Microloans, education and growth

• Date: 01 November 2018
• DOI: http://doi.org/10.1111/rode.12526
• Author: Patrick M. Emerson,Bruce McGough
• Published date: 01 November 2018

REGULAR ARTICLE

Microloans, education and growth

Patrick M. Emerson

1,2

|

Bruce McGough

3

1

Oregon State University, Corvalis,

Oregon

2

Also affiliated to IZA, Bonn, Germany

& C-MICRO, FGV, São Paulo, Brazil

3

University of Oregon, Eugene, Oregon

Correspondence

Patrick M. Emerson, Economics Program,

School of Public Policy, 332 Bexel Hall,

Oregon State University, Corvalis, OR

97331.

Email: patrick.emerson@oregonstate.edu

Abstract

This paper constructs a two‐period overlapping genera-

tions model of human capital investment decisions where

a microloan program designed to finance entrepreneurial

activities is active. It is shown that, in the presence of

human capital externalities, microloans that are small and

have immediate repayment can be growth depressing, and

welfare decreasing, through their effect on the opportu-

nity cost of schooling. By increasing the opportunity cost

of schooling, such microloans divert investment away

from human capital: by failing to internalize the social

returns to education, households’individually optimal

investment decisions in the face of microcredit availabil-

ity act to depress the growth of the economy and result

in suboptimal welfare outcomes. Conditions under which

these negative effects can occur are identified and poten-

tial solutions are suggested.

1

|

INTRODUCTION

Microfinance and microloans are among the most celebrated development success stories of the

last quarter century. The Microcredit Summit Campaign estimates that in 2007, microcredit had

reached almost 155 million recipients, almost 110 million of whom were women and over 106

million were considered to be among the world's poorest (Daley‐Harris, 2009). Despite the popu-

larity and spread of microcredit, relatively little is known about the effect of credit receipt on the

outcomes of individuals and households. The purpose of this paper is to provide insigh t into one

aspect of microloans that has, as yet, received little attention: that the nature of microloans can

cause lower investment in human capital, which can have growth and welfare implications for

economy. It is demonstrated through the use of an overlapping generations’model that even when

microloans are introduced in the form of a new capital injection into an economy there are circum-

stances in which they can lead to welfare loss and negative growth effects. Isolating the aspects of

microloans that can lead to lower welfare and growth can assist in the design of loan programs that

are less likely to encounter such negative consequences.

DOI: 10.1111/rode.12526

e250

|

© 2018 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/rode Rev Dev Econ. 2018;22:e250–e265.

It is, perhaps, hard to understand how microloans—especially those that inject outside capital

into an essentially closed economy—could actually be welfare reducing. In a static setting, the

introduction of microloans represents an expansion of the budget set, which should result in

increased utility. We show that the small and quickly maturing nature of these loans can depress

long‐term investment in human capital.

1

If households have binding time constraints then microlo-

ans can raise the current opportunity cost of schooling, causing households to concentrate more on

entrepreneurial activities and less on education. This can have a beneficial short‐term welfare effect

for households but a detrimental long‐term effect for the household dynasty. By raising the oppor-

tunity cost of schooling and demanding a quick return on investment, microloans might actually

serve to suppress human capital accumulation in the communities in which they are introduced. In

the presence of social returns to education, the long‐term effect of microloans might actually be to

impede economic growth rather than help it, leading to decreased welfare. In the end, by acting to

suppress growth, microloan programs of this type can result in increased rather than decreased

poverty.

The paper also examines aspects of microloans that could contribute to the adverse conse-

quences of the loan program, in particular the size and repayment schedule of the loans, and dis-

cusses possible modifications to avoid these consequences.

2

|

RELATED LITERATURE

The early economic literature on microfinance focused on group liability as a way to overcome the

lack of information and collateral in low‐income country credit markets (Armendáriz de Aghion &

Morduch, 2005; Ghatak & Guinnane, 1999; Stiglitz, 1990). Group lending was shown to have the

potential to overcome these information and collateral problems. With shared liability, the entire

group becomes responsible for repayment and thus group members have an incentive to monitor

each other. Indeed, the evidence suggests that group lending and peer monitoring have been very

effective: repayment rates for microloans average over 90 percent (Grameen Foundation).

Empirical studies of the effects of microloans on the outcomes of the participants are numerous

and often conflicting. Pitt and Khandker (1998) find large positive consumption effects from

Grameen Bank loans, especially for women. Coleman (1999), however, finds little to no impact of

a microcredit program in Northeast Thailand on recipient welfare, but notes that failure to control

for selection would lead to a conclusion of positive impacts. Kaboski and Townse nd (2005) using

a natural experiment approach find positive consumption impacts of microloans, but not on invest-

ment. Karlan and Zinman (2010) use randomization of marginal clients to evaluate the impact of

consumer lending in South Africa, and find that the receipt of microcredit improves the welfare of

the recipients. In a novel approach to address the selection problem, Schroeder (2010) examines

consumption effects from Grameen Bank lending in Bangladesh using an estimation strategy that

relies on second‐moment restrictions and finds positive and significant consumption effects from

microloans.

To address the problem of selection, randomized designs have been used to explore the impact

of microfinance product design such as group lending and repayment schedules (e.g., Field &

Pande, 2008; Giné & Karlan, 2006, 2009). Banerjee, Duflo, Glennerster, & Kinnan (2009) is a

large‐scale randomized experiment that examines outcomes from a microcredit intervention. In this

study, the authors find that durable consumption rises but nondurable consumption does not. More

importantly for the current analysis, they do not find any measurable effects on health or educa-

tional investment.

EMERSON AND MCGOUGH

|

e251