Abstract Performance pay for tax collectors has the potential to raise revenues, but might come at a cost if it increases the bargaining power of tax collectors vis-à-vis taxpayers. We report the first large-scale field experiment on these issues, where we experimentally allocated 482 property tax units in Punjab, Pakistan, into one of three performance pay schemes or a control. After two years, incentivized units had 9.4 log points higher revenue than controls, which translates to a 46% higher growth rate. The scheme that rewarded purely on revenue did best, increasing revenue by 12.9 log points (64% higher growth rate), with little penalty for customer satisfaction and assessment accuracy compared to the two other schemes that explicitly also rewarded these dimensions. The revenue gains accrue from a small number of properties becoming taxed at their true value, which is substantially more than they had been taxed at previously. The majority of properties in incentivized areas in fact pay no more taxes, but instead report higher bribes. The results are consistent with a collusive setting in which performance pay increases collectors’ bargaining power over taxpayers, who have to either pay higher bribes to avoid being reassessed or pay substantially higher taxes if collusion breaks down.

We use unique data from over 600 Indonesian communities on what individuals know about the poverty status of others to study how network structure influences information aggregation. We develop a model of semi-Bayesian learning on networks, which we structurally estimate using within-village data. The model generates qualitative predictions about how cross-village patterns of learning relate to network structure, which we show are borne out in the data. We apply our findings to a community-based targeting program, where citizens chose households to receive aid, and show that the networks that the model predicts to be more diffusive differentially benefit from community targeting. (JEL D14, D83, D85, I32, O12, Z13)

This paper shows that adding a small application cost to a transfer program can substantially improve targeting through self-selection. Our village-level experiment in Indonesia finds that requiring beneficiaries to apply for benefits results in substantially poorer beneficiaries than automatic enrollment using the same asset test. Marginally increasing application costs on an experimental basis does not further improve targeting. Estimating a model of the application decision implies that the results are largely driven by the nonpoor, who make up the bulk of the population, forecasting that they are unlikely to pass the asset test and therefore not bothering to apply.