I use a redesign of the tax system in Ecuador to evaluate the effect of tax enforcement on firms' cost of capital and real decisions. I find that if a firm is included in a group that is audited annually by the tax authority with certainty, then it accesses new bank debt at a significantly lower cost. Additionally, monitored firms increased their investment in human and physical capital. I translate the decision rules that the Ecuadorian tax authority used to choose the group of monitored firms into a fuzzy regression discontinuity design to control for bias from non-random tax enforcement. Finally, I provide evidence that the effect of tax enforcement works through an information mechanism, while the results are not consistent with a behavioral channel. The wider implication is that lowering asymmetric information between borrowers and lenders is an effective and fiscally positive way for the government to increase financial access and firm investment.
Selected Works in Progress
Productivity implications of size-dependent tax enforcement
Unequal enforcement of tax collection is associated with lower aggregate productivity. One mechanism is that firms are discouraged from being large to avoid regulation. Ecuador, over the period 2010-2015, concentrated its tax auditing resources on just 5% of firms. De Simone (2019a) finds that in that setting the average monitored firm was actually better off being monitored because they accessed more and cheaper bank debt. I propose a novel mechanism for the aggregate relationship: Negative spillovers on the firms whose tax payments are not enforced directly. First, the structure of tax collection in Ecuador results in monitored firms enforcing tax collection for their suppliers but not for their customers. I use this setting to implement a differences-in-differences design comparing the tax paid by the suppliers to that of the customers of newly monitored firms. I find that the suppliers of firms monitored by the Ecuadorian tax authority, the Servicio de Rentas Internas (SRI), implicitly experienced an increase in tax enforcement as well, but without the benefit of improved credit access. Next, I test for differences in investment and firm total factor productivity between suppliers and customers of the firms the SRI monitors directly. Second, I estimate the impact of SRI monitoring on the competitors of monitored firms by estimating the effect on firm market share and by asking if lenders view not being chosen for monitoring as a negative signal. I estimate the latter by exploiting variation in industry concentration of monitored firms and by considering competitors close to the decision cutoff that were not chosen. This paper provides evidence for a new mechanism behind the aggregate relationship between size-dependent tax enforcement and aggregate firm productivity.
Labor Income and Household Debt: Evidence from Ecuador (with Poorya Kabir)
We investigate the response of household leverage to income shocks. Standard economic theory predicts households borrow to smooth consumption over a temporary income shock; empirical evidence is mixed. We use the exposure of a household’s employer to a credit supply shock, driven by a bank repression episode in Ecuador, as a plausibly exogenous shock to household income. In 2014, the Ecuadorian government required domestic banks to repatriate foreign assets and invest in government debt at interest rates significantly below those at which the government could have borrowed on international markets. We first document large heterogeneity across banks in how affected they were by the shock, driven by the amount of assets they had previously held in international banks. Next, we construct a new and comprehensive data set that combines information on commercial and consumer loans, employer-employee tax records, and census data in Ecuador over the period 2009 to 2018. The granularity of our data allows us to trace out impacts at the employee level by comparing individuals employed at firms borrowing from differentially affected banks using a differences-in-differences design. This study contributes new evidence on how individuals use their balance sheet to self-insure against labor income shocks in an emerging-markets setting.
Threat of falling high status and tax evasion: Evidence from Ecuador (with Yugin Jeong and Jordan Siegel)
Individuals are more likely to engage in unethical behavior when they perceive that it increases the probability of avoiding a potential loss than if they believe that that same behavior increases the probability of a potential gain. Can governments take this empirical regularity into account when allocating scarce tax enforcement resources? We map the Ecuadorian elite social structure, which is centered around a few families like it is in many emerging market economies. Using a sample of random audits by the Ecuadorian Tax Authority, we test whether poor firm performance relative to the firms of owners’ social peers predicts an increase in tax evasion compared to firms that perform poorly in line with peers. We also ask whether firms are using tax evasion as a substitute for financing, i.e., as an implicit government loan, when social ties weaken. This study advances our understanding of the effect of social status and its dynamics on firm behavior.