Working Papers

  • Firms' (mis)reporting under a minimum tax: Evidence from Guatemalan corporate tax returns [Job Market Paper]


    This paper studies firm responses under a minimum tax in the context of a developing country. Minimum taxes are attractive to governments because under such regimes evasion incentives are expected to be lower than under profit taxation, hence increasing tax collection. The specific focus of the present analysis is the Guatemalan corporate income tax regime faced by firms registered in Regimen Optativo. This regime features a minimum tax scheme in which firms pay the largest tax liability between a tax on profits and a tax on turnover. Moreover, firms can be exempted from this scheme depending on their reported gross margin, a statistic based on the firms’ cost structure. As a result of this complex framework, firms face a tax liability function where two kinks and a firm-specific notch coexist, creating differentiated misreporting incentives. The paper exploits this variation to identify firm behavior consistent with evasion strategies, using bunching analysis. The empirical evidence suggests strong firm responses to the minimum tax and no loss carryforward kinks, as well as to the notch created by the minimum tax exemption rule, most of which seem in accordance with evasion behavior. Upper-bounds for average reported profits are estimated to be as low as 42% of actual firms’ profits, implying an evasion rate of 58% in the absence of the minimum tax scheme. The results presented are consistent with the view that minimum taxes can be an effective mechanism to lower tax evasion in environments with limited enforcement capabilities. The evidence also suggests that exemption rules can play an important role in reducing the effectiveness of minimum tax schemes.

  • Estimating the fiscal impact of extreme weather events


    This paper constructs a panel dataset of 168 countries with information on natural disasters and macroeconomic variables for the period 2000-2015, with the objective of estimating the fiscal impact of extreme weather events. While the literature analyzing the economic incidence of natural disasters has mainly focused on their macroeconomic consequences, the fiscal dimension of this problem remains relatively neglected. Due to their adverse effect on the economy, extreme weather events tend to reduce government revenues and increase public expenditure, creating a negative pressure on the budget balance. According to the results of the present analysis, the occurrence of at least one extreme weather event is associated with an increase in the budget deficit between 0.4% and 0.9% of GDP. This impact comes primarily from an immediate reduction in government revenues, as a percentage of GDP, with some evidence pointing out to a lagged effect on public expenditure two years after the event. Moreover, the fiscal effect is larger for low-income and lower-middle income countries, but is not significant for high-income and upper-middle income countries.

Work in Progress

  • Tax arbitrage and domestic profit shifting in environments with co-existing income tax regimes


    This paper explores the consequences of having two co-existing corporate income tax regimes within a domestic tax system. This scenario is interesting because, in such environments, a simple theoretical model predicts an optimal strategy involving tax arbitrage through income shifting across regimes. The empirical exercise focuses on the case of Guatemala, where firms choose between a regime that taxes profits –Regimen Optativo–, and a regime that taxes turnover –Regimen General–. The Guatemalan setting is particularly useful to analyze this topic for two reasons. First, firms can change their choice of regime before each fiscal year –something generally not observed in other countries. And, second, a recent income tax reform introduced variation in the marginal tax rates of each regime, creating the scope to analyze the differential behavior predicted by the theoretical model. Following a difference-in-difference approach, where treatment and control groups are defined by whether firms belong to a tax arbitrage network or not, the results show differential behavior between the two groups. After the reform, firms that do not belong to a tax arbitrage network faced a decrease of around one percentage point more than the treatment group in the probability of registering in Regimen General. This effect seems to concentrate on low turnover firms. On aggregate, it is estimated that the tax savings obtained by arbitrage networks could be as large as 66% of their tax liability without profit shifting, and equivalent to about 30% of the overall corporate income tax liability in the country.

  • Bunching into VAT: Theory and evidence from China (with Xuan Wang)