Asia Small and Medium-Sized Enterprise Monitor 2021 Volume III

Digitalizing Microfinance in Bangladesh: Findings from the Baseline Survey

• Éditeur: Asian Development Bank
• Année: 2022
• ISBN:
  9789292694784
• Sujet: Economics
• Sous-sujet: Economic Policy

This thematic third volume of the Asia Small and Medium-Sized Enterprise Monitor 2021 focuses on the digitalization of microfinance in Bangladesh. The MSME sector provides much of the income in rural Bangladesh, but its growth is constrained by limited access to affordable finance. This volume reports on a 2021 baseline study carried out in Bangladesh to pilot a randomized controlled trial to determine whether a digitalized group-based credit scheme could be introduced at an affordable price. The findings provide justification for further study on the digitalization of microfinance.

1. Introduction

This report is based on a baseline survey for the purpose of a randomized controlled trial (RCT) on the scope of the digitalization of microfinance in support of microenterprise growth in Bangladesh, with the objective of evaluating financial technology innovations and how financial innovations can help alleviate suffering during the coronavirus disease (COVID-19) pandemic. The objective of this study is to assess (i) the benefits of rural microenterprises in terms of income and productivity, and (ii) the role of microfinance in supporting microenterprise growth in Bangladesh.

Digital financial services have been introduced in recent years to work via web, mobile, and cloud services, among other ways. The most common type of digital financial services in Bangladesh are mobile financial services (MFS), which rely on mobile phone technology to deliver secured, fast, and inexpensive financial transactions such as payments and money transfers. MFS are generally characterized by low marginal costs per account or transaction and can therefore bring efficiencies of scale and cost reduction in financial transactions.

As mobile technology has the potential to enhance the efficiency of group-based credit and savings services for the poor, it is possible that integrating mobile phone technology, in particular MFS, into the financial transactions of microfinance institutions (MFIs) in Bangladesh, who support the poor’s income-generating activities, can save time and money in loan disbursement and collection, cash management, document processing, and verification of potential clients. Consequently, branch operations and staff activities are likely to be more efficient. If much of the cost savings can be transferred to clients, with the likelihood of interest rates being decreased, it is expected that digitalization of microfinance operations could be a win–win scenario for the providers and clients alike. Microfinance supports mostly rural microenterprises, also known as rural nonfarm activities (Khandker et al. 2016). Digitalization of microfinance can thus enhance productivity and growth more effectively in micro, small, and medium-sized enterprises (MSMEs), which currently dominate output and employment in the nonfarm rural sector of Bangladesh.

According to the Bangladesh Household Income and Expenditure Survey (HIES) of 2016, 59% of the income of rural households comes from the nonfarm sector: 33% from non-agriculture wages; 11% from non-agriculture self-employment; and the remaining 15% from rental, remittance (domestic and foreign), transfers, and so on. The same survey finds that the share of employment among rural households in the nonfarm sector is 51%. On the other hand, the 2013 Economic Census and the 2013 Enterprise Survey reveal that 99% of all nonfarm enterprises in Bangladesh were in the micro and small categories, and they provided employment to 20% of the population of 160 million (World Bank 2019).

Inadequate access to finance for MSMEs is a major barrier to the development of the MSME sector and its contribution to the overall financial and economic development of the country. MFIs in Bangladesh in general support six broad types of loans: (i) microcredit for self-employed activities, (ii) microenterprise loans, (iii) loans for the ultra-poor, (iv) agricultural loans, (v) seasonal loans, and (vi) loans for disaster management. Loans up to Tk50,000 are generally considered microcredit, while those of more than Tk50,000 but below Tk1 million are considered microenterprise loans.

