Research

Working Papers:

In this paper, I examine the role of the unemployment insurance system in the United States on labor market dynamics during the Great Recession. I particularly focus on the role of unemployment benefit extension policies and the financing of these policies on employment. In the United States, unemployment benefits are financed by employers through payroll taxes using experience rating methods. Existing experience rating methods are highly non-linear and these non-linearities create unemployment subsidies to some firms. Unemployment subsidies motivate firms to separate more workers, affecting unemployment. Furthermore, each state independently determines its experience rating for collecting unemployment insurance taxes. I construct a novel dynamic measures of unemployment benefits and unemployment subsidies that exploits the systematic differences across states to identify causal effects of unemployment benefits and unemployment subsidies on employment. Empirically, I find large negative effect of subsidized unemployment benefits on aggregate employment. Moreover, using job flows data, I find that increases in unemployment subsidies affect employment negatively through the job separation channel.

In this paper, I study the composition of unemployment based on unemployment insurance (UI) claims and the cost that employers face to provide them. In the United States (U.S.) unemployment benefits are financed through employer UI payroll taxes and these taxes are a function of unemployment benefits, drawn by their previous employed workers. In addition, existing UI taxation methods in the U.S. display non-linearities that result in some firms having zero cost for additional separations; these firms are heavily subsidized when giving UI eligibility to any additional separations, thus affecting the unemployment rate.

I construct a measure that captures the cost of giving UI eligibility in order to study the causal effects of this cost on the composition of unemployment by UI claims. To do so, I exploit systematic differences in state UI legislation and interact them with aggregate unemployment shocks. I provide empirical results to show that firms with higher implicit UI subsidies generate more UI claims, but also fewer separations without UI claims, which is not an obvious result; I rationalize this by proposing a theory of relabeling, where firms relabel a separation as UI eligible when the cost to do so is small. I test the relabeling hypothesis by studying three labor market outcomes that differs by UI claims: unemployment duration, wage changes after re-employment and recall unemployment. The results show evidence that labor market outcomes respond to the firms' implicit UI subsidy in the direction that is predicted by the relabeling hypothesis.

The study of the effects of land tenure security on agricultural investments and agricultural productivity has received great attention in both theory and policy. In a globalized world, a predictable consequence of increasing land tenure security for farmers is improving their ability to sell crops in international markets. This paper presents several results suggesting that increasing land tenure security for small farmers increases their exposure to agricultural exports: farmers with legally titled lands have greater potential to sell their crops internationally and are more likely to cultivate exportable crops than those who are only possessors (without document-certified property rights). As the relationship observed could be influenced by unobserved heterogeneity, the results herein account for possible endogeneity bias by instrumenting land tenure security with data on farmers’ registration in a national Peruvian land-titling and registration program. To isolate the mechanisms tested, the data is constrained on several characteristics potentially correlated with land tenure security. Our findings indicate that even when accounting for these characteristics, land tenure security still significantly increases the probability that a farmer sells yields internationally, as well as the probability that a farmer cultivates an exportable crop. Finally, we use random forest algorithm on satellite images to study the potential effects of land tenure security on the probability to yield exportable crops in the medium term.

In this paper, we assess the effectiveness of the implementation of the Specialized National Justice System (SNEJ) in the quality of care received by victims of gender-based violence provided by the Women’s Emergency Centre (CEM) in Perú. To do that, we use the following measures:  days transpired between the first and last action taken at CEM, the number of actions taken by the victims in the CEM and the average number of days per action taken by the victims. Finally, we seek to evaluate whether this effectiveness is different between the offered service of CEM to national victims and migrant victims in light of a huge wave of Venezuelans migrants received in Peru over the last 5 years. We exploit the differentiated phased implementation of the SNEJ, with a synthetic counterfactual design to conduct a difference-in-differences methodology (where there are no significant pre-existent differences between control and treated units). The results show that the implementation of the SNEJ significantly reduce the average number of days per action taken on CEM by the victims of gender-based violence while increasing the number of actions taken on CEM favourable to migrant victims.