On-Demand Assistance: Experimental Evidence from Indonesia

Given labor informality in developing countries, addressing employment shocks is challenging. We study Indonesia’s new approach for on-demand online applications for temporary cash assistance and online job training. Through 20 lotteries over two years, 11 million beneficiaries were selected. In nationwide representative surveys matched with administrative data, most beneficiaries report not receiving the assistance, with no detectable impact on employment or consumption. In contrast, survey data from applicants with working phones reveal high reported take-up and higher incomes. The results suggest that, while flexible, on-demand assistance has potential, design features that allow for flexibility could hinder receipt and, thus, impact. (JEL E21, E24, I38, J24, J46, O15)