"The authors compare the welfare effects of in situ slum upgrading programs with programs that provide slum dwellers with better housing in a new location. Evaluating the welfare effects of slum upgrading and resettlement programs requires estimating models of residential location choice, in which households trade off commuting costs against the cost and attributes of the housing they consume, including neighborhood attributes. The authors accomplish this using data for 5,000 households in Mumbai, a city in which 40 percent of the population live in slums. The precise welfare effects of resettlement programs depend on assumptions made about the ease with which workers can change jobs and also on the ethnic characteristics of neighborhoods in which new housing is located. To illustrate this point the authors consider a realistic slum upgrading program that could be offered to residents in their sample living in east Mumbai. They summarize the effects of job opportunities and neighborhood composition on welfare by mapping how compensating variation for the program changes depending on where in Mumbai improved housing is located. If program beneficiaries continue working in their original job, the set of welfare-enhancing locations for the upgrading program is small. The set increases greatly if it is assumed that workers can change jobs. The benefits of this program are contrasted with the benefits of in situ housing improvements. "--World Bank web site.