A Matter of Experience? Understanding the Decline in Group Lending
| Author | Matthew Suandi,Christian Ahlin |
| Date | 01 December 2019 |
| DOI | http://doi.org/10.1111/obes.12305 |
| Published date | 01 December 2019 |
1252
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 81, 6 (2019) 0305–9049
doi: 10.1111/obes.12305
A Matter of Experience? Understanding the Decline in
Group Lending*
Christian Ahlin† and Matthew Suandi‡
†Department of Economics, Michigan State University, East Lansing, MI USA
(e-mail: ahlinc@msu.edu.)
‡Department of Agricultural and Resource Economics, University of California Berkeley,
Berkeley, CA USA (e-mail: suandima@berkeley.edu.)
Abstract
Group lending, often considered a key innovation driving the successful expansion of
microcredit across the world, appears to be on the decline. Using MIX data on microfinance
institutions (MFIs), we study this time trend, focusing on macroeconomic predictors of
group lending and MFI experience. Results suggest that any movement away from group
lending is better explained via MFIs gaining experience rather than a secular time trend. We
also find that group lending is used more extensively in poorer and low-growth economies.
We argue that these findings can be understood within established views of group lending,
and carry important implications for the historical and continuing importance of group
lending.
I. Introduction
The microcredit movement has seen dramatic success in increasing formal financial access
worldwide. An unsettled question is the role group lending has played in this movement,
and whether this role has diminished. Early economic research on microcredit focused
on group lending, particularly joint liability, as a key innovation driving the movement’s
success in expanding outreach.1Aspects of group lending beyond joint liability have also
been argued to matter, for example information reporting (Rai and Sj¨ostr¨om, 2004) and
social capital formation (Feigenberg, Field and Pande, 2013).
However, shifts of some high-profile lenders towards individual lending have recently
raised questions about whether group lending remains a useful tool for microlenders. Two
pioneers of microfinance – the Grameen Bank and BancoSol – appeared to shift awayfrom
joint liability in the late 1990s and early 2000s, introducing individual liability loans for
*We thank Jeff Biddle, Todd Elder, Peter Schmidt, Jeff Wooldridge, and two anonymous referees for helpful
comments. All errors are ours.
1For example, see Stiglitz (1990), Besley and Coate (1995), Ghatak (1999, 2000), and Ghatak and Guinnane
(1999).
[Correction added on 11 November 2019, after first online publication: Germany has been removed from Christian
Ahlin’saffiliation.]
Experience and group lending 1253
qualifying borrowers. This anecdotal evidence of a shift was corroborated by Gin´e and
Karlan (2014) and De Quidt, Fetzer and Ghatak (2018a,b), who document a time trend
awayfrom g roup lending in broader samples of microfinance institutions (MFIs).2Indeed,
using within-MFI variation in an earlier version of the dataset constructed for this paper,
De Quidt et al. (2018a) find that the average MFI share of group loans decreased by roughly
5 percentage points between 2008 and 2014.
How should this trend be interpreted? Part of the answer appears to be the increasing
commercialization of the microfinance industry, i.e. the rising share of for-profit lenders
combined with for-profit lenders’ preference for individual lending (Cull et al., 2009, De
Quidt et al., 2018a). But this compositional change does not appear to be the whole story;
for one, the trend is found within MFIs overtime (De Quidt et al., 2018a). To take a concrete
example, few would argue that the shift of Grameen away from group lending is due to a
shift towards for-profitorientation. Is the trend instead par tlyan industr y-wide evolution of
microlending best practices awayfrom g roup lending; or a response to widelyexperienced
changes in the microlending environment; or something else?
The goal of this paper is to contribute to the understanding of when and where group
lending is used, in hopes of shedding further light on how this innovative lending tech-
nique fits into the microcredit puzzle. First, we seek to document how basic features of
the macroeconomic environment predict MFI reliance on group lending. These facts are
independently interesting, and they also could help explain the time trend in group lending,
if these macroeconomic characteristics are trending similarly over time across a broad set
of countries.
Second, we dig deeper into the time trend itself, arguing that a time trend is not in-
dependently identifiable from an MFI age effect when only within-MFI variation is used.
This raises a possibility not yet considered in the previous literature: is the time trend more
accurately described as an experience effect? For example, did Grameen shift away from
group lending as a consequence of experience gained in lending?
Distinguishing between these stories is useful. If experience-based explanations (sev-
eral of which are discussed in section II) fit the data better,this would challenge the arguably
common view that there is a general evolution of lending best practices away from group
lending, i.e. that group lending is being superseded by other lending innovations and/or
obsolescing as a result of a changing lending environment. Further, it would portray group
lending as potentially very valuable historically, in the initial years of microcredit, even
if established MFIs have moved away from it somewhat. It would also suggest that group
lending may still have as much value as it did fifty years ago, for new MFIs and even
for established MFIs that are expanding among new clientele or different environments.
Finally, it would be potentially compatible with both earlier and more recent work portray-
ing group lending as critical for microcredit’s success. A common theme of these analyses
is that group lending harnesses local social assets in order to overcome asymmetric infor-
mation, limited enforcement, or lack of social capital. If so, it is quite plausible that after
an MFI has used group lending in a given environment for some time, group lending has
durably solved or at least mitigated these problems, and thus become less useful.
2Cull, Demirg¨u¸c-Kunt and Morduch (2009) appear to be the first to distinguish between group and individual
lending, using MFI data from MIX, though their focus is not on a time trend.
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial