ABSTRACT. Approximately 135,000 Brazilians die from smoking-related diseases every year (Eriksen et al.,2015). Due to the relevance of smoking as a public health problem, some states and municipalities implemented restrictive smoke-free environments from 2008 onwards. The national regulation became effective only in 2014. In this paper we explore the regional differences in the adoption of this policy to evaluate its impacts on cigarette consumption in Brazil. We propose a differences-in-differences approach to estimate the effects throughout the years on the smoking behavior of two age groups: adults and youths. We also estimate the impacts according to different levels of enforcement among states. We have built our panel using micro data from the National Health Research, collected by the Brazilian Bureau of Statistics (IBGE) in 2013. Our results indicate that smoking bans reduced daily cigarette consumption among young individuals in 7% on average, by reducing smoking initiation. This impact increases to 10% when considering higher enforced smoking bans and vanishes when the law is not enforced. It represents 80 thousand less smokers among youths in the treated capitals. No impacts were found on smoking prevalence and initiation among adults, but high enforced smoking bans are related to 1.2% smoking cessation rate. Robustness tests show that the results are consistent. Our paper contributes to the literature providing evidences to developing countries on the effects of tobacco control policies.

ABSTRACT. Air pollution causes negative externalities on human health, especially on vulnerable groups such as children. We look at hospitalizations due to respiratory diseases for children in São Paulo Metropolitan Area (SPMA) between 2015 and 2017 as consequence of variation in particulate matter (PM 10) levels. We use wind speed as instrument for PM to deal with the endogeneity of air pollution exposure, considering that non-stationary sources of pollution are predominant in the region, which is among the ten largest metropolitan areas in the world. The results show that air pollution positively affects hospitalizations due to all respiratory, pneumonia and asthma in the short term for children between one and five years old. For infant, we only find impact on influenza admission. Additional results suggest the Brazilian public health system is absorbing the increase in hospitalization due to this health shock. Furthermore, delay to visit the hospital may be underestimating our results. We also run a multi-pollutant model, including ozone (O3) as pollutant and solar radiation as instrument. Our coefficients of PM 10 are robust to this specification and we find no impact of O3 on health.

ABSTRACT. We examine the effects of the PBE program, which made the adoption of energy labels mandatory in the Brazilian refrigerator market. Using data from a nationally representative sample of households, we estimate a structural model of appliance choice. We document a modest increase in the valuation of energy costs and reject the equality of pre- and post-PBE valuation distributions as well as the null of their correct valuation, a finding consistent with the existence of an energy efficiency gap. Our policy simulation documents little product switching and substantial heterogeneity in responses. In particular, households aware of the energy label attain improved energy consumption and energy efficiency, the twin targets of PBE. Cognition and attentiveness enhance the effect of the program among informed households, whereas high-income, uninformed, households switch to refrigerators which are more efficient yet more energy-consuming. Increasing information about the label as well as its design are among the ways one can improve the outcomes of the program.

ABSTRACT. This paper provides empirical evidence for the impact of changes in volatility of monetary policy in Brazil using a SVAR where the time-varying volatility of shocks directly affects the level of observed variables. Contrary to the literature, an increase in monetary policy volatility results in higher inflation, combined with reduction in output. The qualitative differences of impulse responses functions, compared to the literature for developed economies, are explained using a calibrated small-scale DSGE model with habit persistence in consumption and stochastic volatility shocks in the Taylor rule. The DSGE model is capable of explaining the increase of inflation in the medium term after a monetary policy volatility shock.

ABSTRACT. This paper studies the macroeconomic implications of simple debt-dependent fiscal rules in emerging countries subject to endogenous financial crises due to a collateral constraint with pecuniary externality. The analysis suggests that debt rules that account for the effects of debt accumulation on asset prices can both reduce the likelihood of financial crises and impact welfare whenever a crisis is likely to happen. Fiscal consolidations based on ad-hoc debt growth targets may be counterproductive in good times and have significant negative effects on welfare during crisis episodes. Simulations suggest that, when carefully designed, fiscal rules based on debt targets improve welfare.

