Rotating Savings and Credit Associations, Credit Markets and Efficiency

Abstract

This paper examines the allocative performance of rotating savings and credit associations (roscas), a financial institution which is observed world-wide. We develop a model in which individuals save for an indivisible good and study roscas which distribute funds using random allocation and bidding. The allocations achieved by the two types of rosca are compared with that achieved by a credit market and with efficient allocations more generally. We find that neither type of rosca is efficient and that individuals are better off with a credit market than a bidding rosca. Nonetheless, a random rosca may sometimes yield a higher level of ex ante expected utility to prospective participants than would a credit market. I. INTRODUCTION Rotating savings and credit associations (roscas) are a widely observed institution for financial intermediation. They are found all over the world, particularly in developing countries, and have heretofore received scant attention from economists.' This paper and its companion piece (Besley, Coate and Loury (1993)) constitute a first attempt to analyse their economic role and performance. Roscas come in two main forms. The first type allocates funds randomly. In a random rosca, members commit to putting a fixed sum of money into a for each period of the rosca's life. Lots are drawn and the pot is randomly allocated to one of the members. In the next period, the process repeats itself, with each previous winner excluded from the draw. The process continues until each rosca member has received the pot once. At this point, the rosca is either disbanded or begins over again. Individuals may also form a bidding rosca in which the pot is allocated via a bidding procedure. The individual who