Pacific country patterns: - Higher financial inclusion is linked to more active use of a policy interest rate and the adoption of price stability and inflation objectives. - Excluding dollarized or currency board regimes, countries with high financial inclusion tend to have more formal and quantified inflation targets and rely more on a policy interest rate as the main instrument. - Countries with lower financial inclusion tend to have more broadly specified monetary frameworks with multiple objectives. - Financial inclusion may be a factor or condition for transitioning to a more price-based inflation-targeting monetary framework. ### Policy implications for monetary authorities - Financial inclusion can enhance central banks’ ability to stabilize economic activity by improving interest rate transmission and reducing reliance on direct/quantitative interventions for countercyclical operations. - Trade-offs: - Financial inclusion accentuates the distributional impact of monetary policy; when central banks tighten policy, adverse impacts on income distribution tend to be amplified as financial inclusion increases. - Structural reforms to enhance both financial inclusion and monetary policy effectiveness: - Establish asset registries. - Expand client coverage of credit bureaus (especially to less established clients). - Enhance payment systems. - Reinforce reforms with financial literacy and reporting for consumers, MSMEs, and farmers. - Strengthen bank regulation and supervision and increase transparency of financial products and services to boost confidence in formal banking. - Remove financial repression and allow interest rates to play a greater role in resource allocation. - Fiscal policy, public financial management, and macroprudential measures can complement structural policies and financial inclusion. ### Interactions between fiscal policy and financial inclusion - Channels through which financial inclusion affects fiscal policy effectiveness (Table 3 summary): - Payments: - Expenditure: Improve targeting; reduce leakage. - Revenue: Reduce tax avoidance. - Savings: - Expenditure: Reduce pressure on social safety net. - Revenue: Reduce tax avoidance. - Credit: - Expenditure: Unlock private complementary investments. - Revenue: Reduce credit constraints for productive investment. - Insurance: - Expenditure: Reduce pressure on social safety net. - Revenue: Encourage business activity. - Indirect channels: - Savings accounts provide less risky, more liquid vehicles for asset accumulation, cushioning shocks and reducing pressure on social safety nets. - Expanded access to credit helps finance private spending on health and education, smooth consumption, and reduce borrowing constraints, thereby increasing productive investment and growth and raising tax revenue. - Insurance services improve risk management and can raise tax revenue. - Greater financial inclusion tends to reduce informal and shadow economic activity, bringing transactions into the formal economy for measurement and taxation. ### Macrofiscal impact and empirical evidence - Simple correlations: - Financial inclusion is associated with higher revenues and expenditure as a share of GDP. - Relationship is broadly stable whether financial inclusion measured by access (share of population with bank accounts) or usage (bank loans per person). - For Asia-Pacific, relationship seems stronger with revenue than with expenditures, based on usage. - Higher financial inclusion is associated with higher VAT efficiency and higher social tax and expenditure efficiency in Asia-Pacific (based on usage), though not always stronger relative to other regions. - More rigorous analysis (OLS with controls for income level and Asia-Pacific-specific factors; robustness checks use other inclusion metrics): - Financial inclusion measured by ATMs or branches per 100,000 people is correlated with higher revenue, expenditure, revenue efficiency, and in some cases expenditure efficiency. - For Asia-Pacific the positive relationship remains significant but at a lower level than in the rest of the world. - Illustration: A country in the highest quartile of financial inclusion (based on access to ATMs and bank branches) has on average 2.4 percent of GDP higher revenue compared with a country in the bottom quartile, after controlling for income. - Financial inclusion and fiscal multipliers: - Financial inclusion can amplify the fiscal multiplier via enhanced public expenditure management or revenue administration. - Empirical example: Estimated fiscal multiplier for Malaysia (51 ATMs [or 81 cash-out points] per 100,000 adults in 2015) is almost twice the size of that for Nepal (9 ATMs). The cumulative effect increases to nearly three times the size for the following two periods. - Finding robust across several measures of financial inclusion. - Note on empirical methods and caveats: - Panel regression control variables include trade openness, exchange rate regime, inflation, level of development, financial depth, capital account openness, Gini coefficient, Asia dummy variables, interaction terms, and fixed effects. - Results are correlations; endogeneity issues may remain, making causal identification challenging. ### Policy implications for fiscal authorities - Virtuous circle: Greater financial inclusion enhances fiscal policy effectiveness through higher revenue collection, better targeting of government expenditure, and improved efficiency; fiscal policy and better public financial management can facilitate financial inclusion. - Fiscal policy design should account for synergies with financial inclusion when allocating budget resources. - Integrate financial inclusion with health and education priorities to reduce income, regional, and gender inequality. - Experimentation opportunities and trade-offs: - Traditional tools like subsidized or directed credit have had mixed results. - Other interventions include gender budgeting, promoting micro-finance and financial literacy, and subsidizing activities such as crop insurance. - Example: China has used tax benefits to promote financial inclusion of micro, small, and medium-scale enterprises. - Such interventions have public resource implications and trade-offs that must be considered. *Source: IMF staff chapter “Interactions between Macroeconomic Policies and Financial Inclusion,” Financial Inclusion in Asia-Pacific.* ### 1. Impact of Financial Inclusion on Fiscal Multiplier ### 1. Impact of Financial Inclusion on Fiscal Multiplier ### Fintech use and technology access: stylized facts - Fintech is playing a growing role in improving financial inclusion by reducing costs and achieving economies of scale, making services commercially viable for previously excluded groups. - Technology access has grown rapidly in Asia-Pacific over the past decade, but on a per capita basis Emerging Market and Developing countries in Asia-Pacific have lower access to technology than counterparts in Europe, Latin America, and small Caribbean island economies; only sub-Saharan Africa fares worse. - Pacific island countries have average per capita mobile phone subscriptions lower than those of sub-Saharan Africa despite recent rapid mobile infrastructure growth. - Asia-Pacific—especially Pacific island countries and low-income and developing countries—have exhibited a greater tendency to “leapfrog” in technology (mobile phone subscriptions relative to fixed line subscriptions). - Meaningful use of the internet to make online purchases is observed only in countries with relatively high internet penetration—at least 60 percent of the population. - China is a global leader in mobile payments, accounting for more than half of total mobile payments in the region as of the end of 2015. - Mobile money and mobile banking have grown markedly in Asia-Pacific; the region is a major center for mobile money service providers and users, second only to sub-Saharan Africa. - Usage gaps remain large: - Rural–urban gap in mobile financial transactions is most stark in Asia-Pacific. - Fintech use among the poorest 40 percent is significantly lower than other regions (except sub-Saharan Africa). ### Empirical evidence on technology’s impact on traditional financial access - Cellular and internet access: - A 1 percentage point increase in mobile phone access is associated with a 4-unit increase in deposit accounts per 1,000 inhabitants in Asia-Pacific. - A 1 percentage point increase in internet access is associated with an increase of deposit accounts in Asia-Pacific by about 19 out of 1,000 inhabitants. - A 1 percentage point increase in the internet user share is associated with 9 additional loan accounts per 1,000 people, on average. - Leapfrogging (mobile as percentage of fixed line) and traditional indicators: - Leapfrogging is negatively related to traditional financial access indicators (number of ATMs, bank branches, and bank accounts per capita) in general, suggesting deposits and loans may move outside the banking system as technology rises. - In Pacific island countries, leapfrogging has a positive impact on traditional banking services: - A 25 percentage point increase in the leapfrogging variable (equivalent to raising the mobile-to-fixed-line subscription ratio from Samoa to Fiji) is associated with an increase of 150 bank loans per 1,000 inhabitants and 325 deposit accounts per 1,000 inhabitants. - Leapfrogging has less impact on the number of ATMs and bank branches in Pacific island countries. - Interpretation: - Broader access to technology helps expand traditional financial services and achieve greater financial inclusion, but overreliance on traditional indicators may mislead when fintech shifts activity outside banks. ### Policy implications and recommendations - Integrate fintech into national financial inclusion strategies: - Link financial inclusion to broader national development strategies to secure resources and align fintech strategies with other development initiatives. - Ensure a fintech strategy is endorsed across government agencies to address emerging regulatory issues. - Invest in technology infrastructure and promote competition: - Continued investment in technology infrastructure is crucial to spread fintech benefits and close cross-country and intracountry gaps, especially for Asian LIDCs and PICs with large infrastructure gaps. - Ensure strong competition in telecom and internet industries and provide support to extend services to disadvantaged groups. - Regulatory and supervisory priorities (key themes for consideration): - Establishing oversight: - Ensure all new fintech entrants are subject to risk-based regulations. - Ensure a relevant regulator is legally empowered and has capacity to provide oversight, acknowledging jurisdictional challenges posed by fintech. - Know your customer (AML/CFT): - Maintain established Anti Money Laundering and Combating the Financing of Terrorism provisions; peer-to-peer lending, blockchain, and electronic payments need special attention. - Require accurate customer recording and clear guidance on correspondent jurisdiction requirements. - Consumer protection: - Combine effective regulations with financial and technological literacy programs to protect vulnerable consumers. - Ensure resolution procedures for failed fintech companies to avoid long-lasting barriers to inclusion from disorganized collapses. - Address additional consumer protection challenges related to cybersecurity and financial data proliferation. - Data and monitoring: - Better and more timely data on use of financial services—including granular data on mobile payments, digital economy, types and value of fintech companies, and fintech transaction volumes—are crucial for regulators to understand business models and assess risks. - Conventional indicators of financial inclusion are inadequate to gauge fintech impacts and should be supplemented with fintech-specific measures. *Source: 46115-dp1817-financial-inclusion-in-asia-pacific, Chapter 1.