The Economics of Informal Saving Groups

Project: Research

Project Details


In the past, we have explored some aspects of the economics of Rotating Savings and Credit Organizations (ROSCAs) in Kenya. We argued that female participation in such groups can be understood as a way to commit the household to a particular saving pattern, and must thus be viewed within the context of a conflicting household decision making process (Anderson and Baland, 2002).
We are currently looking at enforcement in such groups using the same data set (Anderson, Baland and Moene, 2004). While these groups may differ significantly in their organizational structures and functions (insurance, savings, mutual credit, work cooperatives...), they all share in common that (i) participation in these groups is voluntary, and (ii) they do not and cannot rely on external enforcements. However, because of the rotational structure of roscas, the incentive for members who receive the pot earlier in the cycle to default on their later contributions is high. Moreover, the incentives of the member who receives the pot last to contribute to the pot are not at all clear. Although the issue of default is acknowledged in almost any study of roscas, enforcement problems have not been directly addressed in the previous literature. We investigate the role of various punishment schemes, including social exclusion, and then argue that, when social sanctions are weak, the institutional structure of roscas can be designed so as to minimize enforcement problems.
Currently, a new data set has been gathered in Cotonou (Benin) and will be used to test some of the hypotheses discussed above. Finally, we are also pursuing a more modest survey in Cameroon on the determinants of borrowing and lending in the microfinance institutions, where we investigate how borrowing can be used as a way to protect one's own savings against family pressures.
Effective start/end date1/09/04 → …


  • insurance
  • informal saving groups
  • ROSCAs


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.