ABSTRACT. Climate-change forecasts for Brazil points to sizable and spatially heterogeneous changes in temperature. This is likely to spur migration and spatial changes in crop patterns. We develop a unique estimable model that integrates both migration and agricultural adaptation as responses to climate change. Our model considers that workers' locational choice is impacted by climate change through three channels: a direct amenity-value channel, as well as the indirect agricultural wages and housing prices channels. The impacts from indirect channels are affected by how farmers adapt to the new climate. Our simulations predict up to an 18.25% increase in micro-region migration rates, with barely no effect from crop change. On the other hand, agricultural adaptation has important impacts in employment in agriculture. The South macro-region is predicted to have an increase in agriculture share of employment when farmers change crop patterns, this includes a reduction in agriculture area dedicated to the production of Maize and an increase the area dedicated to Coffee in the region. Our results help to identify populations most vulnerable to climate change and the regions where most gains from agricultural adaptation may come from.

ABSTRACT. We investigate how the racial composition of entrepreneurs affects racial differences in labor market outcomes of workers. We create a novel data set that identifies the race of firm owners in Brazil and link it with matched employer employee data that contains information on the universe of formal employment contracts. We first present evidence of assortative matching between nonwhite workers and nonwhite entrepreneurs. Controlling for a wide set of covariates, we find that the share of nonwhite workers is 18.7% higher in firms with at least one nonwhite owner. We then show that the adjusted racial wage gap is 28%smaller in firms with nonwhite entrepreneurs than in white owned firms. This is partially explained by nonwhite workers accessing better occupations in firms with nonwhite entrepreneurs. Lastly, we leverage mass layoff episodes to show that nonwhite workers are more likely to find jobs in labor markets with higher labor demand from firms with nonwhite entrepreneurs.

ABSTRACT. This paper studies the effects of an agricultural commodity price boom on structural transformation. I construct a shift-share measure of exposure to the commodity shock by combining climate- and soil-induced variation in agricultural production patterns among municipalities in Brazil with fluctuations in international commodity prices between 2000 and 2010. I show that labor was reallocated away from agriculture towards the manufacturing sector in locations more exposed to the commodities boom. Using data from the Population and Agricultural Censuses, I argue that the results are consistent with greater use of capital inputs in agriculture, which worked as substitutes for farm labor, and greater land inequality in more exposed locations, which ultimately displaced workers from the agricultural sector.

ABSTRACT. Citizens have used mass protests in democratic countries in order to signal their preferences oreven to show their general dissatisfaction with the incumbent government. We propose a theory of protests as a Bayesian persuasion mechanism and we ask what are the conditions such that protests can be an efficient tool for accountability. We think about accountability in two ways. First, we see accountability purely as persuasion, as incumbents responding to the demands fromthe street. Secondly, we think about accountability in the sense of citizens reelecting incumbentsthat are responsive to voters demands with higher probability. Our model shows that protests thatdon’t have a clear demand – and so may face a noisy communication channel, are not only less efficient, but they can be ex-ante inefficient as persuasion mechanism. Moreover, our model shows that noisy demands also lead to lower chances of reelection. We then examine the effects of the large street protests that took place in Brazil in 2013 in both voters and federal legislators behavior. Using twitter historical data by municipality, we compute the intensity of protests and measure thenoise of protesters’ demands in each municipality. Consistent with the model, we find that there is heterogeneous effect of protests in terms of allocation of amendments related to protests demands, proposal of bills and presence in plenary sessions. Moreover, on average, protests reduced the probability of reelection of the incumbent. The data also allowed us to see interesting features of voters following the protests, such as decrease in turnout, increase in "protest votes" (null votes),and decrease in incumbents’ vote share.

ABSTRACT. This paper examines the consequences of the homeland system set up by the apartheid government in South Africa. This system forced Black people to live in homelands and as consequence several million individuals were moved to such areas during the 1960s and 1970s, resulting in one of history's largest segregation policy experiments. We examine how and why relocation to the homelands affected human capital attainment. We exploit the staggered timing of homeland establishment in a cross-cohort identification strategy that compares migrants and homeland natives. Our basic Finding is that moving to the homelands during childhood significantly reduces educational attainment. The magnitude of this effect is particularly pronounced for early childhood exposure, consistent with the importance of events and circumstances during critical periods of child development. The policy also had important consequences on individual success in the labor market: exposed cohorts are significantly less likely to work and have lower income in adulthood. Our examination of possible mechanisms suggests an important role for place effects. Moving to the homelands at earlier ages implies greater childhood exposure to poorer neighborhoods and it disproportionally reduces human capital attainment. The findings of this paper illustrate how discriminatory policies against specific ethnic groups can have long-lasting consequences and increase ethnic inequality.