* ### Box 1. inclusive Fintech: Emerging issues (continued) ### Box 1. inclusive Fintech: Emerging issues (continued) ### Impact of Technology on Financial Inclusion — Overview - Asia-Pacific countries have made great progress in financial inclusion, but large disparities across and within countries remain. - Large cross-country disparities reflect geography, financial sector structure, and policies, as well as income diversity in the region. - Large within-country disparities disadvantage women, the less educated, small businesses, low-income households, and remote and rural populations. - Gaps extend to the access and usage of fintech, which holds much promise for advancing financial inclusion. - Furthering financial inclusion in the Asia-Pacific region will require leveling the playing field for its most disadvantaged groups, particularly remote communities in Pacific island countries and in low-income and developing countries. - A holistic approach is recommended that encompasses simultaneous actions across macroeconomic, structural, technological, and financial policies, and efforts to improve financial and technology literacy and infrastructure. ### Bangladesh: Building on Success - Historical focus: access to bank accounts and credit for underserved groups, centered on microcredit. - In 2016, outstanding disbursements by the top five micro-financial institutions (MFIs) were equivalent to 10 percent of private sector credit extended by the banking system. - Policy measures broadened over time: no-frill accounts, agent-based banking and mobile financial services, requirement that banks open at least 50 percent of new branches in rural areas, floors on credit to the agricultural and rural sectors backed by credit refinancing lines on concessional terms, and support to MSMEs and women entrepreneurs. - Draft financial inclusion strategy priorities: credit growth in priority sectors, MSME financing, gender priority, promoting access to finance in rural areas, green financing, a life cycle approach toward inclusion, and insurance. - Plans to leverage technology by building a digital ecosystem to make financial services more accessible for the entire population. - Persisting challenges: high level of informal finance, low use of credit and debit cards, and low use of bank accounts. - Suggested measures drawn from other countries: strengthening legal and regulatory framework, establishing a financial ombudsman, requiring payment of subsidized energy via a bank account (piloted elsewhere), and keeping track of progress with tools such as a financial inclusion index. ### Cambodia: Containing Risks while Pursuing Inclusion - Financial inclusion became a strategic priority and part of the Financial Sector Development Strategy 2016–25. - Authorities planning a National Financial Inclusion Strategy, led by the National Bank of Cambodia; new legislative framework for credit guarantees to facilitate access to credit. - Rapid growth in MFIs: in 2011–16, the number of MFIs almost tripled, and the MFI-assets-to-GDP ratio increased from 4 to 20 percent. - In 2016, MFIs provided loans to over 2 million borrowers (20 percent of the adult population). - Risks: MFIs may become a source of financial stability risks—relying on external sources of funding more than the banking sector and expanding loans at a faster pace due to fierce competition. - Despite progress, financial inclusion remains low and uneven: - Share of Cambodian adult population with bank accounts (including mobile accounts) is much lower than the global average of 60 percent. - In 2014 Cambodia had one of the highest shares of the adult population (28 percent) that borrowed from financial institutions, but one of the lowest (4 percent) that saved (Findex 2014). - Financial services skewed toward credit, with relatively lower usage of savings, transactions, and insurance despite rapid credit growth from a low base. ### India: From Directed Credit to a More Holistic Approach - Traditional approach: channeling credit to weaker segments via minimum priority-sector lending requirements and specialized development financial institutions. - Directed credit had limited success; trade-offs between credit-focused inclusion and financial stability. - Priority-sector lending boosted credit to priority sectors but led to higher non-performing loans. - Banks often circumvent targets by investing in other eligible instruments such as development financial institution–issued bonds. - Share of long-term credit flows to the agricultural sector declined between 2006 and 2007 and 2011 and 2012. - Recent policy shift to a more holistic approach emphasizing digital technology: - Priorities: (1) rapidly expand access to formal bank accounts, (2) create incentives for boosting transactional volumes of financial products and expanding add-on products, and (3) expand availability of credit to underserved sectors. - Integration of key enablers: Jan Dhan Yojana (access to formal accounts), Aadhaar (unique biometric identification), and reliance on mobile technologies. - Aadhaar effects: - Introduced unique identification numbers in 2010 to address inability to reliably identify customers. - Streamlined bank account opening and enabled application of “know your customer” norms through electronic platforms. - Helped curtail misuse of subsidy rolls and financial accounts; streamlined welfare program delivery, supported transparency and good governance, and enhanced coverage and usefulness of India’s Credit Bureau. ### Myanmar: Laying Foundations for Leapfrogging - Financial inclusion dominated by state-owned financial institutions in an underdeveloped market with an outdated regulatory regime. - Myanmar Agricultural Development Bank main source of rural credit. - Agricultural cooperatives provide microcredit but have limited resources and have suffered loan losses from natural disasters. - Burgeoning MFI sector limited by funding constraints. - Private banks expanded rapidly but regulations limit lending to one-year overdraft loans. - Bank credit concentrated in large enterprises and urban centers such as Yangon. - Reforms needed alongside inclusion efforts: state-owned bank reform initiated but progress slow; implementation of banking regulatory framework delayed due to capacity constraints and stability concerns. - Mobile banking prioritized: - In 2013 the Central Bank of Myanmar formulated a basic regulatory framework to allow technology service providers, financial service providers, and mobile network operators to partner with banks. - Updated framework aims to ensure interoperability, agent networks (without exclusivity), know-your-customer norms, customer due diligence, and customer protection. - Potential to leapfrog due to rapidly expanding mobile phone usage, declining telecommunication costs, and rapid smartphone penetration. - Caveat: limits to leapfrogging without physical infrastructure (bank branches, roads, electricity) and a solid banking system. ### Samoa: Building Financial Infrastructure - National Financial Inclusion Strategy 2017–20 launched January 2017 with three strategic priorities: - Make a wide range of financial products and services accessible to all, especially low-income segments. - Ensure financial service products are appropriately designed and priced to be widely used. - Achieve strong partnership and collaboration between private and public sectors for a cohesive approach. - Market-driven approach with regulatory oversight by the Central Bank of Samoa. - Drivers of financial exclusion: - Only about one-third of those in the lower 40th percentile of income have access to formal financial services, compared with two-thirds in the top 40th percentile. - Rural population and agricultural workers less likely to have access. - Young population relatively more excluded: 55 percent estimated to be either excluded entirely or relying on informal sources for financial services, reflecting a youth unemployment rate of about 16 percent. - Other barriers: high transaction costs, travel times to formal services, and a strong cultural preference for cash. - Ongoing reforms: - Establishing an effective financial consumer protection regime. - Creating an enabling environment for inclusive insurance markets and microinsurance. - Integrating financial inclusion in school curricula. - Addressing risks related to strains in correspondent banking relationships. - Supporting MSMEs: introduction of a movable property registry, commitment to establish a credit bureau, and encourage economic use of land. ### Agricultural Insurance: Inclusion with a Fiscal Cost - Agricultural insurance is an important instrument for financial inclusion; rural populations in Asia-Pacific are less likely to be financially included despite high shares of agricultural output and employment. - About half of Asia-Pacific countries have had crop insurance schemes. - Asia is the second largest market for agricultural insurance premiums after North America. - China and India have two of the largest agricultural insurance schemes globally: - China’s Agricultural Insurance Program covers $330 billion of agricultural production and receives $6.3 billion in annual subsidies. - India’s Prime Minister’s Crop Insurance Program receives public subsidies capped at 50 percent of the premium value, amounting to about $1.4 billion in 2017–18. - Agricultural insurance is usually not commercially viable and requires substantial fiscal support. - Producer loss ratios (total claims divided by total premium collected) are typically closer to or larger than 1.0. - Considering the premium covers only 25 percent of operating costs, any loss ratio higher than 75 percent implies government subsidies or other supports are required to keep insurers in the market. - Marginal subsidy rate required for increasing coverage rises because farmers’ demand for insurance tends to be inelastic with respect to premium. - Reasons for unviability and reliance on government support: - Crop risks tend to be undiversified and yields