Office of Trade and Labor Affairs Contract Research Program Isolation or Opportunity? Experimental Analysis of a Housing Program for Urban Slum Dwellers in India *

How does giving the urban poor an opportunity to relocate to higher quality but more remote housing influence their well-being in the long run? Slum relocation programs, which move slum dwellers from the city center to the periphery, are widely used, however relatively little is known about their effect. In this paper, we evaluate the long-run impact of a housing program in a large Indian city that offered 110 slum dwellers, chosen by lottery, a mortgage and the opportunity to purchase a new house located in a relatively distant residential complex. Roughly 14 years later, lottery winners and non-winners look similar in terms of income, work and children’s outcomes. Winners are more likely to own a house than non-winners, but also live farther from the city center. Winners are more isolated from traditional family and caste networks and have less access to risk-sharing arrangements. Additionally, winners report both less social insurance and greater collective action to benefit the community than non-winners. These patterns suggest that isolation strengthened winners neighborhood ties but weakened their traditional family and caste relationships. This, in turn, reduced barriers to local cooperation but increased the correlation of risk within informal insurance networks. ∗The authors are from the Institute for Financial Management and Research (Chennai), Duke University, and Harvard University. For financial and field support, we thank the National Science Foundation (grant SES-0752792), The US Department of Labor ILAB (DOL ILAB), the Centre for Microfinance (CMF) at IFMR, the Mossavar-Rahmani Center for Business and Government, and the Real Estate Academic Initiative at Harvard University. We thank Manasee Desai, Janaki Kibe, Keshubhai, Michael Mathai, Sreelakshmi Papineni, Vanya Pasheva, Natalia Rigol, Sarah Bishop and Divya Varma for excellent research assistance, Avdhut Fadanwis and Sachin Srivastava for data management and DOL ILAB staff for detailed comments on an earlier draft. All errors are our own.