ABSTRACT. We propose a framework of regime-switching DSGEs to study financial crises and macroprudential policies dubbed RegGE-EPro, Regime- Switching General Equilibrium with Endogenous Probabilities ("Reggae"). Financial crises are modeled as regime switches and the transition probabilities governing the switch are endogenous (state-contingent). This captures the idea that the probability of financial crises depends on the state of the economy whereas its timing cannot be forecasted. It unifies DSGE modeling with early warning (crises signaling) methods. Because financial markets do not internalize crises risks ("volatility paradox"), the proposed solution for Reggae assumes that agents depart from full rational-expectations by believing regime switches are deterministic. We illustrate the framework with a simple toy model, a new-keynesian DSGE augmented with credit and macroprudential policy. The illustration showcases the functionalities of the framework for studying financial crises, including for identifying the welfare ranking of different policy rules and the optimal coordination between monetary and macroprudential policies.

ABSTRACT. The present work uses social network analysis to estimate peer effects on students' aspiration towards pursuing a college degree. Using data from Brazilian students, I first show that the willingness to pursue a college degree at the last grade of elementary school predicts both students' current effort and their futures outcomes in school - their probability of dropping out of school or being retained at any grade in high school. Second, I address correlated effects and the reflection problem in order to measure causal peer effects on aspiration. For that, I use friends of friends' characteristics as instruments for friends' aspiration. Results show a positive and significant impact of peers' aspirations on students' own aspirations. When investigating the mechanisms behind this impact, I  show that there is no diffusion of information about students' perceived college returns and that conformity to social norms is likely to be in place.

ABSTRACT. In this paper we investigate how students' response to an incentive program varies according to their level of information about the program and to their socioemotional characteristics. We do so by examining the effect of an incentive experiment implemented in Elementary Public Schools in the municipality of Santo Andre, Brazil. The experiment provided both collective and individual non-financial incentives to students based on their/their classroom's performance on numerical and verbal cognitive tests. We found evidence that incentivized students perform worst than students in the control group, consistent with the theoretical prediction that incentives could undermine student's intrinsic motivation. However, by exploiting the fact that students were exposed to the incentive scheme in two different points in time, we find evidence that students that saw more colleges being rewarded in the first intervention respond more to incentives in the second time. We interpret this result as an evidence that students with higher level of knowledge about the incentive scheme are more likely to respond to it. Finally, we exploit variation from students non-cognitive skills and provide evidence that students more open to experiences respond more to the incentive scheme.

ABSTRACT. This paper analyzes the impact of technology use in the classroom on student math learning in Brazil. The technology is a software tool (electronic game) that can be installed on a tablet, smartphone or personal computer that was specifically designed to the student learn and practice the four basic math operations: addition, subtraction, multiplication, and division. Using a cluster-randomized experiment, we find that students who used the software increased their score in a math test by 0.47 of a standard deviation more than the control group in the short term, and by 0.14 s.d. in the medium term. Additional results show that the improvement in the short term occurred for the entire student distribution and that in the midterm students from the left-tail of the student performance were the ones who benefited the most. We also present evidence that only the treated students who were in a competitive environment during the intervention were able to perform better than the control group one year after the treatment has ended.

ABSTRACT. Este artigo tem como foco empírico a previsão em tempo real para a taxa de inflação, Produto Interno Bruto (PIB) e taxa de juros no Brasil utilizando o modelo VAR com coeficientes variando no tempo e volatilidade estocástica. Os resultados indicam uma performance geral satisfatória do modelo comparativamente a modelos com parâmetros constantes e modelos com versões restritas de variação dos parâmetros. Particularmente, os resultados indicam que incorporar a volatilidade estocástica melhora a previsão do modelo, tornando as diferenças menos expressivas em relação ao modelo TVP-VAR(SV). Os resultados indicam que incorporar a mudança estrutural e, especialmente, a possibilidade de choques exógenos ao modelo, melhora os resultados para a previsão em tempo real para a economia brasileira no período recente.

