Failure vs. displacement: Why an innovative anti-poverty program showed no net impact in South India

We analyze a randomized trial of an innovative anti-poverty program in South India, part of a series of pilot programs that provide "ultra-poor" households with inputs to create new, sustainable livelihoods (often tending livestock). In contrast with results from other pilots, we find no lasting net impact on income or asset accumulation in South India. We explore concerns with program implementation, data errors, and the existence of compelling employment alternatives. The baseline consumption data contain systematic errors, and income and consumption contain large outliers. Steps to address the problems leave the central findings largely intact: Wages for unskilled labor rose sharply in the area while the study was implemented, blunting the net impact of the intervention and highlighting one way that treatment effects depend on factors external to the intervention itself, such as broader employment opportunities.