ABSTRACT. Using unique data on between-firm payments and bank to firm lending, we investigate whether bank relationships across firms connected through product market ties affect the provision of loans to these firms. We show that the loans provided by a bank to a firm increase when the bank also lends to one of the firm’s trade partners (customers or suppliers). Negative information about the creditworthiness of a firm’s trading partner is acknowledged by common lending banks, which reduce the loan amount, increases the cost, and reduces the duration of loans provided to the firm. These results suggest that lending to firms connected through the supply chain conveys valuable information to banks.

ABSTRACT. Credit portability has been advocated as an important instrument to promote competition in the banking industry. In 2014, the Brazilian Central Bank (BCB) implemented a regulatory norm to facilitate consumers' credit portability. We explore the spatial local banking concentration in Brazil to investigate how this institutional change affected local credit markets. We show robust evidence that credit portability increased the volume of credit and reduced interest rates for types of loans most benefited by the law.

ABSTRACT. Up to this point, the literature on the issuance of convertible bonds has neglected financial institutions. Contrary to firms, banks not only can issue convertible bonds but also, after the subprime crises, contingent convertible (CoCo) bonds emerged as an alternative. Hence, the purpose of this study is threefold: first, we expand the literature on the motivation to issue convertible bonds in the banking sector; second, we introduce a new proxy (Loans-Deposits Flow) to measure the reinvestment in this sector; and third, we analyze the differences in the motivation for issuing CoCo bonds when compared to convertible bonds. Our results show that the theory of sequential financing is not confirmed for CoCo bonds in the banking sector. Additionally, we provide evidence that banks issue CoCo bonds for regulatory purposes (to increase their capital), while convertibles are issued to allow banks to expand their investments and loan portfolios. The results are robust to several specifications including a propensity-score matching and a difference-in-difference analysis.

ABSTRACT. This paper aims at studying the causal effects of deposit insurance on bank risk and competition. The study uses difference-in-difference approach from period 2010 - 2018 and is separated into two parts: (1) cross country analysis and (2) within country analysis. Cross country analysis explores the effect of deposit insurance implementation. Different countries implemented deposit insurance in different point in time, we use matching on banks' country and banking sector's credit rating, bank size, and bank specialization of the year before implementation. We also conduct the difference-in-difference with multiple time periods by Callaway and Sant'Anna (2021). The results show that treated banks take more risk after implementation due to a decrease in Z-Score. Also, treated banks increase their deposit market share when looking at the ratio between individual bank's customer deposit to total customer deposit of all deposit insurance's member banks in its country. We add to this study by analyzing also the variations of deposit insurance schemes, such as premium method, covered products, and coverage limit. We discover that more relaxing premium method, larger product scope, and higher coverage limit lead to higher risk taking due to an increase in Z-Score and to more potential to attract customer deposit due to an increase in bank's deposit share. Within country analysis investigate the effect of deposit insurance maximum coverage limit changes. We analyze this question by conducting within-country analysis under two exercises: 1)the study of the latest year of limit changes and 2) the study of multiple periods of limit changes. Regarding the first exercise, we implement difference-in-difference to explore behavior of 2 groups of banks, investment and non-investment credit rating, after the limit changes by matching on individual bank's credit rating, size, and specialization of the year before limit changes. Towards the second exercise, we also implement difference-in-difference, but by including all periods of limit changes and controlling for bank's lagged credit rating. The results of two exercises are consistent showing that the increase in maximum coverage limit lead to higher risk taking and attraction of customer deposit of bank with non-investment grade before the year of limit changes.

ABSTRACT. With micro panel data from 32 countries including the poorest and the richest in the world we document (i) a negative relationship between the ability to insure consumption against income shocks and economic development and (ii) a negative relationship between the level of transfer progressivity and the stage of economic development. Importantly, our computation of transfer progressivity includes both public and private net transfers across households—e.g. food transfers. Using an overlapping generations model in which agents differ in permanent productivity, face income shocks and accumulate physical and human capital through learning-by-doing (a labor choice), we find that cross-country differences in transfer progressivity go a long way in explaining the larger ability to insure consumption in poor countries than in rich countries. Then, we use our model to assess the role of transfer progressivity in explaining income per capita differences across countries. We find that decreasing progressivity of poor countries to the levels of rich countries increases income per capita of poor countries by 62%. The opposite experiment, using the progressivity of poor countries on rich countries, reduces income per capita of rich countries by 30%. Since a decrease in transfer progressivity increases the incentives to work and accumulate (physical and human capital) while, at the same time, it reduces social insurance and redistribution, a reduction in progressivity is not necessarily welfare improving. For this reason, we also compute the optimal of transfer progressivity for rich and poor countries separately. We find that optimal progressivity is actually similar (for different reasons) across stages of development which implies that the status quo transfer progressivity for poor (rich) countries is too high (low). Reducing the progressivity of poor countries to optimal levels increases the GDP per capita of the poor by 46% and increases their welfare by 14% in consumption equivalent terms.

ABSTRACT. This paper develops a generalization of the Markov-Switching model with evolving regime-specific means and stochastic volatility proposed Eo and Kim (2016). The model allows for R regimes, B breaks, and time-varying transition probabilities, where the latter are modeled as a multinomial logistic function. We apply our flexible methodology to model the Brazilian GDP growth, which features very complicated dynamics over the last four decades. Our results point to the presence of 3 regimes, at least one break in the long-run trend, and substantial time-variation in the volatility process. Moreover, the selected model features time-varying probability driven by domestic (real interest rate and real exchange rate) and international (commodity prices and global uncertainty) factors. Finally, the results indicate a significant reduction in the Brazilian long-run growth trend.

