Grow the pie or have it? Using machine learning for impact heterogeneity in the Ultra-poor Graduation Model

Abstract

ABSTRACT: Anti-poverty interventions often face a trade-off between immediate reduction in poverty, measured by consumption, and building assets for longer-term gains. An “Ultra-poor Graduation” model, found effective on both dimensions in several rigorous studies, generally leans towards asset building. By using data from a large-scale RCT in Bangladesh, we find significant variation in impact on assets where the top quintile gainers experience asset growth of 344% while asset growth is only 192% for the bottom quintile. Heterogeneity in impact on household expenditures is found to be present but of lower magnitude than that of assets. Importantly, the machine learning techniques we apply reveal contrasts in characteristics of beneficiaries who made the most in assets vs. consumption. The results identify beneficiary characteristics that can be used in targeting households either to maximize impact on the desired dimension and/or to customize interventions for balancing the asset and consumption trade-off