Development as a Process of Social Change: An Application to the Fertility Transition

This paper provides an explanation for two features of the development process that have been observed in many different settings: the slow response to external interventions, and the wide variation in the response to the same intervention. One interpretation of these stylized facts is based on the idea that individual behavior is often socially regulated in a traditional economy. When the economic environment changes, individuals may be uncertain about the new social equilibrium that will emerge in their community. This uncertainty is resolved over time, as individuals interact sequentially with each other, and the community gradually finds its way to a new equilibrium. We apply this view of the development process to the fertility transition in rural Bangladesh, during which contraceptive prevalence changes slowly and there is wide variation in long-run levels of contraception across villages. At the individual level, estimated changes in the contraception decision rule over time match well with our learning-based characterization of social change. Further, women respond strongly to contraceptive prevalence within their own religious group within the village, while cross-religion effects are completely absent, consistent with the view that social interactions among the women, which almost never cross religious boundaries, gradually led to changes in reproductive behavior in these communities. ∗We thank Dan Ackerberg, Orazio Attanasio, Abhijit Banerjee, Sam Bowles, Joytsnamoy Chakraborty, Esther Duflo, Stephen Donald, Jan Eeckhout, Andrew Foster, Jeroen van Ginniken, Michael Kremer, Kevin Lang, Nancy Luke, Dilip Mookherjee, Jack Porter, Andrew Postlewaite, Mark Rosenzweig, Frank Schorfheide, T. Paul Schultz, Frank Vella, Peyton Young, numerous seminar participants and two anonymous referees for helpful comments. We are especially grateful to George Mailath for many helpful discussions that substantially improved the paper. Mohammad Ali expertly assembled much of the data, Aldo Colussi, Nauman Ilias, Poorti Marino and Chun-Seng Yip provided superb research assistance. Munshi’s research was funded by NIH grant R03-HD35516. We are responsible for any errors that may remain. †University of Pennsylvania and MIT ‡International Centre for Diarrhoeal Disease Research, Bangladesh