ABSTRACT. We study the relationship between the political power of agrarian elites and the spread of mass schooling in the early 20th century in Brazil. We use a novel dataset on the occupational structure of the voting elites in 1905 and historical censuses to test whether places, where more voters belonged to the agriculture elite, invested less in schooling. We find that municipalities with a higher share of agriculture voters have a lower literacy rate in 1920 and these effects persist until the 1970s. In the long-run, municipalities that had a higher share of agriculture voters have fewer years of schooling and lower income per capita. We provide evidence that the supply of educational inputs is the main mechanism that explains the long-term persistence.

ABSTRACT. This study is to look at competition, productivity, and resource misallocation. By including common features of developing countries, we built an endogenous growth model in an open economy with input wedges and exogenous terms of trade. Our model shows that international trade pressure potentiates the Schumpeterian effect, and resource misallocation inhibits the escaping competition effect. We seek empirical evidence on these effects on firm dynamics in a high-friction country, thus, we use administrative databases from Brazilian industry, and a competitiveness promotion policy as unanticipated exogenous variation on input wedges. In the short run, our empirical results suggest that product market competition increases by 1.1-1.4 p.p, productivity increases by 0.8-4.6%, and capital misallocation decreases by 0.7-2.5% in the treated group than that in the control group. In the medium-run, there is no effect on productivity or capital misallocation, while product market competition increases by 0.05-0.70 p.p, and labor misallocation decreases by 1.0-2.0% in the treated group than that in the control group. Overall, we show that sources of resource misallocation and non-competitiveness barriers might reduce the role of creative destruction in business dynamism.

ABSTRACT. How can we measure the welfare benefit of ongoing stabilization? We develop a methodology to calculate the welfare cost of business cycles taking into account that observed consumption is partially smoothed. We propose a decomposition that disentangles consumption in a mix of laissez-faire (absent policies) and riskless components. With a novel identification strategy, we estimate the span of stabilization power. Our results show that the welfare cost of total fluctuations is 5.81 percent of lifetime consumption, in which 80 percent is smoothed by the status quo, yielding a residual 1.05 percent to be tackled by policy.

ABSTRACT. We develop an endogenous production network economy model coupled with incomplete information, where the level of information and the linkage-formation interact distorting the producers' decision and the aggregate allocations of the economy. Producers get to be more or less informed depending on how edges far he knows of the equilibrium network. In our model, firms use different inputs from whom they buy through a decentralized search producer model, which endogenizes the linkage-formation of the network, and the final ones sell to a representative consumer. Then, under this setting, we establish the existence, uniqueness and efficiency of the network equilibrium for a given level of information. We show in our model that: higher the level of information (i) the more stable is the network; (ii) the lower is the density of the network, and (iii) the higher is the spillover impact of a productivity shock on the aggregate output. We also design an optimal contract to show that the combination of information-enhancing policies and tax-subsidies is able to mimick a Walrasian full information equilibrium. Finally, we use a proprietary dataset that covers a large share of Brazilian financial transactions and simulate our model with it.

ABSTRACT. This paper examines if and how creditor rights reforms affect banking market competition. By decreasing the expected loss given default of creditors, policies aimed at improving creditor protection might also change the banking market structure. We study a Brazilian bankruptcy reform in 2005 that improved the recoverability of secured creditors in bankruptcy proceedings. We find that local banking concentration decreases in Brazilian municipalities that were more affected by the reform. This result is explained by non-leader banks gaining market share over local leader banks due to the reform. These results are robust to controlling for financial constraints variables. Furthermore, consistent with stronger collateral value reducing information asymmetry, more opaque firms benefit the most from the reform. Overall, our results highlight the role of creditor rights reforms in breaking the information monopoly of incumbent banks.