While MFIs are major sources of credit for microenterprises, funding is observed to be inadequate to meet demand. The economic slowdown due to the COVID-19 pandemic over the last 2 years has perhaps worsened the situation and limited microenterprises’ access to finance. Like many other economic sectors in Bangladesh, microfinance was not spared during the pandemic, which severely affected microfinance customers and service providers alike. During the peak of the lockdown, the activities supported by microcredit were affected in a number of ways. First and foremost, customers of microfinance products and services—who are a vulnerable segment of the population to begin with—lost their job or their income was reduced, limiting their ability to consume such products and services. Second, lockdowns disrupted the country’s supply chains, which affected microfinance customers’ ability to carry out their income-generating activities. Restricted supply chains also raised the price of consumer goods in general, affecting everybody, including microfinance borrowers.

These constraints also decreased their ability to repay microloans, resulting in increased loan defaults. Because of reduced loan disbursements and collections, MFI operations also suffered, many branches were closed down and many MFI staffs lost their jobs. Thus, COVID-19 had a spiraling effect on both the borrowers and service providers of microfinance. To mitigate the crisis, the government offered a range of stimulus packages for microenterprises, with a total amount of $1.2 billion. However, the amount was not enough for the large demand for funding from microenterprises. It was estimated that MFIs needed at least an additional $2.2 billion (ADB 2021).

Microenterprises, generally defined as those activities employing 1–9 people, dominate the rural nonfarm sector in Bangladesh—about 98% of rural nonfarm activities are microenterprises. As per the projection of the country’s largest on-lending funding agency of microfinance, Palli Karma Sahayak Foundation (PKSF), $170 million of additional financing was deemed necessary for post-COVID-19 economic recovery to meet microenterprises’ demand for working capital. It is an open issue to determine if MFIs and borrowers managed (and, if so, how) to cope with the negative consequences of this historical global health shock. In general, while the Government of Bangladesh provided payouts for vulnerable populations, the amount was modest and mostly included a one-time transfer payment. However, MFIs in Bangladesh showed resilience in the past by supporting their borrowers with emergency or disaster loans, loan moratoriums, and other schemes such as training programs on coping strategies. It is an important research question to determine to what extent government assistance was provided and how successful it was in mitigating the crisis in the MSME and MFI sectors.

This study addresses several related issues surrounding microfinance, MSMEs, and COVID-19. In order to study these interlinked issues, baseline data collection was carried out during October–December 2021. The purpose of this baseline is to help design an RCT study to assess how a digitalized version of microfinance can be a win–win situation for both MFIs and MSMEs. However, the data collected were for households engaged in microenterprises, as well as other rural activities including farming, and for branch-level information of MFIs serving these households. The data therefore have the potential to facilitate investigation of a number of outstanding related issues: (i) how MFI borrowers were affected by the COVID-19 pandemic health-wise and economically; (ii) how their microfinance transactions were affected (receiving loans, repayments, and savings); (iii) how their income-generating activities, which are supported by MFIs, were affected; (iv) how MFI branch activities were affected by COVID-19; (v) whether MFI borrowers utilize MFS services to secure and repay loans or deposit savings; and (vi) whether microenterprises received any stimulus money offered by the government or the MFI on-lending agencies. Moreover, the scope of this data collection under the baseline survey is to facilitate examining how MFI operations were likewise affected by the pandemic and whether they received any stimulus money for their expected losses in loan collection and savings mobilization due to COVID-19.

Our approach in this study is consistent with a large literature on the subject of access to finance and its role in microenterprise growth. A large body of literature has already documented that better access to affordable finance (in terms of better terms and conditions of loans as well as reliable source) is essential for improved productivity and growth in any economy (e.g., Butler and Cornaggia 2011, Cull and Xu 2005, Sawada and Zhang 212, Wang 2008). In fact, while other obstacles matter, a lack of access to finance consistently emerges as the single most important and robust factor constraining firm enterprise growth (e.g., Aterido et al. 2011, Beck et al. 2006, Beck et al. 2005, Buyinza and Bbaale 2013, Deininger and Jin 2007, de Mel et al. 2008, Rand 2007).

Our study uses baseline data collected from households, communities, and MFI branches active in the study villages. Data analysis confirms some stylized facts. Microenterprise dominates the rural nonfarm sector in terms of income–more than 90% of nonfarm income and about 70% of total household income is drawn from MSME activities. MSMEs consist of cottage and small manufacturing and other businesses and services that play an important role in the rural economy. Startup capital for these enterprises comes primarily from their own savings or the support of family members—not from banks or even MFIs. However, MFIs extend support to existing MSMEs—they lend about 45% of their loan portfolios to MSMEs once they have been established using their own resources.

