Staying for Benefits
利用孟加拉国一个健康与家庭计划项目的准随机布局，发现该项目使迁出人口流降低了约15%，且效果持续存在，表明福利项目能长期改变人口分布，对项目评估和经济发展有启示。
Abstract There is concern that benefit programs may alter out-migration patterns. We exploit the quasi-random placement of a health and family planning program in Bangladesh to examine changes in out-migration patterns. Using individual-level migration data from 1979–91, we find that the flow of out-migration was approximately 15 percent lower for women and men in the treatment versus comparison area. We find comparable changes in outmigrant stock, showing that effects persisted even after allowing for return migration. Our findings suggest that benefit programs can lead to longer run changes in population, with consequences for program evaluation design and economic development.