While it is well known that individual behavior may be predicated on underlying social influences, the specific mechanism through which such influences operate remains poorly understood. Social effects are known to have played an important role in transforming traditional agriculture and changing reproductive behavior in many third world countries over the past four decades. In the United States, social or neighborhood effects are believed to be important determinants of contraceptive use and sexual behavior, especially among teenagers and population-groups at high-risk of contracting AIDS and other sexually transmitted diseases. External information programs also play a significant role in motivating changes in behavior and practice, in conjunction with the social effects described above. The effectiveness of such programs would be greatly enhanced if their design took account of the social externality associated with changes in individual behavior. The proposed research attempts to distinguish between two alternative social effects, peer-influence and social learning, as independent determinants of changes in individual behavior. The theory is applied to study changes in contraceptive use in Matlab district, Bangladesh. The International Center for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) launched a family planning program covering seventy villages in Matlab district in 1977. While the ICDDR,B family planning program has clearly contributed significantly to the increased prevalence of contraceptive use, there remains substantial variation in the rate of adoption across villages in the treatment area which we attempt to explain through the social effects mentioned above. The Matlab data-set provides us with a unique opportunity to study the role of social effects in individual decision-making since detailed household-level information is available within each social unit (village) over an extended period of time. In addition to the richness of the data, a real advantage of studying rural households in a developing country is that the composition of the village remains relatively stable over time, allowing us to accurately measure the social effects of interest and their role in changing individual behavior. While the empirical results will be obtained in a particular institutional environment they are likely to be more broadly relevant; providing us with an understanding of the relative importance of the alternative social effects as well as the mechanism through which such social effects operate. The results of the study could consequently be extended to contribute to the optimal design of information programs aimed at changing sexual behavior and practice in the United States.
