ABSTRACT. This work provides partial identification of the average treatment effect on the treated (ATT) in a "difference in differences" (DID) setting when there is sample selection. Adapting the strategy of Lee (2009), under a monotone response of treatment on selection, we derive sharp bounds on the ATT for the sub-population that would be sampled irrespective of the treatment effect on selection. Our partial identification results are based on parallel trends modified to hold conditional on treatment invariant selected sub-populations. Our approach is of a semi-parametric nature as we identify selection probabilities and trimming proportions by imposing a parametric linear index with a constant latent treatment effect on selection. We discuss when these assumptions are reasonable and what limitations the DID design imposes on our strategy. We show that the direction of the monotone treatment response on selection has greater implications in our framework if the latent treatment effect on selection is not constant. We use data from Bastian (2020) to provide an empirical application for our bounds estimators, assessing labor supply intensive margin effects of the 1975 EITC introduction and highlighting a real case where our assumptions are plausible.

ABSTRACT. We build a model for simultaneously now-casting economic conditions in the euro area and its three largest member countries---Germany, France, and Italy. The model formalizes how market participants and policymakers monitor the euro area by incorporating all market moving indicators in real time. We find that area-wide and country-specific data provide informative signals to now-cast the economic conditions in the euro area and member countries. The model provides accurate predictions of economic conditions in real time over a period that covers the past three recessions.

ABSTRACT. In this study, we focus on a generalized nonparametric scalar-on-function regression model for heterogeneously distributed and strongly mixing data. We provide almost complete convergence rates for the local linear estimator of the regression function. We show that, under our conditions, the pointwise and uniform convergence rates are the same on a compact set. On the other hand, when the data is dependent, it is proved that the convergence rate can be slower than that of obtained for independent data. A simulation study shows the good performance and finite properties of the functional local linear estimator (FLL) in comparison to the local constant estimator (FLC). In addition, a one step ahead energy consumption forecasting exercise illustrates that the forecasts of the FLL estimator are significantly more accurate than those of the FLC.

ABSTRACT. This paper presents evidence on the existence of a role model effect in the election of black mayors in Brazil. Taking advantage of the fact that, starting in the 2016's Brazilian mayoral elections, the Brazilian Superior Electoral Court demanded the self-declaration of race from all candidates, I use a regression discontinuity design to compare interracial close elections, that is, elections in which, between the first two candidates, one self-declared as being white and the other as black ("preto" or "pardo", according to the Brazilian Institute of Geography and Statistics' classification). I obtain evidence that the election of black candidates in such elections raises the participation of young black students in the National High School Examination (ENEM) in the following year and increases considerably the proportion of black people born in the municipality running for state deputy in the following election. These results do not seem to be driven by administrative changes performed by the mayor, such as larger investments in education or in social programs, nor are they driven by the support given by the mayor to his political group in subsequent elections. Therefore, a plausible mechanism driving the results is a role model effect, i.e., the example given by black people in positions of leadership can be relevant to motivate and inspire others who identify with them, which can potentially be a powerful mechanism to reduce racial inequality.

ABSTRACT. Government policies may impact economic outcomes directly but also indirectly through effects on political equilibria. This paper examines the effects of the PPCDAm – a centralized environmental policy that synced real-time satellite deforestation data with enforcement on the ground – on the behavior and electoral outcomes of a powerful special-interest group operating in the Brazilian Amazon: farmers. Exploiting close elections, we document that municipalities governed by farmer mayors had higher deforestation rates and CO2e emissions, earmarked more resources to agriculture, and experienced more land-related conflict before, but not after, the PPCDAm was implemented. Any electoral advantage these mayors had before the policy also disappears with the introduction of the PPCDAm. Our findings are consistent with a political agency model where candidates use their occupation to signal commitment to deforestation.