highly correlated across farms, limiting cross-sectional risk pooling. - Adverse selection and moral hazard more pronounced due to limited actuarial information. - Low financial literacy in rural areas. - Alternative schemes and approaches: - Self-insurance schemes (examples described): special savings accounts with government matching contributions (withdrawals when yields or income fall below a threshold); farm management deposit schemes with tax deferral. - Advantages of these schemes: less prone to moral hazard and adverse selection, lower management costs than standard agricultural insurance. - Implementation requires some level of financial development, challenging for many developing countries in Asia-Pacific. - Policy implications: - Need to balance costs and benefits of agricultural insurance and explore more cost-effective approaches. - Containing costs may require greater targeting (e.g., prioritize small farm holdings over large commercial plantations). - Agricultural insurance should be paired with other inclusion measures: financial literacy and access to bank accounts. - Technology may reduce risk premiums by improving monitoring and mitigating moral hazard. ### Regional Policy Recommendations and Next Steps - Pursue a holistic approach to close gaps in financial inclusion: - Build on structural reforms and financial sector infrastructure progress. - Recognize opportunities from rapid development of digital and mobile technology to reach the “last mile.” - Actions to continue and prioritize: - Strengthen financial infrastructure and regulations: establish/expand coverage of credit bureaus, enhance payment systems, improve asset registries, and strengthen financial sector transparency. - Improve physical and digital infrastructure: build affordable telecommunication and internet networks that can reach rural and remote communities. *Source: Box 1. inclusive Fintech: Emerging issues (continued), Financial Inclusion in Asia-Pacific, IMF staff.* ### Conclusions ### Conclusions ### Policy conclusions and recommendations - Financial inclusion strategies should prioritize underserved groups, including the poor, women, MSMEs, rural populations, and the urban poor. Regulatory policies should ensure strong competition in operating these networks. - Fintech should be a part of the financial inclusion strategy. Fintech has significant potential for addressing the needs of excluded groups, as seen in the use of mobile payments in overcoming barriers to financial inclusion, especially for remote communities. - Fintech policies should aim to reduce gaps in access to and usage of fintech-enabled services and avoid the digital divide. - Fintech regulation needs to strike the right balance between fostering innovation and ensuring systemic stability and consumer protection. Strengthening anti–money laundering regulations will also be important. - Improving financial and technology literacy is a priority. The focus should be on MSMEs, women, and low-income and rural populations. Financial education should include fintech innovations to keep up with technological progress. - Better and broader data on financial inclusion will strengthen monitoring and policy design. Surveys should not substitute administrative data collection, which may better capture the rapid growth of fintech and other dimensions of financial inclusion, such as gender and income distribution. - Countries should explore the feasibility of unlocking big data to better understand consumer needs, develop tailored financial services, and better assess potential borrowers’ creditworthiness. - Design policies need to consider interactions between macroeconomic policy and financial inclusion. Higher financial inclusion can increase the effectiveness of interest rate policy as well as amplify its distributional impact. Greater financial inclusion can also enhance the effectiveness of fiscal policy and help create space for needed public spending. Fiscal policy can help promote financial inclusion through targeted assistance. - Asia-Pacific’s experience suggests that each country should adopt its own holistic approach to financial inclusion. Countries should encourage social experimentation (for example, Bangladesh’s microcredit approach and East African popularization of mobile payments). - An effective holistic approach calls for joint efforts by central banks, finance ministries, regulatory bodies, telecom regulators, and education authorities, in partnership with the private sector. - The IMF can continue to support financial inclusion efforts in Asia-Pacific through policy advice and capacity development. The IMF has made financial inclusion an integral part of its policy advice and capacity development. Further work is needed to strengthen the IMF’s toolkit for financial inclusion analysis through a better understanding of the roles of macro, structural, financial, and technology policies. The IMF and other development partners should continue to collaborate closely in providing tailored policy advice and capacity development. - Malaysia, for example, has developed an online Separately Managed Account financing aggregator, which enables digital submission of documents, facilitates multiple financing applications seamlessly, and reduces the information asymmetry between banks and Small-Medium Enterprises. ### Data and measurement: sources and priorities - Main data sources used in the report: - Financial Access Survey (FAS): collected by the IMF from regulatory data submitted to central banks and other regulators, documenting the extent of financial services provided. - Global Findex: a triannual survey conducted by the World Bank, which asks respondents (sampled from the general population) on their use of financial services. - Table coverage (as presented): - Global Findex (20)Financial Access Survey (35) - Key additional data sources used in production of the report and APD Financial Inclusion Database include: - Financial Access Survey (FAS) (http:// data .imf .org/ FAS) - Global Findex (http:// www .worldbank .org/ en/ programs/ globalfindex) - International Financial Statistics (IFS) (http:// data .imf .org/ IFS) - World Economic Outlook (WEO) (http:// www .imf .org/ external/ pubs/ ft/ weo/ 2017/ 01/ weodata/ index .aspx) - World Development Indicators (WDI) (http:// data .worldbank .org/ data -catalog/ world -development -indicators) - The Standardized World Income Inequality Database (SWIID) (http:// fsolt .org/ swiid/) - IMF’s Information Notice System (INS) - Global Financial Development Database (GFDD) (http:// www .worldbank .org/ en/ publication/ gfdr/ data/ global -financial -development -database) - IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAERs) (http:// www .elibraryareaer .imf .org/ Pages/ Home .aspx) - Global Competitiveness Index (GCI) (http:// reports .weforum .org/ global -competitiveness -index/ #topic = data) - World Governance Indicators (WGI) (http:// info .worldbank .org/ governance/ wgi/ #home) - Global Microscope (Economist Intelligence Unit) (http:// www .eiu .com/ public/ thankyou _download .aspx ?activity = download & campaignid = Microscope2016) - Enterprise Surveys (World Bank) (http:// www .enterprisesurveys .org/) - Financial Development and Structure Dataset (World Bank) (http:// www .worldbank .org/ en/ publication/ gfdr/ data/ financial -structure -database) - Public Sector Ownership Data - Cull et al. IMF Working Paper No. 17/60: ’Bank Ownership: Trends and Implications’ - Consolidated Banking Statistics (Bank of International Settlements) (http:// www .bis .org/ statistics/ consstats .htm) - Global System for Mobile Communications Association (GSMA) Mobile Money Metrics Dataset (https:// www .gsma .com/ mobilemoneymetrics/ #global ?y = 2017 ?v = overview ?g = global) ### Indices and quantitative findings - An index can help combine financial inclusion (FI) indicators and provide a more rounded measure of financial inclusion. Three indices are presented, based on factor analysis and data from the Financial Access Survey (FAS). The indices show that FI in Asian and Pacific countries is diverse but improving. - Index construction process (as presented): - Data normalized for scale invariance by expressing each variable relative to its maximum value across countries. - Factor analysis used to group variables into sub-indices (factors) and to obtain factor-loadings as weights. - Sub-indices weighted and aggregated to form a composite index using each factor’s share of explained data variance as the weights. - Index design notes: - Three indices constructed with varying levels of cross-country coverage. The goal of broad country coverage for Asia limits the number of indicators that can be used. - Index 1 focuses on financial inclusion “access” or “outreach.” - Indices 2 and 3 incorporate “usage”; Index 2 uses account numbers, Index 3 uses numbers of depositors and borrowers. - A small number of high and variable FI countries may be excluded from reported index values to reduce noise (examples given: for Index 1, San Marino, Hong Kong SAR, and Macao SAR were excluded; for Index 2 San Marino and Malta were excluded). - Volatility in maximum FI indicators can induce volatility in normalized indicators and index values; indices were calculated twice (full sample for rankings; recalculated excluding a small number of high and variable FI economies for reported values). - Index specification summary (as presented in Table 2.2): - Index 1Index 2Index 3 - DimensionsN/AN/AAccess, Usage - Indicators(1) ATMs per km2(1) ATMs per km2(1) ATMs per km2 - (2) Bank branches per km2(2) Bank branches per km2(2) Bank branches per km2 - (3) ATMs per 100,000 adults(3) ATMs per 100,000 adults(3) ATMs per 100,000 adults - (4) Bank branches per 100,000 adults(4) Bank branches per 100,000 adults(4) Bank branches per 100,000 adults - (5) Bank depositors per 1,000 adults(3) Bank depositors per 1,000 adults - (6) Bank borrowers per 1,000 adults(4) Bank borrowers per 1,000 adults - Total countries (2015)1638870 - Asian countries (2015) 311913 - Retained factors 2 3 2 - Index findings and trends (summary as presented): - Financial inclusion rankings across Asia were diverse, but Asian countries have generally been stable or have improved relative to other countries. Table 6 summarizes changes in Asian country rankings and index levels by year. - In most cases, rankings improved or remained the same as the previous year, with the EMs and LICs accounting for most cases of catch up. - Index rankings were broadly sensible, but index values should be interpreted with care where efficiency gains or technological substitution may result in the same level of FI at lower values of FI indicators (numbers of ATMs or bank branches). Results for Australia and NZ may be biased by the inclusion of indicators by area, given the geographic size of these countries. - Selected index change counts (as presented in Table 2.6): - Index 1 - Total Asian Countries3030303128303030 - Higher than previous