Working Papers
Taxation, Information and Withholding: Evidence from Costa Rica
With Marco Hernandez
Coverage from La Nación
Winner of the World Bank Innovation Prize
Transnational Terrorist Recruitment: Evidence from Daesh Personnel Records
With Mohamed Abdel Jelil, Kartika Bhatia, Quy-Toan Do and Clément Joubert
Coverage from Le Monde, Times of Israel, Times of India, Alarabiya, World Bank Blog
Selected Works in Progress
Audit Selection Under Weak Fiscal Capacity: A Field Experiment in Senegal
With Pierre Bachas and Bassirou Sarr
Supported by two EDI grants (pilot and scale-up)
AEA RCT Registry
Should low-income countries leave discretion to tax inspectors to select firms for audit or should selection be determined by risk-scoring algorithms? Does the optimal amount of discretion depend on the quantity of third-party information available and its ease of access for tax inspectors? In a context with weak fiscal capacity, leveraging inspectors’ private information could be valuable but opens the door to discrimination and corruption. This project tests which selection method (discretionary inspector selection or algorithm selection based on risk scores) is most effective in detecting firm non-compliance and increasing audit yield. An experiment in collaboration with the tax administration in Senegal is currently in the field.
Supported by two EDI grants (pilot and scale-up)
AEA RCT Registry
Should low-income countries leave discretion to tax inspectors to select firms for audit or should selection be determined by risk-scoring algorithms? Does the optimal amount of discretion depend on the quantity of third-party information available and its ease of access for tax inspectors? In a context with weak fiscal capacity, leveraging inspectors’ private information could be valuable but opens the door to discrimination and corruption. This project tests which selection method (discretionary inspector selection or algorithm selection based on risk scores) is most effective in detecting firm non-compliance and increasing audit yield. An experiment in collaboration with the tax administration in Senegal is currently in the field.
Taxing Property in Developing Countries: Theory and Evidence from Mexico
With Alejandro Estefan, Juan Carlos Suárez Serrato and Karina Ramirez
We study the design and effectiveness of property taxation in Mexico City -- a context with household credit constraints and limited government enforcement capacity. We first evaluate the effects of three policy levers used by the government to increase tax revenues: tax rates, late fees, and enforcement. Using regression discontinuity and difference-in-differences research designs, we show that tax rate increases raise revenue but also result in non-compliance. We estimate revenue elasticities of 0.3-0.7, suggesting that the current tax rate is below the revenue-maximizing rate. We then use time variation in late fees to estimate the likelihood that households pay on time. Households are very responsive to late fees, with a payment elasticity of 1. Finally, we use a field experiment that delivered an enforcement message to show that households are also very responsive to enforcement efforts. Indeed, the share of households paying their outstanding liability increases by 7-10 percentage points. We then provide guidance on how to best make use of these policy levers. From a revenue-maximizing perspective, all tools are underutilized and the optimal policy mix would ensure the revenue elasticities of these policy instruments are equal to each other. In contrast, under a welfare-maximizing perspective, which takes into account households' credit constraints, enforcement efforts and excessive late fees reduce welfare relative to gains in revenue. This is because credit-constrained households comply with the property tax by reducing consumption. The welfare-maximizing perspective suggests that tax rates should be used to collect revenue from non-constrained households and that late fees can be used to provide liquidity to credit-constrained households. These results show that governments in developing countries can raise substantial revenues using property taxes. In addition, improving the design of property taxes can make these taxes a more effective source of revenue.
We study the design and effectiveness of property taxation in Mexico City -- a context with household credit constraints and limited government enforcement capacity. We first evaluate the effects of three policy levers used by the government to increase tax revenues: tax rates, late fees, and enforcement. Using regression discontinuity and difference-in-differences research designs, we show that tax rate increases raise revenue but also result in non-compliance. We estimate revenue elasticities of 0.3-0.7, suggesting that the current tax rate is below the revenue-maximizing rate. We then use time variation in late fees to estimate the likelihood that households pay on time. Households are very responsive to late fees, with a payment elasticity of 1. Finally, we use a field experiment that delivered an enforcement message to show that households are also very responsive to enforcement efforts. Indeed, the share of households paying their outstanding liability increases by 7-10 percentage points. We then provide guidance on how to best make use of these policy levers. From a revenue-maximizing perspective, all tools are underutilized and the optimal policy mix would ensure the revenue elasticities of these policy instruments are equal to each other. In contrast, under a welfare-maximizing perspective, which takes into account households' credit constraints, enforcement efforts and excessive late fees reduce welfare relative to gains in revenue. This is because credit-constrained households comply with the property tax by reducing consumption. The welfare-maximizing perspective suggests that tax rates should be used to collect revenue from non-constrained households and that late fees can be used to provide liquidity to credit-constrained households. These results show that governments in developing countries can raise substantial revenues using property taxes. In addition, improving the design of property taxes can make these taxes a more effective source of revenue.
Electronic Payment Technology and Tax Compliance: Evidence from a Financial Inclusion Reform
Economic development leads to the expansion of electronic transaction technology such as credit and debit cards. This may increase firms’ tax compliance, as electronically recorded sales are more difficult to hide than cash transactions. Exploiting quasi-experimental variation from a large financial inclusion reform, we study the effect of firm-level incentives (subsidies for the rental fee of credit card machines) and consumer-level incentives (VAT-rebates for payments with a credit/debit card) on technology take-up, firm’s reported sales and tax liability.
Do Tariff Reforms Pay for Themselves? Evidence from Production Networks in Rwanda and Costa Rica
With Pierre Bachas, Anders Jensen and Gabriel Tourek