ABSTRACT. The rise of computerization and fixed and mobile broadband has been an important factor for changes in the labor market of rich countries. However, as the world of tech rapidly evolves and even faster mobile internet emerges, there is still little evidence about the impact of current mobile technologies in developing countries. Brazil, in particular, during the last decade, has seen a remarkable increase in mobile internet access. In this paper, we explore this rapid expansion of 4G coverage to estimate the causal effects of fast mobile internet in formal local labor markets in Brazil. Since 4G technologies are not rolled out randomly, provision commitments in contracts between 4G national operators and Brazilian regulatory agency of telecommunications (ANATEL) work as an instrumental variable to internet presence. The average increase of mobile operators observed in sample is responsible for a reduction of 4% in formal employment contracts (about 20 % of the variation in period). This reduction is explained by (i) reduced turnover and by (ii) displacement effects. Consistent with the routinization hypothesis, 4G expansion shifts labor demand away from manual and routine tasks. This, in turn, has benefited more high-skilled workers, increasing the share of college graduates in the labor force and the average skill level index of local labor markets. Effects are driven by the extremes of the distribution of technological intensity but there are no changes in the industry-composition of labor markets. We also show suggestive evidence that mobile broadband is associated with job polarization.

ABSTRACT. Service inputs are a key component of the costs of exporting, and contribute to explain the process of internationalization of firms. A new dataset on the participation of French firms in global value chains reveals that firms with longer export experience in a market are more likely to source service inputs from there. We rationalize this fact in a model where firms are initially uncertain about how successful they are as exporters, but learn their export profitability as they keep selling abroad. Because offshoring requires larger sunk costs than domestic sourcing, some firms decide to offshore only when they become sufficiently confident about their export prospects, i.e., once they acquire enough export experience. More export experience in a foreign destination also induces firms to offshore within the boundaries of the firm rather than at arm’s length. The model further implies that firms are more likely to offshore when frictions in the provision of services between the domestic and the foreign market are greater. In turn, offshoring firms sell greater volumes, display less volatility, and are less likely to exit foreign markets. Exploiting our novel dataset, we provide strong empirical support for each of these predictions.

ABSTRACT. This paper investigates the economic consequences of the 2011 Rio de Janeiro landslides, considered the worst weather-related natural disaster that occurred in Brazil. We evaluated the effect of the disaster on the firm performance and investigated the role of the main recovery policy designed for exposed firms: the subsidized credit provided by the Brazilian Development Bank (BNDES). To define the affected establishments, we use georeferenced data of landslide points and calculate a simulated coverage radius of each specific point based on a prediction model of maximum run-out distance. Using a Difference-in-Differences approach, we show that firms exposed to the disaster reacted to the unexpected shock by reducing the number of employees rather than closing their business or cutting the wages. Also, the firms affected by the shock increased their probability of obtaining subsidized credit. This increase in subsidized credit was partially effective in mitigating the negative effects of the shock since it prevented the exposed companies from closing but did not revert the job losses.

ABSTRACT. The Brazilian “Organizações Sociais de Saúde” model of healthcare provision consists in transferring the administration of public hospitals to not-for-profit organizations, under a pay-for-performance scheme. In this paper we test whether this blend of contracting-out with pay-for-performance increases hospital productivity, and whether this comes at the expense of treatment quality. Synthetic control estimates indicate that this change in hospital incentives increases productivity without increasing in-hospital mortality rates. Effects are not driven by changes in hospital infrastructure or patient characteristics, and output targets seem to explain part of the results. Changes in physician incentives may also play a role. The results support the idea that some incentive-ownership schemes can overcome the healthcare quantity-quality tradeoff, in developing country settings.

ABSTRACT. Nós desenvolvemos um modelo multissetorial de equilíbrio geral aplicado para a análise de reformas tributárias no Brasil. Além de trabalho e capital, as firmas representativas de cada setor utilizam os produtos dos demais setores da economia como insumos na produção. O modelo é calibrado para replicar a matriz de contabilidade social e financeira do Brasil em 2017. A reforma analisada visa equalizar as alíquotas implícitas dos impostos sobre a produção PIS e Cofins entre os 67 setores da economia e migrar completamente a base tributária para valor adicionado. As novas alíquotas são fixadas de forma a manter a arrecadação tributária constante. Partindo de um estado estacionário com tributação exclusivamente sobre a receita, a reforma gera um aumento de 2,4% no consumo privado e 1,4 no PIB real no longo prazo. A maior parte desse ganho advém da equalização da alíquota, que sozinha gera um aumento no consumo privado de 1,5.