ABSTRACT. This is the first paper to study daily unambiguous and disentangled mutual fund inflows and outflows, and their heterogeneous sensitivity to past fund performance. Using a comprehensive, novel sample of all daily purchases, redemptions and returns from Brazilian investment funds, I first characterize the time-series properties of daily mutual fund flows. Then, I find daily inflows and outflows are subject to strong within-month daily seasonality -- one order of magnitude large than monthly seasonality of net flows. Inflows and outflows are highly predictable across different classes of mutual funds. Evaluating the daily flow-performance sensitivity, I show that inflows are more sensitive to one-year past fund performance than outflows, in consonance with extant asymmetric investor search-cost hypotheses. Finally, at market-wide aggregate level, very large daily high outflows contribute more to negative daily net flows, than very large daily inflows contribute to positive daily net flows.

ABSTRACT. We assessed volatility transmissions that occurred (inter and intra) commodities of metals and energy from October 16, 1998 to October 17, 2018. With a total of 5220 price observations for each commodity, we estimate spillover indexes developed by Diebold and Yilmaz (2012). We measure such transmissions among twelve commodities and computed the parcels of shock's effects in each pairwise of assets. We also compute how much of these volatilities transmissions was absorbed by each asset in these markets. And using complex network statistics metrics, we describe statistically and graphically these transmissions, following a research line originally developed by Allen and Gale (2000). Our results pointed that total connectivity is 53% and, through the network analysis, we realize that the greatest interactions occur involving oil and nickel and gold and silver. We believe that these findings are useful for the literatures of financial networks and metalic and energetic commoditie's market. As well for investors, risk managers, fund managers, policy makers and metals and energy's producers and buyers.

ABSTRACT. We propose a model of collaborative work in pairs. Each potential partner draws an idea from a distribution that depends on their unobserved ability. The partners then choose to combine their ideas, or work separately. These decisions are based on the intrinsic value of their projects, but also on signaling payoffs, which depend on the public's assessment of individual contributions to joint work. In equilibrium, collaboration strategies both justify and are justified by public assessments. When partners are symmetric, equilibria with symmetric collaborative strategies are often fragile, in a sense made precise in the paper. In such cases, asymmetric equilibria exist: upon observing a collaborative outcome, the public ascribes higher credit to one of the partners based on payoff-irrelevant ``identities." Such favored identities do receive a higher payoff relative to their disfavored counterparts conditional on collaborating, but may receive lower overall expected payoff. Finally, we study a policy that sometimes (but not always) clarifies the ordinal ranking of partners' contributions, and find that such disclosures can be Pareto-improving and reduce the scope for discrimination across payoff-irrelevant identities.

ABSTRACT. We investigate the efficiency of dynamic random matching and bilateral bargaining markets with adverse selection. We take a detail-free approach to the bargaining game, assuming only that: (a) each agent's actions are optimal given the equilibrium market conditions and the equilibrium strategy of the opposing agents; (b) bargaining reveals information about the quality of the object to the uninformed buyer, who cannot commit to acquire the object at a price that exceeds her expected reservation price conditional on such information. We characterize the equilibrium that maximizes the realized gains from trade among all decentralized equilibria. We show this equilibrium features full information revelation during bargaining. This equilibrium also realizes strictly more gains from trade than any equilibrium of trade under a price system, which we use as a benchmark for trade in centralized markets. However, as trading frictions vanish, the optimal decentralized market allocation converges to the optimal centralized market allocation.

ABSTRACT. Recently, Ota, Kato, and Hara (2019) proposed to estimate the conditional mode of a response, given a vector of covariates, using a computationally scalable estimator derived from the linear quantile regression model proposed by Koenker and Bassett (1978). Alternatively, we propose to estimate the conditional mode by maximizing the conditional density estimator of Fernandes, Guerre, and Horta (2021). This approach offers at least two benefits: computational efficiency and good asymptotic behavior which, in particular, “bypasses” the curse of dimensionality.

ABSTRACT. We consider the problem of inference in Difference-in-Differences (DID) models when there are few treated units and errors are spatially correlated. We first show that, when there is a single treated unit, existing inference methods designed for settings with few treated and many control units remain asymptotically valid when errors are strongly mixing. However, these methods are invalid with more than one treated unit. We propose asymptotically valid, though generally conservative, inference methods for settings with more than one treated unit. These alternative inference methods are valid even when the relevant distance metric across units is unavailable. Moreover, they may be relevant alternatives even for settings in which the number of treated units is usually considered as large enough to rely on cluster-robust standard errors. We also present an empirical application that highlights some common misunderstandings in the use of randomization inference in DID applications, and illustrates how our results can be used to provide proper inference.