ABSTRACT. A inflação implícita é a diferença entre as taxas de juros nominal e real. Esta medida pode ser obtida a partir de títulos públicos prefixados e indexados à inflação. No entanto, tal tarefa apresenta duas dificuldades: a) os títulos indexados à inflação possuem defasagem de indexação; b) a sazonalidade da inflação implica sazonalidade da taxa real. Araujo e Vicente (2017) mostram como resolver esses problemas com a ajuda do mercado futuro. O objetivo deste trabalho é propor uma metodologia alternativa para estimação da inflação implícita a partir somente de títulos públicos que trate desses dois problemas. Assumindo que o prêmio de risco da inflação é pequeno no curto prazo, avaliamos a capacidade de previsão da inflação implícita confrontando-a com expectativas baseadas em pesquisas, com a inflação implícita negociada no mercado futuro e com a inflação implícita por títulos públicos que não trata dos dois problemas descritos acima (implícita naive). Os resultados mostram que a inflação implícita é competitiva em relação às previsões de analistas de mercado divulgadas no Boletim Focus e em relação a implícita via mercado futuro e que ela é bem superior à implícita naive no curto prazo. Uma vantagem da inflação implícita é o fato de ela permitir um acompanhamento mais estreito das expectativas do que as pesquisas, uma vez que ela é atualizada continuamente.

ABSTRACT. How much were the average real equity and short-term interest returns in Brazil since 1968? How much is the equity premium in Brazil and how does it compare with the U.S.? We show that the annual averages real returns and the equity premium have been higher in Brazil than in the U.S. However, the much higher Brazilian equity volatility discourages heavier investments in stocks. MODELS. A by-product of our calculations is information about the history of the Brazilian financial market since the creation of the Ibovespa (the main Brazilian Stock Market Index).

ABSTRACT. In this paper we introduce a new machine learning (ML) model for nonlinear regression called Boosting Smooth Transition Regression Tree (BooST). The main advantage of the BooST is that it estimates the derivatives (partial effects) of very general nonlinear models, providing more interpretation than other tree based models concerning the mapping between the covariates and the dependent variable. We provide some asymptotic theory that shows consistency of the partial derivatives and we present some examples on simulated and empirical data.

ABSTRACT. We formulate a continuous-time price discovery model and investigate how the standard price discovery measures vary with respect to the sampling frequency. We find that the component share measure is invariant to the sampling frequency, and hence a continuous-time price discovery measure can be identified from discrete sampled prices. We also contribute by proposing a novel estimation strategy for the continuous-time component share. We establish consistency and asymptotic normality of a kernel-based estimator that compares favourably to the standard daily VECM regression. Finally, we compute daily estimates of price discovery for 30 stocks in the U.S. from 2007 to 2013.

ABSTRACT. This paper studies robust procurement policies under adverse selection and moral hazard. We present a model set up in Laffont and Tirole's (1986) canonical procurement environment, in which the procuring authority conditions payments on observed costs realized by the firm, and trades off productive efficiency with information rent in the optimal contract. However, the principal is assumed to be non-Bayesian, in the sense she cannot form a prior belief over the distribution of private information, but only knows the first two moments of the distribution. We call the solution to the related maxmin problem a ``robust procurement contract'', and characterize it analytically for a special case formulation. This framework is deployed to motivate a mean-variance structure for the principal's value function, which can be interpreted as a robust project selection criterion, i.e., how the principal chooses between production technologies under moment conditions and asymmetric information.

