Savings, sanctions, and support: Essays on collective action and community organizations in Kenya

This thesis consists of three essays on collective action in local organizations, drawing on primary data and research conducted by the author in western Kenya. The thesis is organized in four chapters. Chapter 1 presents an overview and synopsis of the thesis; chapters 2 through 4 are the main essays.; Chapter 2 poses the question: why do individuals choose to join rotating savings and credit organizations (roscas) that are characterized by inflexibility, high costs, and low rates of return? The chapter shows how current theories cannot explain the structure of many roscas and presents an alternative hypothesis: saving requires self-discipline, and roscas provide a collective mechanism for individual self-control in the presence of time-inconsistent preferences and in the absence of alternative commitment technologies. The chapter tests this hypothesis using detailed primary data on 70 roscas in western Kenya and finds that the structure of these roscas is consistent with the commitment hypothesis.; Chapter 3, jointly authored with Ted Miguel, investigates how ethnic diversity affects community participation in rural primary schools. A theoretical model proposes that social sanctions imposed within ethnic communities are the key mechanism used to overcome free-rider problems in village economies. School committee meeting records from western Kenya indicate that ethnically diverse schools do impose significantly fewer sanctions on parents. The empirical results also show that school committees in ethnically diverse areas are plagued by low parent participation in school activities and low teacher attendance and motivation.; Chapter 4 uses the results of a prospective, randomized evaluation to examine the impact of a development program explicitly targeted at promoting agricultural output and strengthening organizational capacity among rural women's groups. The results indicate that funding does not lead to sharply higher agricultural output among women's groups receiving funding, but does lead to more turnover among group members, as well as increased entry into groups and group leadership by younger, more educated women, by women employed in the formal sector, and by men. These results raise the possibility that assistance to indigenous organizations of women may change the very characteristics of these organizations that made them attractive to funders in the first place.