year1410181810122223 - Same as previous year 4 9 5 6 0 0 0 0 - Lower than previous year1211 7 71818 8 7 - Index 2 - Total Asian Countries1718191915171819 - Higher than previous year 3 41414 6101314 - Same as previous year 3 4 3 1 0 1 1 0 - Lower than previous year1110 2 4 9 6 4 5 - Index 3 - Total Asian Countries11121313 8111213 - Higher than previous year 1 2 3 9 6 510 7 - Same as previous year 4 8 4 3 1 0 0 0 - Lower than previous year 6 2 6 1 1 6 2 6 _Conclusions (Chapter 7), Appendix 1 and Appendix 2 material from the provided IMF content unit._ ### Appendix 2. Financial Inclusion Indices for Asia-Pacific Country Analysis ### Appendix 2. Financial Inclusion Indices for Asia-Pacific Country Analysis ### Empirical approach and sample - Empirical approach follows IMF (2015), Park and Mercado (2016), and Honohan (2008). - Growth regressions follow standard growth regression as in Barro and Sala-i-Martin (1992) and Aghion and others (2005). - Financial inclusion is represented by the financial inclusion index type-1 constructed in Appendix I. - Poverty and income inequality regressions follow specifications similar to Honohan (2008) and Park and Mercado (2016). - Dependent variables: - Poverty measures: ratio of the number of people whose income falls below the poverty line, and percent of population living under 1.90 and 3.10 a day in 2011 dollars in terms of purchasing power parity. - Income inequality: GINI coefficients and income gaps as ratios of income levels at top and bottom 10 and 20 percentiles. - Sample: includes 188 countries covering the period between 1990 and 2015; five-year averages of all variables are used to smooth out cyclical variations. Panel is unbalanced. - Estimation: General Method of Moment (GMM) estimation with instrumenting (STATA code uses Roodman (2009)); instrumental variables include lags of dependent and explanatory variables. Constant term, country and time fixed effects included. Standard serial correlation and identification tests performed; tests robust to choice of instruments. Number of observations for each country is less than 10. ### Key estimation results (Table 3.1 — global sample) - Growth of per capita income (specifications (1)–(4)): - Initial income: 20.018***, 20.017***, 20.018***, 20.017*** (t-statistics shown in table). - Financial inclusion: 0.014***, 0.010***, 0.013***, 0.011***. - Financial inclusion*per capita income: 20.000**, 20.000***, 20.000*, 20.000***. - Education: 0.025***, 0.026***, 0.022***, 0.026***. - Financial development (selected coefficients): 20.026*, 20.040***, 20.018, 20.044***. - Poverty (ratio below poverty line) (specifications (5)–(8)): - Initial income: 20.584, 0.037, 0.943, 1.519. - Financial inclusion: 21.400***, 21.372***, 21.357***, 21.206**. - Financial inclusion*per capita income: 20.000, 0.000*, 20.000, 0.000. - Education: 27.835*, 210.632*, 210.522*, 213.126**. - Income inequality (GINI) (specifications (9)–(12)): - Initial income: 4.012*, 3.990, 5.255**, 5.061**. - Financial inclusion: 22.637**, 22.241*, 23.107*, 22.530*. - Financial development (selected): 10.087*, 10.299**, 9.887**, 9.048**. - Regression diagnostics and sample sizes reported in table (e.g., Observations: Growth regressions 428; Poverty regressions 204; Inequality regressions 312). AR2 and Hansen test p-values reported per specification in table. ### Asia-specific estimation results (Table 3.2 — Asia sample / Asia dummies) - Growth of per capita income (specifications (1)–(4)): - Initial income: 2 0.016***, 2 0.015***, 2 0.015***, 2 0.014*** (table presents these values exactly as shown). - Financial inclusion: 0.013***, 0.008***, 0.011***, 0.008***. - Financial inclusion*per capita income: 2 0.000, 2 0.000*, 2 0.000*, 2 0.000* (as presented). - Education: 0.023***, 0.022***, 0.019***, 0.019***. - Poverty (specifications (5)–(8)): - Financial inclusion: 2 1.490***, 2 1.490***, 2 1.442, 2 1.437***. - Financial development: 2 0.000, 2 0.000***, 2 0.000, 2 0.000***. - Asian dummy and interaction terms reported in table (coefficients and t-statistics as shown). - Income inequality (specifications (9)–(12)): - Financial inclusion: 2 3.292**, 2 2.609**, 2 3.379**, 2 2.734*. - Education: 2 9.394***, 2 9.268***, 2 9.225***, 2 9.065***. - Observations, AR2, Hansen p-values and other diagnostics are reported in table for each specification. ### Financial inclusion and interest rate transmission (Appendix 4 summary) - Method for transmission effectiveness follows Mehrotra and Nadhanael (2016) and Bilbiie and Straub (2012); estimate real interest rate sensitivity of output using Euler equations across different financial inclusion levels. - Two Euler specifications: - Hybrid model (with persistence/habit formation): y_t = a + a1 y_{t-1} + a2 y_{t-2} + m E_t y_{t+1} + b(i_t − E_t π_{t+1}) + ε_t. - Forward-looking no-persistence model: y = a + m E_t y_{t+1} + b(i_t − E_t π_{t+1}) + ε_t. - Sample for panel Euler regressions (Asia): Australia, Bangladesh, China, Fiji, Hong Kong SAR, India, Indonesia, Japan, Korea, Malaysia, New Zealand, Philippines, Singapore, Taiwan Province of China, Thailand. Countries grouped by whether they are above or below the average on the FI index (results robust to grouping using median). - Output equations estimated separately for each group after controlling for level of economic development, openness, and financial depth. Coefficients for the two groups are found to be significantly different in both Euler specifications. ### Monetary policy shocks and inequality — data and methodology - Data: - Sample 2004–2013, 33 advanced and emerging market economies. - Income inequality: SWIID 5.1 net Gini (post-tax, post-transfers). - Policy rate (or 3-month short term rate where policy rate not available): Haver Analytics. - Inflation and output growth: IMF WEO database. - Forecast errors for interest rates, inflation and output growth computed using forecasts from Consensus Economics. - Financial inclusion index: staff calculations including financial access and inclusion variables from FAS database. - Constructing exogenous monetary policy shocks: - Forecast error of policy rate: FE_t^i = SR_t^i − CR_t^i (SR = actual policy or 3-month short rate; CR = Consensus Economics forecast as of October of same year). - For each country, regress FE_t^i on FE_t^{inf} and FE_t^{g}: FE_t^i = β_0 + β_1 FE_t^{inf} + β_2 FE_t^{g} + ε_t. Residual ε_t is the exogenous monetary policy shock (MP). - Estimating impact on inequality: - Local projection method (Jorda, 2005). For horizons h = 1,...,4 estimate: - y_{i,t+h} − y_{i,t} = α_{i}^h + θ_{t}^h + β^h MP_{i,t} + γ^h X_{i,t} + ε_{i,t+h}. - y_{i,t+h} − y_{i,t} is accumulated change in 100 times the log of net inequality. α_i^h are country fixed effects; θ_t^h are time fixed effects; X_{i,t} includes lagged monetary policy shocks and lagged changes in inequality. - β^h used to generate impulse response functions. - Role of financial inclusion: - Nonlinear specification with smooth transition function G(z_{i,t}) using normalized financial inclusion index z (zero mean, unit variance): - y_{i,t+h} − y_{i,t} = α_{i}^h + θ_{t}^h + β_{1}^h G(z_{i,t}) MP_{i,t} + β_{2}^h (1 − G(z_{i,t})) MP_{i,t} + γ^h X_{i,t} + ε_{i,t+h}, - G(z_{it}) = exp(−γ z_{i,t}) / (1 + exp(−γ z_{i,t})), γ > 0. - Similar exercise conducted for financial depth. ### Financial inclusion and output-price volatility - Volatility regression specification: - V_t = β_0 + β_1 FI_t + B X_t + ε_t, - V_t is output-price volatility defined as the log of the ratio of output gap to inflation variance; FI_t is the financial inclusion index; X_t includes controls such as level of financial development and interaction terms. - Reported results (Table 4.1 — Inflation Output Volatility): - FI Index24.70***34.86***18.48**39.13***24.73**24.36* (coefficients and significance exactly as reported in table). - Index * APD Dummy 219.29**216.66*20.93921.873 (coefficients as reported). - Real GDP per Capita2.44e-05***2.45e-05*** (coefficients as reported). - Credit (% of GDP)0.0133***0.0179***0.0125***0.0122*** (coefficients as reported). - Index * Credit 20.243**20.179**20.167 (coefficients as reported). - Constant 20.566***20.652***20.998***21.284***21.321***21.313*** (coefficients as reported). - Observations: 174, 174, 150, 150, 150, 150 across specifications. R-squared: 0.110, 0.132, 0.194, 0.222, 0.267, 0.267. Standard errors in parentheses. *** p,0.01, ** p,0.05, * p,0.1. ### Monetary regimes and financial inclusion in Asia (Table 4.2 highlights) - Table reports monetary regimes, main policy instruments, AFI membership, and financial inclusion index quintile grouping for Asia-Pacific economies. - Quintile groupings (examples preserved as in table): - Financial inclusion, index quintile 1: Australia (IT 2–3% Cash rate), Hong Kong SAR (No (Currency board) Maintain currency stability—), Japan (IT 2% Short and long term policy interest rates), Korea (IT 2% Base rate), New Zealand (IT 1–3% Official cash rate), Singapore (Implicit IT Primary objective to promote low inflation). - Financial inclusion, index quintile 2: Brunei Darussalam (No (Currency board)), India (IT 5% Benchmark repo rate), Indonesia (IT 4% 1/2 1 ppt BI policy rate), Sri Lanka (Other (CLA) Repo and reverse repo rates), Thailand (IT 2.5% 1/2 1.5 ppt BOT 1-day bilateral repo rate). - Financial inclusion, index quintile 3: Bangladesh (Monetary aggregate), China (Monetary aggregate Includes PBC benchmark rate, 7-day repo rate), Malaysia (Other (OM) BNM overnight policy interest rate), Philippines (IT 3% 1/2 1 ppt Overnight reverse repo rate). - Financial inclusion, index quintile 4 and 5: table lists countries including Bhutan, Cambodia, Mongolia, Nepal, Lao PDR, Myanmar, Papua New Guinea, Solomon Islands, etc., with their regime class and main instruments. - Sources for table: IMF, AREAERS/Article IV reports, central bank websites, Alliance for Financial Inclusion (AFI). Grouping note: Australia and NZ added to quintile 1 index group. For missing index values: Palau and FSM set equal to RMI; Kiribati set equal to Vanuatu; Bhutan set equal to Nepal. *Source: IMF Staff Estimates (Appendix 2 and Appendices 3–4 of the provided chapter).