ABSTRACT. This paper analyzes a short-run competitive equilibrium in electricity markets with thermal, hydro and intermittent power sources. Thermal generators have positive and increasing costs and use a marketable and non-renewable input. Hydrogenerators use a free and uncertain input, but storeable. Intermittent renewable (solar or wind) generators use a free, uncertain and non-storeable input. We develop a simple framework and use it to characterize the market equilibrium. We prove that the competitive equilibrium is Pareto optimal for any given technology mix. We then extend the model to incorporate intermittent sources. Specifically, we simulated an economy where hydro capacity is replaced by renewable capacity. The numerical results show that when we increase the share of intermittent sources (i) the profitability of the thermal and hydro plants follow a U-shape relationship, (ii) the mean price drops due to technology diversification and then rise for high shares of intermittent sources, (iii) the profitability of the renewable sources drops steadily. These findings corroborate the idea of a balanced technology mix in long term equilibrium and that of a "clean energy curse".

ABSTRACT. We show that investors who are inattentive to a very simple and well-known tax-exemption opportunity in the Brazilian stock market also display worse out-of-sample stock-picking performance and stronger behavioral biases. The results hold even among experienced investors who trade large volumes. The results are consistent with lab experiments which show that less reflective investors display stronger biases and poorer performance.

ABSTRACT. We propose a novel way to assess information processing in a complex environment of market fragmentation. We take a different angle from the price discovery literature, and investigate information processing in the stochastic process driving stock's volatility (volatility discovery). We show that our volatility discovery framework successfully identifies the leading market in the volatility process, whereas price discovery measures are unable to capture the dynamics of the market-specific volatilities. We compute volatility discovery for 30 stocks and find significant differences in how exchanges impound information into the efficient volatility, as ARCA and NYSE are more important than NASDAQ. Interestingly, price discovery measures suggest different results for nearly half the sample.

ABSTRACT. This paper examines the labor reallocation across agriculture, manufacturing and services, and their impacts on aggregate labor productivity in Brazil from 1950 to 2010. We use a multisector model that features nonhomothetic preferences and constant elasticity of substitution to decompose the drivers of the labor reallocation. For the entire 1950-2010 period, the income effect accounts for most of the reallocation of labor away from agriculture towards manufacturing and services. On the other hand, if we focus only in the 1980-2010 sub-period, the relative price change is now the main driver of the reallocation of labor, but income effect is still quantitatively important. In addition, we explore two important aspects of the Brazilian economy: the fast growth of manufacturing productivity from 1950 to 1980 and its subsequent sluggish behavior, and the decline of services labor productivity after 1980. We find that the fast growth of manufacturing productivity between 1950 and 1980 is responsible for 15% of the aggregate productivity level in 1980. We also find that if services labor productivity had stayed constant at its 1980 level, aggregate labor productivity in 2010 would be 28% higher than observed.

ABSTRACT. Este trabalho utiliza um modelo TVP-VAR para analisar as mudanças ocorridas na dinâmica da inflação entre os mandatos de Henrique Meirelles e Alexandre Tombini à frente do Banco Central. As evidências apontadas por este modelo indicaram um aumento da persistência do hiato da inflação no período de Tombini, em comparação com o de Meirelles. Houve indícios, também, de aumento da tendência e da volatilidade da inflação na gestão de Tombini. Para investigar as causas relacionadas a estas mudanças na dinâmica inflacionária, estimou-se um mesmo modelo DSGE para cada uma das gestões. As análises derivadas deste modelo se mostraram em linha com as do TVP-VAR, tal que apresentaram, também, aumento da persistência do hiato, bem como da volatilidade e da tendência da inflação entre os dois mandatos. A partir de uma análise contrafactual conduzida com o modelo DSGE, verificou-se que a política monetária na gestão de Tombini se mostrou como o fator preponderante para explicar o aumento da volatilidade e da persistência do hiato da inflação.

