Determinants and Mechanisms of Tariff Evasion in Latin America: Some Empirical Evidence

David A. Ortiz-Escobar

Major Professor: Thomas Stratmann, PhD, Department of Economics

Committee Members: Carlos Ramirez, Alex Tabarrok

Carow Hall, Conference Hall
December 04, 2015, 11:00 AM to 08:00 AM


This dissertation explores some determinants and mechanisms of import-tariff evasion in Latin America. In the first two papers I measure the elasticity of tariff evasion with respect to tariff rates. The proxy for tariff evasion is the annual or monthly statistical gap between reported exports and reported imports. I use different empirical models to assess the importance of factors such as corruption and law enforcement in explaining tariff evasion elasticities. In the third paper I rely on exports and re-exports statistics to test the hypothesis that indirect trade is a mechanisms to evade import-tariffs.

Paper 1. In this paper I measure the sensitivity of tariffs evasion with respect to duty rates in Latin America. I match exports reported by the US and China with imports reported by some Latin American countries during the period 2003-2013. I infer the effects of tariff rates on tariff evasion by comparing this disaggregated statistical gap with duty rates. In most cases this effect is positive and significant, although it varies widely across Latin American countries. I find evidence that misinvoicing of values and quantities is an important mechanism for tariff evasion in these countries. I also find that tariff evasion elasticities are correlated with corruption at the country-year level and, more specifically, with the efficiency of cross-border trade procedures.

Paper 2. In this paper I measure the elasticity of tariffs evasion with respect to duty rates using monthly international trade figures. To proxy for tariff evasion, I calculate disaggregated discrepancies between reported exports and reported imports. I examine the behavior of these statistical gaps in the context of free trade agreements (FTAs). My identification strategy relies on the fact that FTAs are exogenous shocks to bilateral tariff structures. Through difference-in-difference models I compare the statistical gap tendency of the treatment group (products that enjoyed tariff cuts due to FTAs) with the tendency of the control group (products with no tariff cuts due to FTAs). I apply these models to the concrete case of Peru’s FTAs with China and Mexico. I find that, for the treatment group, the introduction of FTAs resulted on an average reduction of misinvoicing of 10% for imports from China and 30% for imports from Mexico.

Paper 3. Important volumes of exports to Latin America enter recipient countries indirectly via an entrepôt such as Panama. In this paper I rely on exports and re-exports statistics to test the hypothesis that indirect trade is a mechanism to evade tariffs. For the concrete case of Colombia I find that, for non-agricultural products, a ten-percentage point increase on an item's duty rate raises its re-export intensity by 8 percentage points. These results are driven by re-exports from Panama and by non-homogeneous products, and are robust to the inclusion of industry effects. As there is no legal tax advantage of shipping these goods through Panama, the strong correlation between tariffs and re-export intensity is evidence of tariff evasion.