Credit constraint is found to be an important factor for explaining lower returns to productivity. Nonetheless, the estimated rate of return to capital is about 20%. Many MFIs charge more than 20% interest for loans extended to microenterprises. The banks charge less than 15% interest (more recently, about 10%) and often extend a loan beyond 1 year, whereas the average duration of an MFI loan is 1 year. Banks are reluctant to finance MSMEs despite the fact that loan recovery rates are observed to be more than 95%. (The loan recovery rate was close to 90% even during the pandemic.) However, pandemic-related stimulus did not reach most of these enterprises and, on average, they lost as much as 20% of their income during this period.

While as much as 70% of rural households have MFS membership, their (MFS) impact on enterprise productivity is not much. On the other hand, as microenterprises depend on microfinance (they draw more than 60% of their loans from MFIs), there is merit for digitalizing microfinance. This is justified for a number of reasons such as the cashless and speedy transfer of money. But a very small percentage of MSME households (about 4%) are found to use MFS to pay loans or deposit money as savings, although digitalizing lending of MFIs has not been yet introduced for their clients. Hence, there is scope for an experimental research design evaluating whether and how microfinance lending (as well as savings and loan repayment and insurance) can be digitalized. The baseline study design to study this question is therefore justified on this ground.

The report is organized as follows. Section 2 discusses the survey design and salient features of the data collected from three major stakeholders of MSMEs: (i) households that are engaged in MSME activities as compared with those who are not—MSMEs are largely cottage and small enterprises employing up to 10 persons (household survey); (ii) MFIs supporting these enterprises (MFI survey); and (iii) communities where these households and enterprises are situated (community survey). This section also discusses the sampling frame for data collection, which was the basis for the random drawing of 124 surveyed villages (communities), 2,993 households, and 285 MFI branches.

Section 3 discusses how the MSME sector is defined and its current state, both in terms of coverage and sector distribution in Bangladesh, as per government statistics such as the latest round of the Economic Census of 2013 and different rounds of the HIES. It also discusses the financing and regulatory authorities governing the sector’s performance.

Section 4 presents and discusses the trend of microfinance growth in terms of outreach (measured by membership and loan disbursement, and savings mobilized), and financial efficiency indicators (profitability and interest rates on lending and savings) using aggregate time-series national-level as well as branch-level data of the three largest MFIs covering some 80% of MFI transactions. Branch-level data permits a comparison of MFI performance before and after COVID-19 hit the economy and impacted the MSME sector.

Section 5 discusses the alternative sources of finance to MSMEs. The financial services consist of four major sources—banks, MFIs, MFS, and informal finance. Data show that the largest sources of MSME finance are MFIs and MFS. While MFIs are an outlet for MSMEs to save and borrow, MFS are the only source for transferring money for different purposes; they are used to remit and receive money. Less than 1% of MFI users use MFS to repay loans and save money. Analysis of household- and enterprise-level data shows that as much as 70% of households and 80% of MSME owners have both MFI and MFS accounts.

Section 6 discusses the sources of startup capital and estimates rates of return to capital for all three categories of microenterprises. It also examines the role of improved access to financial services in MSME productivity and income. Both financial accounting data and a production function approach is used to estimate the return on assets (ROA). Results are encouraging—returns to capital are as high as 37% and as low as 22%. Despite such high returns to investment for MSMEs, commercial banks are reluctant to finance them—only 19% of MSMEs have an account with banks. In contrast, MSME access to MFIs is as high as 60%. Yet, MSMEs are credit constrained and thus the question remains of how to provide capital at affordable prices to MSMEs. Banks charge about a 9% rate of interest against loans extended for a period longer than 1 year, while MFIs extend loans at a rate of about 24% for a short duration (often 1 year). So, improved access to long-term loans with lower interest rates is a way to enhance productivity and growth in the MSME sector.