ABSTRACT. Existing distortions not only create a wedge between the marginal productivity of inputs and their rental price, but they also impact the cost of entry into entrepreneurship. We investigate misallocation in a multisector general equilibrium model with input-output linkages, in which returns to scale are decreasing and occupational choice is endogenous. As in standard models without endogenous occupational choice, such distortions misallocate labor and inputs among sectors at the intensive margin, but in our environment they also affect aggregate productivity through two additional wedges: a "labor-entrepreneurship" wedge, which misallocates agents between entrepreneurship and the labor force; and a "between-sector" wedge, which misallocates entrepreneurs among the different sectors of economic activity. We show analytically and numerically the importance of the production network and the endogenous occupational choice in amplifying the aggregate effects of distortions.

ABSTRACT. We derive a model in which firms operate in an epidemic environment and internalize infections among their employees in the workplace. The model is calibrated to fit the properties of the Covid-19 epidemic. We show that firms have incentives to fight against infections and can do so very effectively by increasing teleworking and rotating employees between on-site work, teleworking, and leave. The fight against infections in firms flattens the aggregate infections curve. Subsidies to teleworking reduce infections and save lives. Subsidies to sick leave reduce the cost of sick workers and raise workplace infections. Firms delay and weaken the fight against infections during economic downturns. We also consider a planner that values output and lives. We show that the planner prefers to severely restrict the epidemic by tolerating short-term output losses if it takes a short time to develop a vaccine.

ABSTRACT. The widespread adoption of automation technologies that substitute for routine labor potentially alters the costs and benefits of investing in non-routine skills provided by post-secondary education. We study how the unprecedented advance in a leading automation technology - industrial robots -affected individuals’ college decisions in the United States. Our research design exploits geographic variation in adoption intensities and the timing of cohort exposure in a difference-in-differences strategy. We find that individuals growing up in robot-exposed markets are more likely to have a Bachelor’s degree and tend to major in subjects where the prevalence of routine-related occupations is lower. We also find evidence that the gains in college attainment come largely from individuals who otherwise would have specialized in routine-intensive occupations. These patterns coincide with a sizeable increase in the college premium induced by the adoption of robots, suggesting that this factor may play an important role in driving our findings. We find limited support for other potential mechanisms, including changes in the supply-side of education or government educational investments. We then show that the improvements in college attainment map onto labor market outcomes: exposed cohorts are more likely to be employed and have higher incomes in adulthood. These findings suggest that the negative effects of robots on employment could be outweighed by the positive effects in the coming decades, as new generations of workers with different skills replace older ones.

ABSTRACT. This paper investigates the long-run effects of historical pathways on the distribution of economic activity. With records on Brazil’s official pathway networks in 1873, we construct the historical pathways using least-cost-path methods and show that regions crossed by these historical networks feature higher development levels in 2010 than their neighbors, measured by population density and nightlight incidence. Importantly, we show that these effects are significant only a long time after this transportation method becomes obsolete. Our results further suggest that stronger state capacity and economic diversification might be important mechanisms.

ABSTRACT. This paper investigates if women are more vulnerable to intimate partner violence (IPV) in the aftermath of earthquakes in Peru. By combining household-level data on IPV and spatial data on all earthquakes that happened between 2000 and 2008, we show that exposure to very strong earthquakes increases the incidence of IPV by 12.7 percentage points. We document that the effect is less pronounced for women employed at the time of the earthquake, for households with less conservative gender norms and those residing in districts with protective institutions. Further evidence suggests that the increase in IPV following earthquakes is induced by a drop in female intra-household economic power and a rise in male psychological distress. Our findings shed light on a relatively neglected aspect of post-disaster settings and highlights the role of protective policies for women following large-scale disasters.