Microfinance competition and multiple borrowing: Evidence using panel data from Bangladesh

• Published date: 01 May 2022
• Author: Minhaj Mahmud,Yasuyuki Sawada,Mari Tanaka
• Date: 01 May 2022
• DOI: http://doi.org/10.1111/rode.12869

1164 

|

     wileyonlinelibrary.com/journal/rode  Rev Dev Econ. 2022;26:1164–1188.

© 2022 John Wiley & Sons Ltd

1

|

INTRODUCTION

Microcredit (small- scale loans) has emerged as one of the most powerful policy tools for provid-

ing financial assistance  to the poor. The number of  poor households  with access to  microcredit 

programs has increased globally by more than 18 times over the past 20years, from 7.6 million in 

Received: 18 March 2021 

|

  Revised: 14 January 2022 

|

  Accepted: 15 January 2022

DOI: 10.1111/rode.12869  

REGULAR ARTICLE

Microfinance competition and multiple

borrowing: Evidence using panel data from

Bangladesh

MinhajMahmud1*

|

YasuyukiSawada2

|

MariTanaka3

*This work has been completed when he was at the Bangladesh Institute of Development Studies. 

1Asian Development Bank, Manila, 

Philippines

2University of Tokyo, Bunkyo- ku, Tokyo, 

Japan

3Hitotsubashi University, Kunitachi, 

Tokyo, Japan

Correspondence

Minhaj Mahmud, Asian Development 

Bank, Manila, Philippines.

Email: mmahmud@adb.org

Funding information

JICA Research Institute

Abstract

We  examine  the  causes  and  consequences  of  multi-

ple  borrowing  in  rural  Bangladesh  using  long- term 

household  and  village  panel  data  covering  the  years 

2000– 2014.  We  find  that  the  sharply  growing  number 

of microfinance institutions (MFIs) in a wider set of vil-

lages over time coincides with a corresponding increase 

in household borrowing from multiple MFIs and house-

holds  accessing  loans  generally  as  well  as  significant 

rises in  the  total values  of  household assets,  especially 

in the form of agricultural equipment. Although the in-

creasing number  of MFIs  resulted in  some households 

borrowing to repay  previous loans, the  fraction of such 

households is relatively small.

KEYWORDS

Bangladesh, household borrowing, microbusiness, microfinance 

institutions

JEL CLASSIFICATION

G21; D14; O12; O16

|

1165

MAHMUD et al.

1997 to 139.9 million in 2018 (Microfinance Barometer, 2019). There has been rapid expansion of 

the microfinance sector, with increasing numbers of  both microfinance institutions (MFIs) and 

borrowers. Bangladesh  has been  the global  center of  this  trend, witnessing  rapid expansion  of 

MFIs and their branches.1 Despite the 2006 Nobel  Peace prize being jointly awarded to both the 

Grameen Bank and its founder  Professor Muhammad Yunus, there remains little consensus  on 

the impact of microcredit programs. While there has been some discussion in the literature about 

the positive role of microcredit in household welfare, including its cost- effectiveness,2 recent evi-

dence based on randomized control trials provides rather weak evidence of their positive impact 

(e.g. Banerjee etal.,2015).

The overall impact of  improved access  to credit on the  welfare of  borrowers ultimately rests 

on whether the microcredit programs relax the binding credit constraints on borrowers (i.e. mit-

igating credit market imperfections) or merely induce over- borrowing (i.e. borrowing more than 

the borrower has the capacity to repay). It may also depend on the complementarity or substitut-

ability between a loan  from an MFI and other credit sources  due to the nature  of liquidity  con-

straints, risk- sharing, borrower  production functions, and mechanisms producing  downstream 

outcomes (Banerjee etal.,2015).

Notwithstanding the fact that major MFI lenders had agreed not to work with the same clients, 

in the absence  of effective  credit bureaus, their  rapid expansion inevitably  results in  borrowers 

borrowing from multiple  lenders, including MFIs— a  phenomenon widely  known as  “overlap-

ping borrowing” or “multiple  borrowing.” This raises the important question of  whether inten-

sifying competition among  MFIs is a  predictor of  increasing overlapping borrowing,3  and if  so, 

what are the consequences of this for household welfare. However, to the best of our knowledge 

there has been limited rigorous empirical research looking into the causes and consequences of 

overlapping borrowing, and,  in particular, the role  of competition  between MFIs on  borrowing 

behavior and household welfare.

In this paper we aim to at least partially bridge this gap in the literature by analyzing the asso-

ciation between the increasing number of MFIs and household borrowing behavior (e.g. multiple 

borrowing, total amount of loans, and the purpose of  loans) and household asset accumulation. 

To this end, we use four waves of  representative household panel survey in Bangladesh covering 

62 villages from 2000 to 2014. To link this data set to the number of MFIs in the village and other 

village characteristics, we combine the data with our own retrospective survey conducted in 2014 

on village- level  information for the  same period.  Our data indicate  that the  average number of  

MFIs per village increased from 3.16 in 2000 to 4.34 in 2014, alongside a rapid rise of the share of  

multiple borrowing households from 7% in 2000 to 23% in 2014.

Overall, we find that an increase in the number of MFIs in a village predicts a higher probabil-

ity that households will borrow from multiple MFIs and  a higher probability of their borrowing 

from at least  one MFI;  however, there is  no change  in the  probability of  borrowing from  other 

sources. In addition, increasing  MFI competition in the  village predicts larger household  assets 

that are mostly used  for productive purposes  (such as agricultural equipment)  rather than only 

for repayment or  consumption. Consistent  with  this evidence,  the growing  number of  MFIs  is 

associated with increasing multiple borrowing for agricultural or  business purposes rather than 

consumption purposes. This  result is  consistent with  the hypothesis  that combining  small and 

inflexible loans from multiple MFIs serves to satisfy the high demand for borrowing for the pur-

pose of purchasing  productive assets.  On the  other hand,  we find  that  the growing  number of  

MFIs is also associated with a small but statistically significant increase in households borrowing 

for repayment for previous loans. We identify that these households borrowing for repayment are 

small in fraction (2% of our sample) and have less assets and lower income.