* ### Appendix 4: Financial Inclusion and Monetary Policy ### Appendix 4: Financial Inclusion and Monetary Policy ### Estimation approach and caveats - Method: OLS panel regression with fixed effects on an unbalanced panel of 190 countries covering 2004–2015. - Purpose: To estimate associations between financial inclusion and fiscal outcomes (quantitative: revenue and expenditure as a share of GDP; qualitative: revenue efficiency and tax efficiency). - Controls: trade openness, exchange rate regime, inflation, level of development (GDP per Capita), financial depth (Credit to Private Sector), capital account openness (Chinn-Ito), Gini coefficient, Rule of Law, interaction terms, and Asia dummy (APD). - Data limitations: Introduction of controls reduces sample size; sophisticated fiscal indicators have less coverage. - Identification: Results are correlations; endogeneity issues may remain, limiting causal claims. ### Key regression results — financial inclusion and fiscal outcomes (selected coefficients and significance) - ATMs per 100,000 People / Adults - Table 5.1a (ATMs per 100,000 People): - Revenue (% of GDP): 0.0748*** (standard error 0.0118) - Expenditure (% of GDP): 0.129*** (0.0112) - Tax Efficiency (VAT, %): 0.405*** (0.0388) - Observations: 1,140 (revenue), 1,139 (expenditure), 481 (tax efficiency) - R-squared: 0.291 (revenue), 0.182 (expenditure), 0.292 (tax efficiency) - Table 5.2a (ATMs per 100,000 Adults, robustness): - Revenue (% of GDP): 0.0355*** (0.0067) - Expenditure (% of GDP): 0.0342*** (0.0065) - Tax Efficiency (VAT, %): 0.0619*** (0.0134) - Observations: 830 (revenue), 829 (expenditure), 610 (tax efficiency) - R-squared: 0.730 (revenue), 0.767 (expenditure), 0.618 (tax efficiency) - Interaction with Asia Dummy (Table 5.1a): - ATMs per 100,000 People * Asia Dummy: −0.0126 (0.0300) for revenue; −0.0309 (0.0285) for expenditure; 0.309*** (0.1070) for tax efficiency. - Branches per 100,000 People / Adults - Table 5.1a (Branches per 100,000 People): - Revenue (% of GDP): 0.0781*** (0.0163) - Expenditure (% of GDP): 0.115*** (0.0158) - Tax Efficiency (VAT, %): 0.496*** (0.0627) - Observations: 1,278 (revenue), 1,275 (expenditure), 361 (tax efficiency) - R-squared: 0.304 (revenue), 0.145 (expenditure), 0.186 (tax efficiency) - Table 5.2b (Branches per 100,000 adults, robustness): - Revenue (% of GDP): 0.0797*** (0.0126) - Expenditure (% of GDP): 0.0776*** (0.0120) - Tax Efficiency (VAT, %): 0.165*** (0.0256) - Observations: 915 (revenue), 914 (expenditure), 666 (tax efficiency) - R-squared: 0.718 (revenue), 0.766 (expenditure), 0.603 (tax efficiency) - Interaction with Asia Dummy (Table 5.1a): - Branches per 100,000 People * Asia Dummy: 0.0638 (0.0527) for revenue; 0.0866* (0.0512) for expenditure; −0.593 (0.5710) for tax efficiency. - Depositors per 1,000 People / Adults - Table 5.1b (Depositors per 1,000 People): - Revenue (% of GDP): 0.00272*** (0.000485) - Expenditure (% of GDP): 0.00419*** (0.000478) - Tax Efficiency (VAT, %): 0.0162*** (0.001980) - Observations: 830 (revenue), 829 (expenditure), 354 (tax efficiency) - R-squared: 0.273 (revenue), 0.212 (expenditure), 0.247 (tax efficiency) - Table 5.2d (Depositors per 1,000 adults, robustness): - Revenue (% of GDP): 0.00204*** (0.0004) - Expenditure (% of GDP): 0.00228*** (0.0004) - Tax Efficiency (VAT, %): 0.00147* (0.0009) - Observations: 511 (revenue), 510 (expenditure), 388 (tax efficiency) - R-squared: 0.712 (revenue), 0.773 (expenditure), 0.545 (tax efficiency) - Interaction with Asia Dummy (Table 5.1b): - Depositors per 1,000 People * Asia Dummy: −0.0024* (0.00126) for revenue; −0.00208* (0.00124) for expenditure; 0.0201*** (0.00566) for tax efficiency. - Borrowers per 1,000 People / Adults - Table 5.1b (Borrowers per 1,000 People): - Revenue (% of GDP): 0.0106*** (0.00245) - Expenditure (% of GDP): 0.0188*** (0.00223) - Tax Efficiency (VAT, %): 0.0705*** (0.00749) - Observations: 735 (revenue), 735 (expenditure), 292 (tax efficiency) - R-squared: 0.312 (revenue), 0.181 (expenditure), 0.327 (tax efficiency) - Table 5.2c (Borrowers per 1,000 adults, robustness): - Revenue (% of GDP): 0.00368* (0.0019) - Expenditure (% of GDP): 0.00495*** (0.0018) - Tax Efficiency (VAT, %): 0.00566 (0.0059) - Observations: 373 (revenue), 373 (expenditure), 248 (tax efficiency) - R-squared: 0.687 (revenue), 0.741 (expenditure), 0.629 (tax efficiency) - Interaction with Asia Dummy (Table 5.1b): - Borrowers per 1,000 People * Asia Dummy: 0.00608 (0.00661) for revenue; 0.00249 (0.00601) for expenditure; 0.210*** (0.04140) for tax efficiency. - Asia Dummy (baseline differences) - Across many specifications Asia Dummy coefficients are large and statistically significant, e.g. Table 5.1a shows: - Asia Dummy: 26.283*** (1.087) for revenue in ATM specification; 25.324*** (1.033) for expenditure. - Indicates that APD group differs substantially from rest of the world in levels of fiscal outcomes in these models. ### Country fixed effects specifications (selected) - Table 5.3a (ATMs per 100,000 Adults; country fixed effects) - ATMs per 100,000 Adults: - Revenue (% of GDP): 0.0207* (0.0111) - Expenditure (% of GDP): 0.0272** (0.0133) - Tax Efficiency (VAT, %): 0.0652* (0.0340) - GDP per Capita (USD): −0.000624*** (0.000151) for revenue. - Observations: 1,432 (revenue); Number of Countries: 158; R-squared: 0.276 (revenue). - Table 5.3c (Borrowers per 1,000 Adults; country fixed effects) - Borrowers per 1,000 Adults: - Revenue (% of GDP): 0.00668** (0.0028) - Expenditure (% of GDP): 0.00944*** (0.0031) - Observations: 690 (revenue); Number of Countries: 82; R-squared: 0.278 (revenue). ### Fiscal multiplier estimation: motivation and method - Hypothesis: Financial inclusion could raise fiscal multipliers via: - Improved targeting of transfers/subsidies and reduced leakage (less crowding-out; more spending in formal economy). - Broader access to financing strengthening crowding-in effects for private investment. - Broader account coverage increasing bank savings from transfers, boosting funds for intermediation. - Empirical approach: OLS panel regression with interaction terms following Espinoza and Senhadji (2011). - Dependent variable: Real GDP growth. - Fiscal variable: Growth of real expenditure (current and lags). - Financial inclusion measures: ATMs per 100,000 adults, Branches per 100,000 adults, Borrowers per 1,000 adults, Depositors per 1,000 adults. - Cumulative short-term multiplier: partial derivative formulas shown in the source: - First period: ∂ RealGDPgrowtht,i / ∂ RealExpendituregrowtht,i = β1 * FinancialInclusiont,i + β3 - Second period adds β4 * FinancialInclusiont−1,i + β5 ### Fiscal multiplier results (Tables 5.4a and 5.4b) — selected coefficients and interactions - Baseline coefficient on Growth of Real Expenditure: - Table 5.4a (four specifications): 0.102*** (0.0135) in specification (1); 0.00936** (0.0036) in (2); 0.00463 (0.0034) in (3); 0.0877*** (0.0151) in (4). - ATMs per 100,000 adults (Table 5.4a, spec. with ATMs) - ATMs per 100,000 adults: −0.0454*** (0.005450) - Growth of Real Expenditure * ATMs per 100,000 adults: 0.00137*** (0.000500) - Growth of Real Expenditure (t-1) * ATMs per 100,000 adults: 0.00148*** (0.000436) - Growth of Real Expenditure (t-2) * ATMs per 100,000 adults: 0.000302 (0.000443) - Observations: 900; R-squared: 0.224 - Branches per 100,000 adults (Table 5.4a, spec. with Branches) - Branches per 100,000 adults: −0.0552*** (0.008710) - Growth of Real Expenditure * Branches per 100,000 adults: 0.005040*** (0.000664) - Growth of Real Expenditure (t-1) * Branches per 100,000 adults: 0.00238*** (0.000738) - Growth of Real Expenditure (t-2) * Branches per 100,000 adults: 0.000746 (0.000687) - Observations: 1,046; R-squared: 0.118 - Borrowers per 1,000 adults (Table 5.4a, spec. with Borrowers) - Borrowers per 1,000 adults: −0.00716*** (0.0011) - Growth of Real Expenditure * Borrowers per 1,000 adults: 0.000518*** (0.000103) - Growth of Real Expenditure (t-1) * Borrowers per 1,000 adults: 0.000323*** (0.000100) - Growth of Real Expenditure (t-2) * Borrowers per 1,000 adults: 0.000087 (0.000098) - Observations: 556; R-squared: 0.110 - Depositors per 1,000 adults (Table 5.4a, spec. with Depositors) - Depositors per 1,000 adults: −0.00108*** (0.0003) - Growth of Real Expenditure * Depositors per 1,000 adults: 0.0000635** (0.000021) - Growth of Real Expenditure (t-1) * Depositors per 1,000 adults: 0.0000297 (0.000019) - Growth of Real Expenditure (t-2) * Depositors per 1,000 adults: −0.0000215 (0.000019) - Observations: 546; R-squared: 0.180 - Robustness checks (Table 5.4b) generally confirm: - Growth of Real Expenditure baseline coefficients: 0.119*** (0.0168) in spec. (1); 0.00685* (0.0037) in (2); 0.00366 (0.0033) in (3); 0.116*** (0.0182) in (4). - ATMs and Branches retain significant interaction terms in some robustness specs: - Growth of Real Expenditure * Branches per 100,000 adults: 0.00508*** (0.000814) - Growth of Real Expenditure * ATMs per 100,000 adults: 0.000709 (0.000569) (not significant in this robustness spec), Growth of Real Expenditure (t-1) * ATMs per 100,000 adults: 0.000795* (0.000477) ### Interpretations from results reported - Financial inclusion indicators (ATMs, branches, depositors, borrowers) are generally positively associated with higher revenue and expenditure shares of GDP and with tax efficiency measures in multiple specifications. - Interaction terms imply financial inclusion amplifies the short-term fiscal multiplier associated with real expenditure growth: - Positive and statistically significant coefficients on Growth of Real Expenditure * FinancialInclusion in many specifications (notably Branches and Borrowers; also ATMs and Depositors in several specifications). - The direct coefficients on financial inclusion variables in the multiplier regressions are negative (e.g., ATMs per 100,000 adults: −0.0454***; Branches per 100,000 adults: −0.0552***; Borrowers per 1,000 adults: −0.00716***; Depositors per 1,000 adults: −0.00108***), while the interaction terms with expenditure growth are positive, implying financial inclusion modifies the response of GDP growth to expenditure growth. ### Limitations and caution - Results are associative, not causal: financial inclusion tends to change gradually and reforms coincide with other developments, making causal identification difficult. - Sample coverage varies substantially across indicators and more sophisticated fiscal measures have less data. - Standard errors and significance levels: standard errors reported in parentheses; significance denoted by *** p,0.01, ** p,0.05, * p,0.1. *Source: Appendix 4: Financial Inclusion and Monetary Policy, from the provided IMF content unit.