ABSTRACT. This paper documents new evidence on price-setting behavior using Brazilian microdata from 1989 to 2007. The dataset is extensive in time (222 months), inflation variability (from hyperinflation to monthly deflation), and basket of goods and services (8,353 brands). We find evidence of marked differences between frequency and size of price changes immediately after Plano Real. The plan put an end to hyperinflation and substantially altered price-setting behavior in Brazil. The empirical findings can be summarized as: (i) during hyperinflation, an average of 80.9% of all prices change every month (vs 38.4% after Plano Real); (ii) prices increases are more frequent during hyperinflation, although a small share of prices drops every month (mostly food); (iii) once inflation reaches lower levels, price decreases are almost as likely as price increases; (iv) price changes are less frequent on Services items, but this only stands out during lower inflation; (v) the mean frequency of price changes is close to 80% for all sectors of goods and services during hyperinflation; (vi) the size of nonzero price changes during hyperinflation is 36%, while monthly inflation averages 25.3%; (vii) the size of price decreases is similar during hyper and low inflation (8.3% vs 9.8%); (viii) the intensive margin (size) explains most of the inflation variation from 1989-1993; (iv) while the extensive margin (frequency) gains importance during the 1995-2007 period.

ABSTRACT. The classical approach to preference theory assumes that agents have preferences that are transitive and complete. There are many contexts, however, when one or both of these assumptions may be too stringent. In this paper we explore a characterization of a choice correspondence based on nontrasitive complete preference relations, which we call the Generalized King Chicken Representation. We find that our results are closely related to representations proposed by the tournament literature, which we figure are often special cases of the representation we propose.

ABSTRACT. This paper develops a representation of preference over menus in which the decision-maker chooses within consideration sets. Endowed with a strict preference over the alternatives, he chooses a menu today knowing that tomorrow he will pick the best alternative in the menu that belongs to a randomly draw consideration set. The uncertainty about the relevance of each consideration set in the act of choice leads to a preference for fexibility. We work with both ordinal and additive representations and show that additivity imposes additional restrictions only when the consideration sets are pairwise disjoint. The present representation nests the standard rational representation but it is a particular case of the representation of Kreps (1979).

ABSTRACT. A law abiding citizen is a fully rational agent who has a strong disutiliy for breaking the law. The standard consumer is a special case of a law abiding citizen. However, the decisions of a law abiding citizen can be more complex than those of a consumer. This follows because the law may not be ordered or rationalizable. Our results demarcate the structure of the law that permits a characterization of rational choice under the law with a counterpart of the Weak Axiom of Revealed Preferences. Hence, under suitable conditions, law and economics can have compatible axiomatic foundations.

ABSTRACT. In this paper I study the impact of pretrial detention on rearrest rates and crime activity exploring the natural experiment provided by the introduction of detention hearings in the city of Rio de Janeiro, Brazil. The new procedures were designed to fight the overuse of pretrial incarceration in the Brazilian criminal-justice system and the overcrowding of detention centers. Using novel data about in flagrante delicto apprehensions, I compute rearrest rates per police precinct and asses the reform's impact on them applying a difference-in-differences estimation strategy. I further investigate if the introduction of detention hearings affected crime activity by looking at several crime indicators. I document that the policy change had no significant impact on any of these outcomes, implying that shorter and less certain pretrial custodial sanctions did not significantly affect the incentives to criminality.

ABSTRACT. Many theoretical models have emphasized the importance of legal capacity for reducing crimes. Using a novel database on judges’ and courts’ productivity, this paper investigates whether improvements in judicial system performance impacts violent crimes. In order to identify causal effects, we exploit discontinuities in judicial district classification determined by a discontinuous function of the number of voters. We use a Regression Discontinuity Design and show that changes in judicial district classification increases judicial performance through the selection of more experienced and productive judges. This change is associated with a decrease in violent crimes due to the raise in arrests with warrant.

ABSTRACT. We estimate the impact of fire-armed police on violent crimes (homicides) and acts of aggression in a quasi-natural experiment with small and medium-sized non-metropolitan municipalities. In 2003, Brazilian legislators approved a law that regulates the use of firearms by the municipal police. We explore the population eligibility criterion as an instrumental variable of police firearm possession. We find robust results that a municipal police force with firearms significantly reduced homicides and acts of aggression between 2002 and 2012. Moreover, our additional investigation on mechanisms of deterrence suggests that this result depends strongly on incapacitation (i.e., fire-armed police make more arrests).