Micro nance games

Microfinance Games∗
Xavier Gin´ e The World Bank Dean Karlan Yale University May 1, 2009 Pamela Jakiela Washington University in St. Louis Jonathan Morduch New York University

Abstract Group-based lending contracts offer potential solutions to credit market imperfections in populations too poor to provide collateral. Group-based mechanisms drive the global microfinance movement, but, byproviding borrowers with implicit insurance against investment losses, the contracts are also vulnerable to free-riding and collusion. We created an experimental economics laboratory in a large urban market in Lima, Peru, and over seven months conducted eleven different experimental economic games that allow us to unpack microfinance mechanisms in a systematic way. We find that risk-taking broadly conformsto theoretical predictions, with dynamic incentives strongly reducing risk-taking even without group-based mechanisms. Group lending increases risk-taking by pushing risk-averse borrowers to take greater risks than they otherwise would. The effect is moderated, however, when borrowers have the opportunity to form groups on their own, and we show sorting by risk-aversion. The results clarify costsand benefits of group-based contracts. Group contracts raise loan repayment rates by creating an implicit insurance mechanism, allowing borrowers to remain in good standing with lenders despite investment losses. Thus, group-lending can facilitate profitable risk-taking while maintaining high rates of loan repayment. However, the costs are borne by fellow borrowers and fall most heavily on the mostrisk averse participants. The work provides an example of how to use framed field experiments as a methodological bridge between laboratory and field experiments.
We thank Shachar Kariv, Edward Miguel, Owen Ozier, Chris Udry, and participants in the University of Groningen 2005 conference on microfinance, NEUDC 2005, Pacific Development Conference 2007, and seminars at Yale and UC Berkeley. AntoniCodinas, Gissele Gajate, Jacob Goldston, Marcos Gonzales, and Karen Lyons provided excellent research assistance. We thank the World Bank, the Social Science Research Council Program in Applied Economics, the UC Berkeley Institute for Business and Economic Research, and Princeton University for funding.
∗

1

Introduction

Banking in low-income communities is a notoriously difficult business.Banks typically have limited information about their customers and often find it costly or impossible to enforce loan contracts. Customers, for their part, frequently lack adequate collateral or credit histories with commercial banks. Moral hazard and adverse selection, coupled with small transaction sizes, limit the possibilities for banks to lend profitably. Despite these obstacles, over the pastthree decades microfinance practitioners have defied predictions by finding workable mechanisms through which to make small, uncollateralized loans to poor customers. Repayment rates on their unsecured loans often exceed 95 percent, and by 2007 — the year after Muhammad Yunus and Grameen Bank won the Nobel Peace Prize — microfinance institutions were serving about 150 million customers around the world.This achievement has been exciting to many, and advocates describe microfinance as a revolutionary way to reduce poverty (Yunus 1999). From a theoretical perspective, though, the success has puzzling elements. Many microfinance mechanisms rely on groups of borrowers to jointly monitor and enforce contracts themselves. However, group-based mechanisms tend to be vulnerable to free-riding andcollusion; in the absence of “dynamic incentives” which raise the costs of default by denying borrowers in arrears access to future loans, it is not obvious that group-lending mechanisms should out-perform individual liability alternatives. In this paper, we explore the impact of a variety of individual and group lending mechanisms on investment decisions within a controlled laboratory environment,...