* ### Appendix 5. Financial Inclusion and Fiscal Policy ### Appendix 5. Financial Inclusion and Fiscal Policy ### Model specification and data - Empirical approach: simple OLS regressions on end-2015 cross-sectional data for countries in the IMF’s Financial Access Survey (FAS) database using the specification: - Yi = 휶 + 휷1 Xi + 휷2 Ci + ui - Yi: measure of financial inclusion - Xi: set of explanatory variables - Ci: set of control variables - Sample notes: - Dataset contains about 190 economies; sample varies by indicator availability. - Financial inclusion data run from 2004 until 2015, with many series starting in 2008. - No controls for cross-sectional dependence because of short time-series (about 5–6 years in many cases). - Robustness check: country fixed effects specification employed where feasible; caution advised due to limited time series. Second specification not possible for depositors per 1,000 adults and borrowers per 1,000 adults due to lack of data. ### Dependent variables (Yi) used to measure financial inclusion - (i) number of ATMs per 1,000 inhabitants - (ii) number of commercial bank branches per 1,000 km2 - (iii) number of bank loans per 1,000 inhabitants - (iv) number of deposit accounts per 1,000 inhabitants ### Explanatory variables (Xi) - (i) cellular access: per capita mobile phone subscriptions - (ii) internet access: number of secure internet servers per capita - (iii) technological leapfrogging: ratio of mobile to fixed line phone subscriptions - Interaction terms included between the Asia-Pacific regional dummy and cellular access, internet access, and technological leapfrogging to test for region-specific effects. - Regional dummies include Africa and Asia-Pacific to highlight comparative performance. ### Control variables (Ci) - (i) GNI per capita (level of income) - (ii) regional dummies - (iii) financial depth proxied by credit-to-GDP ratio ### Interpretation guidance - Positive, significant coefficients: technology complements inclusion in the traditional financial system. - Negative, significant coefficients: technology substitutes for inclusion in the traditional financial system. - Insignificant coefficients: phenomena are independent. ### Key regression results (high-level findings from Table 6.1) - Leapfrogging: - ATM (per 100,000 adults): 20.03*** (0.01) - Bank Branches (per 1,000 km2): 20.01** (0.01) - Bank Loans (per 1,000 adults): 20.12** (0.05) - Deposit Accounts (per 1,000 adults): 20.64*** (0.19) - GNI per capita: - ATM: 0.00** (0.00) - Bank Branches: 20.00** (0.00) - Bank Loans: 0.00*** (0.00) - Deposit Accounts: 0.00 (0.00) - Asia-Pacific dummy (APD): - ATM: 23.45 (2.73) - Bank Branches: 3.83 (3.04) - Bank Loans: 22.09 (31.78) - Deposit Accounts: 2108.75 (86.11) - Africa dummy (AFR): - ATM: 22.03 (1.91) - Bank Branches: 27.28*** (1.79) - Bank Loans: 280.27*** (29.56) - Deposit Accounts: 2353.31*** (59.27) - Leapfrogging * Asia Dummy: - ATM: 20.06*** (0.02) - Bank Branches: 0.03 (0.04) - Bank Loans: 20.96*** (0.30) - Deposit Accounts: 21.33 (1.01) - Cellular Access: - ATM: 0.15*** (0.03) - Bank Branches: 20.09*** (0.03) - Bank Loans: 0.50 (0.34) - Deposit Accounts: 3.92*** (1.10) - Cellular * Asia Dummy: - ATM: 0.08 (0.06) - Bank Branches: 0.15*** (0.06) - Bank Loans: 20.30 (0.44) - Deposit Accounts: 21.95 (1.69) - Cellular * Africa Dummy: - ATM: 20.04 (0.03) - Bank Branches: 0.06* (0.03) - Bank Loans: 0.63 (0.45) - Deposit Accounts: 22.88** (1.18) - Internet Access: - ATM: 0.62*** (0.06) - Bank Branches: 0.45*** (0.07) - Bank Loans: 7.12*** (0.74) - Deposit Accounts: 8.24*** (2.32) - Internet * Asia Dummy: - ATM: 20.11 (0.12) - Bank Branches: 20.56*** (0.10) - Bank Loans: 1.35 (1.13) - Deposit Accounts: 9.71** (4.42) - Internet * Africa Dummy: - ATM: 0.08 (0.10) - Bank Branches: 0.49** (0.23) - Bank Loans: 22.19* (1.26) - Deposit Accounts: 24.38*** (4.01) - Constant terms and model fit: - ATM constant: 1.79 (1.81); Observations: 1,137; R-squared: 0.58 - Bank Branches constant: 7.81*** (1.57); Observations: 1,250; R-squared: 0.19 - Bank Loans constant: 41.22 (29.07); Observations: 653; R-squared: 0.54 - Deposit Accounts constant: 390.56*** (56.96); Observations: 807; R-squared: 0.44 - Robust standard errors in parentheses. Significance levels: *** p,0.01, ** p,0.05, * p,0.1. ### Technological leapfrogging and financial inclusion (Tables 6.2 and 6.3 — selected coefficients and diagnostics) - Table 6.2 (All countries; Number of countries and observations shown): - Technological Leapfrogging: - Branches per 100,000 adults: 0.00 (0.0065) - ATMs per 100,000 adults: 20.03*** (0.3900) - Depositors per 1,000 adults: 0.31 (0.0200) - Borrowers per 1,000 adults: 20.02 (0.0000) - Internet Access: - Branches: 0.07*** (0.040000) - ATMs: 0.00067*** (0.000035) - Depositors: 9.93*** (2.510000) - Borrowers: 2.07** (0.810000) - Cellular Access: - Branches: 0.03*** (0.010000) - ATMs: 20.000726*** (0.000231) - Depositors: 0.89 (0.810000) - Borrowers: 0.47* (0.240000) - GNI per Capita and Financial Development reported with exact coefficients and standard errors as in source. - Observations: 830, 829, 739, 634 respectively. R-squared: 0.73, 0.767, 0.46, 0.32. Number of Countries: 133, 127, 93, 77. - Table 6.3 (Pacific Islands subset; interaction terms with Pacific reported): - Technological Leapfrogging * Pacific: - Branches per 100,000 adults: 0.03*** (0.010) - ATMs per 100,000 adults: 0.69*** (0.390) - Internet Access * Pacific: - Branches per 100,000 adults: 0.03*** (0.010) - ATMs per 100,000 adults: 0.61*** (0.140) - Cellular Access * Pacific: - Branches per 100,000 adults: 20.02 (0.040) - ATMs per 100,000 adults: 0.03 (0.140) - Observations and R-squared identical to Table 6.2 for comparability: Observations 830, 829, 739, 634; R-squared 0.73, 0.767, 0.46, 0.32. Number of Countries: 133, 127, 93, 77. - Standard errors in parentheses. Significance levels: *** p,0.01, ** p,0.05, * p,0.1. ### Case study highlights — Bangladesh (Appendix 7 excerpt relevant to financial inclusion and fiscal policy) - Historical context: - Financial inclusion in Bangladesh has long relied on microcredit; Grameen’s influence noted. - Bangladesh Bank (BB) created a Financial Inclusion department and included financial inclusion goals in its Strategic Plan (2015–19). - Microfinance scale: - As of June 2016: 680 licensed MFIs (an additional 191 with provisional licenses) catering to over 26 million clients. - Outstanding disbursements by the top five MFIs in FY16 compare to a substantial 10 percent of private sector credit extended by the banking system. - No-frill accounts and targeted programs: - No-frill Taka 10 accounts introduced in 2010 targeting farmers to enable receipt of government transfers; accounts have no fees or minimum balance requirements. - No-frill accounts extended to garment workers, small footwear and leather industry workers, city corporation cleaning workers, the physically challenged, and school children; in 2014 banks instructed to provide similar services to impoverished children under NGO guidance. - Growth in number of accounts and balances indicated in source figures (time series shown but exact series preserved in source). - Priority-sector credit and refinance facilities: - Eight refinance facilities for banks and NBFIs to provide personally guaranteed uncollateralized loans to women entrepreneurs, new entrepreneurs and cottage and agro-based industries at preferential rates. - Banks encouraged to offer SMEs a minimum of 20 percent of their lending, and women entrepreneurs at least 15 percent of total SME funding. - Disbursements to women entrepreneurs rose from a little over 0.2 percent of GDP in 2010 to close to 0.3 percent of GDP in 2016. - Agricultural and rural credit: - BB’s Agricultural and Rural Credit Policy sets disbursement targets with interest rate ceilings (currently 10 percent). - Interest on certain crops and certain areas is further subsidized by the government. - Since July 2015, BB stopped compensating banks on blocked undisbursed amounts held in BB’s accounts. - Digital and agent channels: - Mobile financial services introduced in 2011; agent based banking in 2013. - Five mobile operators and eighteen banks provide mobile financial services; twelve banks offer agent based banking. - Agents providing mobile banking increased from 51,000 in 2012 to 772,109 in 2017. - Active mobile money accounts increased from 0.5 million to 28 million between 2012 and 2017. - Agent based banking agents grew from 1,646 in December 2016 to 1,847 in June 2017. - More than 9,000 ATMs and close to 36,300 point of sales machines exist around the country. - Number of rural bank branches now exceeds urban branches; rural deposits shown to be growing (time series preserved in source figures). - Risks and regulatory responses: - Non-performing loans (NPLs) decreased in agriculture sector from three years ago but increased sharply in the SME sector. - Authorities plan to revisit set targets and implement better Know Your Customer (KYC) policies: - Mandatory national identity card for accessing credit. - Dedicated window/desk for women and SME entrepreneurs in creditor institutions to assist with KYC. - Strategic planning and future priorities: - National Financial Inclusion Strategy being drafted jointly by Ministry of Finance, BB and Department for International Development (DFID, U.K.). - Strategy will define products, extent, time frame, prioritization, and coordination for implementation. - Priorities include: credit growth in priority sectors, SME financing, gender priority, rural access, green financing, life-cycle approach to inclusion, insurance, and building a digital ecosystem to connect the first and last mile. - Recommendations to strengthen use of formal finance: strengthen legal and regulatory framework, consider a financial ombudsman, pilot innovations (example referenced from Papua New Guinea), conduct periodic surveys, develop a financial inclusion index to track progress, and implement the finalized Financial Inclusion strategy as planned. - Gross nonperforming loans ratio (Table 7.1) — In percent of total loans outstanding in that sector: - Agriculture: 2013: 22.8; 2014: 23.4; 2015: 14.3; 2016: 17.9 - SME: 2013: 9.9; 2014: 9.2; 2015: 18.7; 2016: 13.2 - Other sectors: 2013: 7.8; 2014: 8.9; 2015: 8.3; 2016: 8.6 ### Cross-cutting implications from empirical and country evidence - Technology effects: - Internet access and cellular access show several positive, significant associations with financial inclusion indicators (ATMs, branches, loans, deposit accounts) in cross-sectional regressions, but regional interactions matter (Asia-Pacific and Africa coefficients and interactions often significant). - Technological leapfrogging has mixed associations depending on the indicator and region; some negative coefficients on ATMs and loans are significant. - Policy trade-offs: - Rapid expansion of access (microcredit, mobile accounts, agent networks) increases inclusion but presents financial stability risks when combined with rapid credit growth, particularly in MFI and SME sectors. - Complementary macroprudential and regulatory measures (KYC, monitoring NPLs, revisiting lending targets, interest rate ceilings, subsidization policies) are needed to balance inclusion with stability. *Prepared from Appendix 5 and Appendix 7 excerpts of the source PDF.