Section 7 estimates the role of microfinance beyond microenterprises. Households that are engaged in MSME activities are also often involved in other income-generating activities. Therefore, the role of microfinance cannot be limited to evaluating its impact on MSMEs only. Rather, the MFI role needs to be evaluated against overall income generation. This section precisely estimates the net effect of MFIs on household income from farm and nonfarm sources as well as on overall income. Findings support the fact that MFIs play a positive role in raising both farm and nonfarm income. MFIs’ impact is significant on nonfarm income for MSME households and on farm income for non-MSME households.

Section 8 examines the extent of COVID-19’s impact on microenterprise productivity. Data analysis shows that microenterprises were hit hard by the pandemic. Households engaged in MSME activities lost income and suffered unemployment more than those who were not engaged in MSME activities. Such findings contradict the government perception that stimulus money must have reached the target groups. In fact, the econometric analysis does not find any such impact from stimulus measures, nor does it find that MFIs played any role in mitigating the negative impact of COVID-19. The pandemic impacted MSME productivity adversely, which could not be mitigated by external forces such as government stimulus money.

Section 9 concludes the report with implications for the future study of the digitalization of microfinance. The baseline survey data analysis justifies undertaking future study that assesses how the financial services of MFIs can be digitalized to help boost growth in the MSME sector, which provides the lion’s share of rural income in Bangladesh. As MFIs largely support MSME activities, it follows also that the way financial institutions, including commercial banks, currently operate is not of much use unless their services are digitalized.

2. Salient Features of Census, Household, and Village Sample

In this section, we briefly discuss some of the salient features of the baseline data we collected through a village census, household surveys, and MFI branch surveys. As part of the sampling framework of the study, we first conducted a census in 150 selected villages in five districts of the Rangpur division in Bangladesh. The region is regarded as the poorest in terms of headcount ratios according to the latest 2016 HIES conducted by the Bangladesh Bureau of Statistics (see Figure A.1 in the Appendix). Later, from the census list of 124 villages we randomly draw 3,000 households for the baseline study, keeping proportionality of mainly MFI participation and microenterprise households.¹ In these selected regions, we also conduct MFI branch surveys among the MFIs who are reported to have been serving the census households. Also, we conduct key informant interviews in all the selected villages to understand village-level characteristics. We report detail information in the Appendix.

A total of 24,708 households participated in the census across all five districts. Most of these (31%) were four-member households, similar across all districts. More than 70% of household heads were in the age range of 31–60 years and this percentage is also consistent across all five districts. A majority (93%) of surveyed households were headed by men. More than 41% of household heads had no formal education, while 14.5% of the household heads had completed primary education. Less than 4% of household heads had tertiary education. Almost 63% of households have no landownership. This percentage is highest in Kurigram and Rangpur (both around 67%) and lowest in Lalmonirhat (57%). Most of the households (more than three out of every four) do not own or operate any enterprise. This percentage is the lowest for Kurigram villages (66%) while relatively similar (about 75% to 79%) for the other four districts. Wholesale and retail business are the most common type of business (13%) and this is similar across all districts, followed by transport and communication (5.7%), which is highest in Gaibandha (almost...

Table des matières

• Front Cover
• Title Page
• Copyright Page
• Contents
• Tables and Figures
• Foreword
• Acknowledgments
• Abbreviations
• Executive Summary
• 1. Introduction
• 2. Salient Features of Census, Household, and Village Sample
• 3. Defining MSMEs and Their Relevance to Meeting the Sustainable Development Goals
• 4. State of Microfinance and Microfinance Institutions
• 5. Microenterprise Access to Financial Services
• 6. Assessing the Impact of Financial Services on Microenterprise Productivity
• 7. Assessing the Impact of Microenterprise on Rural Income: Does Access to Microfinance Matter?
• 8. Mitigating Risk in Microenterprise Production and Rural Income
• 9. Conclusion
• Appendix: Sample Design and Data Collection and Description
• References
• Footnotes
• Back Cover