UC Berkeley

At a time of growing inequality and under-investment in public infrastructure, my re- search has focused on understanding governments’ constraints in raising tax revenue and providing redistribution. These challenges are particularly important for low and middle- income countries: despite improvements in their institutional capacity in the last decades, their ratios of tax revenue to GDP remain much lower than OECD countries’ (Besley and Persson 2013a), and their tax and transfer systems are often distributionally neutral, instead of progressive.

In the first chapter, I ask whether developing countries with limited information and tax capacity can use the corporate income tax to raise additional revenue, and design it optimally given these constraints. I explore this question for Costa Rica, using the universe of corporate tax returns and a novel methodology which exploits the country’s unique tax design: firms with marginally different revenue face discontinuously higher average tax rates. This notch feature allows me to first estimate the elasticity of profits with respect to the tax rate, and second to separate it into its components, namely the revenue and cost elasticities. I find that firms facing a higher tax rate slightly decrease reported revenue, but considerably increase reported costs, leading to a large drop in reported profits. Using additional data sources, firms’ behavioral responses appear to occur through tax evasion, with no evidence of production responses. Taken together, this implies that Costa Rican firms evade taxes on a massive 70% of their profits when faced with a 30% tax rate. In this context, lowering corporate tax rates could increase tax revenue, since we estimate the revenue maximizing rate to be below 25%. Alternative tax rules, that limit the deductibility of costs could be preferable since they would reduce evasion opportunities on this crucial margin. The results highlight the limitations of standard business taxation as an instrument to raise revenue in developing countries.

The first chapter points to limitations for revenue collection, when given the current enforcement environment. In the Second Chapter, I study how third-party information might spread to the government, in order to improve tax enforcement. Firm level tax compliance depends on the stock of information accessible to the government. Theoretical work

(Gordon and Li 2009b, Kleven, Kreiner, and Saez 2016) highlights two specific information trails: access to formal finance and the number of employees. I test whether firms with more employees and with access to formal finance are more likely to be audited and less prone to evading taxes. I use firm-level data on 108,000 firms across 79 countries in the World Bank Enterprise Surveys and construct instruments for finance and worker-size at the industry level, using an out of sample extrapolation strategy related to Rajan and Zingales (1998). The instruments isolate variation in industry technological demand for labour and formal finance by taking the US industry distributions as undistorted benchmarks. I find that firms with more employees are more likely to be audited and to comply, but find no evidence that firms using the financial sector are under higher scrutiny.

Finally, in the third chapter, I turn to the redistributive role of governments, and study how a new technology to deliver cash transfers can be used to impact transfer beneficiaries’ trust in financial institutions and their savings behavior. It is well documented that trust is an essential element of economic transactions, however trust in financial institutions is especially low among the poor, which may explain in part why the poor do not save formally. Debit cards provide not only easier access to savings (at any bank’s ATM as opposed to the nearest bank branch), but also a mechanism to monitor bank account balances and thereby build trust in financial institutions. I study a natural experiment in which debit cards were rolled out to beneficiaries of a Mexican conditional cash transfer program, who were already receiving their transfers in savings accounts through a government bank. Using administrative data on transactions and balances in over 300,000 bank accounts over four years, I find that after receiving a debit card, the transfer recipients do not increase their savings for the first 6 months, but after this initial period, they begin saving and their marginal propensity to save increases over time. During this initial period, however, they use the card to check their balances frequently; the number of times they check their balances decreases over time as their reported trust in the bank increases. Using household survey panel data, I find the observed effect represents an increase in overall savings, rather than shifting savings; I also find that consumption of temptation goods (alcohol, tobacco, and sugar) falls, providing evidence that saving informally is difficult and the use of financial institutions to save helps solve self-control problems.