* ### appendix 8. Cambodia: Balancing ### Appendix 8. Cambodia: Balancing Financial inclusion and Stability ### Context and dollarization - Cambodia’s financial system is highly dollarized: "95 percent of deposits are denominated in U.S. dollars." - Average loan-to-deposit (LTD) ratio of commercial banks reached "100 percent by end-2016", up from "79 percent in 2010." - The financial system is dominated by banks, which account for "90 percent of total assets." ### Financial inclusion developments - Rapid growth of MFIs has expanded formal financial services, especially in rural areas. - MFI assets-to-GDP ratio expanded from "4 percent to 20 percent between 2011 and 2016." - As of 2016, MFIs provided loans to over "2 million borrowers (20 percent of adult population)", more than three times as many clients as commercial banks served. - MFIs credit stock and flows account for about "20 percent" of those in the financial system. - As of 2016, there are "seven MDIs out of 71 MFIs." ### Mobile financial services - Mobile money accounts have increased rapidly; in terms of number of mobile accounts as percentage of population, Cambodia is the highest among regional peers. - Growth is supported by demand from domestic migrant workers for low cost money transfer services and rising mobile phone penetration. ### Remaining gaps and specific inclusion metrics - Share of adult population holding accounts (including mobile accounts) is lower than suggested by Cambodia’s level of income and financial depth (based on World Bank Global Findex, 2014). - Cambodia has one of the highest rates of adults borrowing from a financial institution: "around 28 percent." - Cambodia has one of the lowest rates of adults saving at a financial institution: "only 4 percent." - Finscope 2015 (based on its definition) reports "59 percent of adults formally included", higher than Laos and Myanmar. - Use of financial services is skewed: transactions, credit, savings, insurance usage differ regionally (chart notes). ### Financial stability risks from MFI expansion - Liquidity risks: - MDIs have average LTD ratios "over 200 percent." - MFIs rely largely on external sources of funding and commercial borrowing from overseas, creating rollover and interest-rate risks if global financial conditions tighten. - Foreign exchange risks: - Around "90 percent of loans denominated in U.S. dollars." - Changes in the riel/U.S. dollar exchange rate can impact unhedged borrowers’ debt servicing capacity, particularly for rural borrowers earning mostly in riel. - Credit risks: - Rapid expansion of MFI lending—especially in rural areas—to maintain growth targets in a highly-saturated industry has raised credit risks. - The average MFI loan size increased from "around US $800 in 2013 to almost US $1600 in 2016." - NPL ratio and average loan size increases suggest an upward trend in credit risks. - Interconnectedness: - MFIs are predominantly funded through foreign borrowing but also borrow from domestic banks, increasing interconnectedness risks. - Regulatory asymmetry: - MFIs are subject to less stringent regulation than banks: looser capital and reserve requirements, exemptions for provisions on restructured loans, and looser asset classifications. ### Government strategy and policy intentions - Financial inclusion is a strategic priority in the Financial Sector Development Strategy 2016−2025. - Government plans: - Develop a National Financial Inclusion Strategy led by the National Bank of Cambodia (NBC). - Draft laws and regulations to establish a credit guarantee scheme to facilitate increased access to credit and further financial inclusion. - MFIs’ funding largely comes from parent companies and international or bilateral development financial institutions, which the authorities consider to be a stable form of funding. ### Prudential and supervisory recommendations - Strengthen regulation and supervision to eliminate regulatory arbitrage between MFIs and banks: - Standardize asset classification for all banks and MFIs. - Unify accounting and provisioning standards between MFIs and banks. - Adjust reserve requirements for MDIs to be in line with banks. - Improve monitoring of systemic linkages. - Align solvency ratio requirements for MFIs and commercial banks more closely with Basel II and Basel III principles. - Apply different prudential treatment to riel and U.S. dollar intermediation consistent with different risk profiles. - Monitor unhedged exchange rate risk exposure in banks and MFIs’ clients and ensure prudent limits via maximum debt-servicing to income ratio. - Shift supervisory approach from compliance-based to more risk-based supervision. ### Measures to address dollarization and promote riel use - NBC measures to promote riel use: - Liquidity-Providing Collateralized Operations established in October 2016 to provide lower-cost riel liquidity. - Further prudential measures to factor in dollarization risks can improve local currency use and reduce vulnerabilities. ### Capacity and staffing constraints - Supervisory capacity is constrained: largely unchanged number of NBC staff over time impedes adequate and effective supervision of the fast-growing financial sector. - Alleviating strains on supervisory capacity and resources needs to be prioritized. *Appendix 8. Cambodia: Balancing Financial inclusion and Stability — IMF staff report content as provided.* ### Appendix 9. Financial Inclusion in India ### Appendix 9. Financial Inclusion in India ### Recent financial inclusion policy steps: key aspects - As recently as 2011, a Census survey revealed that only about 59 percent of households had bank accounts, with low access for rural and low-income populations, and women. - The launch of the Pradhan Mantri Jan Dhan Yojana (PMJDY) in August 2014—backed by the rapid spread of Aadhaar—has led to the addition of more than 280 million bank accounts by previously unbanked individuals. - Close to half of new PMJDY accounts were opened by females and 60 percent by persons in rural areas. - Penetration of bank deposit accounts in India was more than 1.5 per adult as of end-2015. - Caveats on account penetration: - Certain individuals have multiple bank accounts. - Average balances are small (at less than Rs. 2,500). - Transactional volumes are still limited. - Infrastructure constraints affecting use: - Relatively low density of bank branches and ATMs. - Limited mobile money transactions across India. - PMJDY implementation notes: - PMJDY accounts are largely channeled through public banks, which account for 80 percent of enrolment. - Transactions are facilitated by fixed-point agents (Bank Mitra) that act on behalf of banks. - Linkage to social transfers and incentives: - Delivery of subsidies and social spending through PMJDY accounts has supported more targeted and efficient government social transfers and can yield additional fiscal savings. - Add-on services and incentives include debit cards, digital direct benefit transfers (direct cash payments for social programs, certain food and fertilizer subsidies, and health insurance), bundling with free life and accident insurance and pension products, and overdraft facilities. - Digital and payments infrastructure: - The Government’s Digital India initiative (adopted in 2015) is expected to give India’s digital economy a considerable push by 2019. - Unified Payments Interface (UPI) platform—an instant, smartphone-based electronic payment system introduced in 2016—permits digital linkage of multiple accounts to banking services, including person-to-person transfers and merchant payments, improving speed and traceability. - Credit expansion initiatives: - MUDRA Bank and the Pradhan Mantri MUDRA Yojana (PMMY) scheme were set up in 2015 to bridge funding gaps for micro-, small- and medium-sized enterprises (MSME). - MUDRA Bank provides refinancing to micro-finance and nonbank financial institutions which on-lend to enterprises at different growth stages, including to enterprises led by women, scheduled castes (SCs), scheduled tribes (STs) or other backward classes (OBC). - PMMY specifically targets micro enterprises. - Stand-Up India scheme (launched in 2016) targets provision of bank loans specifically to women entrepreneurs, SCs and STs. - Per the Mor Committee estimate (2014), 90 percent of India’s small firms are estimated not to have access to bank services. - Women-targeted initiatives: - Self-Help Groups (SHGs) mostly comprise women and enhance women’s access to credit and income-generating capacity. - SHG loans to women are set at a preferential rate of 7 percent with banks compensated through interest subventions. - There are cases when such loans are interest-free, e.g., funding for 78,000 women-run SHGs in Maharashtra covered by a 100 percent subvention from the state government. - Under PMMY, credit to women-led businesses accounted for about half of the total amount lent and about 80 percent of the number of loans. - Financial literacy efforts: - RBI’s Project Financial Literacy promotes banks’ financial literacy centers and financial literacy camps. - SEBI cooperates with Sanchayan and set up a panel of qualified Resource Persons to disseminate financial literacy. ### Access and use — key statistics (end-2015) - Automated Teller Machines (ATMs) per 100,000 adults: 2024 - Branches of commercial banks per 100,000 adults: 1446 - Deposit accounts with commercial banks per 1,000 adults: 1,542 - Outstanding deposits (in percent of GDP): 66 - Loan accounts with commercial banks per 1,000 adults: 154 - Mobile money transactions: number per 1,000 adults: 193 - Mobile money transactions: value, in percent of GDP: 1.0 - Source for these figures: IMF, Financial Access Survey (as presented in the chapter). ### Economic and financial stability impact — empirical findings - A two-sector general equilibrium model for India (formal and informal finance) shows policies enhancing women’s access to formal finance improve gender equality in entrepreneurship. - Overall increase in entrepreneurship raises investment, labor force participation and employment, strengthening GDP growth. - The considerable positive effect on output is mostly linked to relaxation of collateral constraints. - The reduction in income inequality is primarily related to the associated decline in funding costs. - Policy-Specific Lending (PSL) experience: - Counterfactual policy analysis indicates PSL has boosted financial inclusion and enterprises’ access to credit but has led to a slight uptick in nonperforming loans (NPLs). - As firms gained first-time access to funding or raised borrowing, investment, aggregate output and productivity increased, and inequality outcomes were favorable; however, rising NPLs warrant close monitoring of borrower credit quality and balancing of inclusion and macroprudential measures. ### Risks and the road ahead — constraints and recommendations - Remaining challenges: - Account dormancy has declined, but use of services and access to credit still need to be boosted. - Transactional volumes and average account balances remain small despite improvements. - Expansion of points-of-access (ATM availability), digitalization, mobile banking and greater usage of direct benefit transfer schemes are needed to stimulate account use and deliver efficiency gains. - Data security must be emphasized to support customer confidence in mobile banking and UPI solutions. - Further geographic expansion of micro-finance schemes should be carried out under tight regulatory oversight to avert overly aggressive growth and risk buildup. - Policy priorities and recommendations: - Encourage higher transactional volumes on existing accounts through digitalization and incentives. - Channel more subsidies and social payments through Aadhaar-linked bank accounts to realize fiscal savings and increase usage. - Enhance data security and consumer protection to build trust in digital payment platforms. - Strengthen regulatory oversight of micro-finance expansion to prevent risk buildup. - Continue and expand financial literacy programs, including workshops aimed at building rural communities’ trust with the banking system, and conduct surveys to understand costs and barriers to usage of formal financial services. - Facilitate healthy demand for formal finance by addressing structural constraints to female labor market participation; empirical evidence suggests relaxing funding constraints for females is considerably more effective when accompanied by policies to alleviate labor market rigidities. - Measure success of policy interventions by: - Pace of improvement in female labor market participation. - Extent of rise in formal female entrepreneurship. - Mobility of female-led firms from small to medium or large sizes. - Rationalize schemes for agricultural credit: - Improve coordination between center and state governments to reduce implementation inefficiencies. - Address high share of overdue loans in the agricultural sector. - Prioritize addressing long-term structural bottlenecks and improving agricultural productivity to ensure scheme success. - Industry outlook on payments growth: - Industry estimates suggest India’s mobile transactions are expected to grow very rapidly (12-fold in 2015−2020). *Appendix 9. Financial Inclusion in India — IMF staff chapter content as provided in the source PDF.* ### 1. Myanmar: Mobile Transformation ### 1. Myanmar: Mobile Transformation ### Mobile adoption and device landscape - Rapid decline in SIM card price and sharp rise in mobile penetration between 2011 and 2017 (chart noted). - Household ICTs and mobile phone ownership: figures presented for households and for 15–65 population (chart noted). - Mobile handset type distribution among mobile owners indicated (chart noted), with one chart element showing: 76% / 2% / 22% (as presented in the source). ### Regulatory framework and market developments - Central Bank of Myanmar (CBM) has established a basic regulatory framework for mobile financial services: - 2013 Directive on Mobile Banking (bank-based model): allows an agent to partner with a bank to sign up customers for cash transactions (albeit in a “closed-loop network”). - 2016 Regulation on Mobile Financial Services (non-bank-based model): provides payments and financial services using mobile technology infrastructure, including kyat-denominated cash-in and cash-out transactions, money transfers and domestic payments. The regulation permits mobile network operators (MNOs) to offer those services through their own platforms. - The 2016 regulation draws on international good practice and provides a number of improvements over the 2013 directive, including: - i) interoperability; - ii) agents network (without exclusivity); - iii) know-your-customer (KYC) and customer due diligence (CDD); - iv) customer protection. - Market entry and product rollout: - In 2016, the CBM issued the first license to Wave Money, a tie-up between Yoma Bank and Telenor (Norwegian-owned network operator). Wave Money initiated person-to-person transfers (P2P) and airtime top-up for mobile phones, and plans to expand its services to bill payments. - Microfinance institutions could use the bill payments system to collect their loans. - Other telecom operators, Ooredoo and Myanmar Post & Telecom (MPT), lined up for a license, indicating mobile money would be part of their strategy. - MPT also aims for government salary payments and government-to-persons (G2P) transfers, given its history as a state economic enterprise (SEE). ### Opportunities for leapfrogging and innovative options - Mobile ecosystems enable new designs for financial services, including: - i) biometric passwords; - ii) camera-integrated payment systems (e.g., QR Code scan & pay); - iii) social networking-integrated payment system (e.g., Viber); - iv) social network-integrated virtual savings groups (Pwint and others, 2017). - These innovative options could expand financial services to every segment of societies if appropriately designed and introduced. ### Remaining challenges and areas needing further effort - Success of mobile financial services depends on: - i) a continued expansion of mobile phone network and usage; - ii) an adequate branch network run to reach the disadvantaged areas; - iii) an appropriate regulatory framework to ensure system stability and growth. - Additional areas requiring attention: - Financial literacy and departure from cash-based economy: - Many Myanmar consumers have limited understanding of the benefits of mobile money. Given the long history of a cash-based economy, it will take time and effort to change the mindset. Awareness campaigns such as those envisaged under the Mobile Financial Service Providers (MFSPs) would help enlighten customers about the benefits of mobile money. - Informal sector: - Myanmar’s informal cash transfer system (e.g., Hundi system) has a long history and is prevalent and competitive. Mobile network operators may collaborate with the Hundi dealers by nominating them as their agents, though they must be first brought into the formal system. - Supervision: - CBM’s supervision team has limited capacity to measure current metrics on the amount of money in the system, transactions per month, number of active users and other basic data. This limits the quantifiable data available from the banks and limits the transparency of the bank-led systems. - Financial access: - It is unlikely that bank-led options are going to have any incentive to create interoperability or access for the poor and rural people. Microfinance service (MFS) providers led by MNOs have access to wide distribution networks and thus are perhaps better positioned to serve the rural population. *Appendix 10. Myanmar: Digital Leapfrog and Mobile Finance.* ### appendix 12. Case Study on ### Appendix 12. Case Study on Agricultural Insurance ### Fiscal exposure and subsidy arrangements - 55 percent of these policies are provided by the state-owned People’s Insurance Company of China. - India’s ’Prime Minister’s Crop Insurance’ Program contains provisions for public subsidies, with the amount capped at 50 percent of the premium value. - The country budgeted $1.4 billion to fund the program in 2017–18. - Agricultural insurance expansion typically requires substantial fiscal support. ### Actuarial performance and sustainability - Producer loss ratios, defined as total claim divided by total premium collected from farmers, are typically closer to or larger than 1. - Sample period: United States (1999–2015), India (1999–2009), Korea (2001–2009), all other countries (2003–2007). - Example: in the United States, the average loss ratio of crop insurance from 1995 to 20015 was about 2.2, i.e. insurance companies paid $ 2.2 claim for each dollar of premium revenue collected from farmers. - Considering operating costs of insurance policies usually amount to over 25 percent of the premium, a larger-than-75 percent loss ratio implies that government subsidies are required to prevent insurers from exiting the market. - One notable exception is Australia where governments do not subsidize insurers and insurers are still able to make positive profits; however, the size of crop insurance industry is substantially smaller than Europe and the United States. ### Marginal subsidy dynamics and demand elasticity - The marginal subsidy required for further expanding agricultural programs tends to be increasing. - Farmers’ demand for insurance product tend to be inelastic with respect to premium (Coble and Barnett 2013). - To encourage more farmers to participate, governments have to provide exponentially more financial incentives. - In both the United States and China, the governments increased subsidy rates to achieve larger coverages. ### Structural reasons why crop insurance relies on fiscal subsidy - Undiversified Risk: yield risks can be highly correlated across farms; aggregate risk is hard to eliminate through cross-sectional risk pooling (OECD 2009). - Information cost: actuarial analysis is mainly based on cross-time instead of cross-sectional data; limited sample size increases information cost, especially in Asian countries with limited historical weather and yield data. - Adverse selection: farmers with larger exposure to yield risks are more likely to purchase insurance policies (Glauber 2004); the problem is intensified by large information cost. - Moral hazard: farmers may change behavior (e.g. use of fertilizer, seeds, efforts, choice of plants) after the purchase of insurance policies (Glauber 2004, Young and others 2001). - Basis risks: index-based policies (e.g., regional rainfall) can suffer from low correlation between individual yields and the exogenous index (Clarke and others 2012). ### Alternatives: self-insurance and saving-account schemes - Schemes encourage farmers to accumulate savings to cope with yield risks (self-insurance). - Canada: farmers can set up individual special saving accounts for future yield loss; government provides matching contributions up to a limit; withdrawals only when yields or income fall below a threshold (Coble, 1995). - Australia: farmers can claim a tax deduction for income deposited to saving accounts under the farm management deposit (FMD) scheme; income taxes for withdrawals are due when farmers make withdrawals (typically in years with natural disasters), but overall farmers still pay less taxes due to the progressive income tax scheme. - In 2016, the total size of FMD savings reached AUD 40 billion. - Considering the variations of annual agriculture output in Australia is below AUD 4 billion, the savings under FMD should be able to largely insure against the yield risks. - Advantages of self-insurance schemes: - Less prone to moral hazard and adverse selection. - Lower management costs than crop insurance programs (AAFC, 2012). - Subsidizing saving accounts tends to be less regressive than subsidizing crop insurance. - Implementation challenges: - Require some level of financial development (e.g., bank accounts or tax identification), which can be challenging for developing countries in Asia. ### Policy implications and recommendations - Agricultural insurance can be a useful financial inclusion instrument to help farmers mitigate agricultural risks, but it can be costly. - Policymakers should consider alternatives (e.g., schemes based on self-insurance) that may be more cost-effective. - Policymakers seeking to promote financial inclusion should be mindful of the potential fiscal costs and distributional effects (e.g., under crop insurance subsidy schemes, rich farmers tend to get more subsidy because they have more acres). *Appendix 12. Case Study on Agricultural Insurance, Financial Inclusion in Asia‑Pacific* --- _Source: https://www.imf.org/-/media/files/publications/dp/2018/46115-dp1817-financial-inclusion-in-asia-pacific.pdf_