ABSTRACT. We estimate a DSGE model for Brazil that includes both anticipated and unanticipated fiscal shocks. The model contains a relatively detailed public sector, which allows us to investigate the effects of anticipation for a much wider array of fiscal instruments than previously considered in the literature; indeed, besides shocks to government spending and/or capital and labor taxation, we also analyze important budget components such as public investment, employment, and transfers. Instead of fixing in advance the degree of anticipation of fiscal shocks (generally assuming that they are anticipated in several quarters), we estimate it through a selection scheme based on Bayes Factors. We confirm the literature’s result that fiscal shocks are not the main drivers of business cycles. However, we find that anticipated shocks are less relevant in Brazil when compared to other countries, and that the degree of anticipation varies between only one and two quarters, depending on the fiscal instrument.

ABSTRACT. An important stylized fact about precautionary savings in a labor market characterized by formal and informal workers is that the former saves more compared to the latter. In this paper, we build a search and matching model that incorporates informality in an incomplete markets framework to see the role of Unemployment Insurance (UI) in such behavior. Thus, we add in our UI design the possibility of becoming eligible and exhaust the benefit after some specific period. Moreover, based on empirical evidence, we add the possibility of worker to receive the benefit while working in informality. Then, we calibrate the model to be consistent with micro and macro evidence for Brazil.

ABSTRACT. O artigo tem por objetivo identificar os efeitos da incerteza em um modelo com agentes heterogêneos. Construímos um modelo TANK (Two-Agent New Keynesian) para uma economia pequena e aberta e postulamos um processo de volatilidade estocástica na política fiscal e monetária. Percebeu-se que dado um choque de incerteza a dinâmica no mercado de trabalho é essencial para diferenciar as respostas em ambos os tipos de agentes na economia. De um lado, a restrição de crédito impede que as famílias suavizem o consumo, de outro a empresas diminuem as taxas de investimento. Como resultado, isso se traduz em uma queda dos salários, levando as famílias ricardianas a reduzirem sua oferta de trabalho mais do que as famílias hand-to-mouth. Em geral, verificou-se uma retração do consumo, do produto e do investimento. No mais, percebeu-se que dado um choque de incerteza o câmbio deprecia e a balança comercial reagem de forma positiva, impactando o produto agregado.

ABSTRACT. In a labor market characterized by directed search, unemployed workers search for jobs that are flexible with respect to how much effort they require. Assuming separable preferences, we characterize the optimal unemployment contract for both the case in which savings are observed and, if preferences are of the CARA-GHH type, for the case in which they are hidden. It is always optimal for the government to distort downwards effort through positive marginal tax rates on labor earnings. In the case with hidden savings, we show that optimal contracts take a very simple form, thus showing that Shimer and Werning’s (2007) findings for a McCall search model have a counterpart in a directed search environment with intensive margin adjustments.

ABSTRACT. Using rich administrative data from Brazil, I study how worker reallocation in recessions impacts sorting, match quality, and aggregate productivity in heterogeneous sectors. My first finding is that the percentage of new hires who are high-type workers increases in all sectors except farming during recessions. Therefore, I find evidence for a cleansing effect of recessions: new hires are better. The second finding, however, is that the share of employment that accrues to high types decreases during recessions for all sectors as well. Thus, there is also a sullying effect of recessions: employers are worse. I then investigate the evolution of match quality and find that overall sorting improves for all sectors. In conclusion, I also examine how the reallocation of workers and jobs and changes in match quality affect aggregate productivity. A decomposition exercise shows that reallocation during recessions impacts sectors heterogeneously. A counterfactual analysis shows that the services sector, which relies more on labor, sees a higher adverse impact in recessions from the sullying effect, while the cleansing effect prevails in manufacturing.

ABSTRACT. Some criminal groups are able to avoid confrontation and diversify their economic activities while others are constantly contested by rival groups and the state. To understand why, we propose a model to investigate criminal groups' incentives to invest in military capacity and the state's response to arming. Typically it is difficult to map criminal groups' presence and document their economic activities. We build a novel panel dataset to map presence and characterize the economic activities of criminal groups in Rio de Janeiro, where drug factions and militia groups have controlled territories for at least three decades. We provide evidence that both groups are multi-product enterprises that explore a wide range of licit and illicit goods and services. We empirically test the predictions of our model. Our findings suggest that more groups in the territory increase conflict and state repression and reduce economic diversification. In addition, we argue that the ability to collaborate with the state is crucial for groups to exploit more markets. To the best of our knowledge, this is the first study that uses systematic data to show how criminal groups in Rio de Janeiro evolve over time and diversify their economic activities.

ABSTRACT. This paper studies the short- and medium-run impacts of a policy that decreased exposure to violence of children and teenagers in one of the largest and most violent cities in the world. I exploit the staggered introduction of a place-based policy that dislodged drug traffickers out of the slums they ruled and introduced community-oriented policing. The Pacification Police Units Program (UPP) changed an important feature of these neighborhoods by reducing homicides and police killings by more than 50\% in treated areas, while arguably holding other unobservables such as social networks and interactions, and informal institutions constant. First, in the short-run, children attending schools in treated favelas perform 0.1sd better in Math and 0.07sd better in Reading in national 5th and 9th grades standardized exams. I find no evidence that this finding is driven by compositional changes for students or teachers and I present suggestive evidence of a reduction in exposure to violence \emph{at school} and an increase in teachers' aspirations for students. Second, using individual links using administrative data in the medium-run, when children from UPP phase-in communities are in their early 20s, I find that boys who were less than 13 years old when treatment commenced in their favelas of residence are 5\% more likely to be in the formal labor market in 2018 and 46\% less likely to be incarcerated. I find no evidence that human capital accumulation can explain impacts on adult outcomes and conclude instead that these impacts are more likely driven by a decrease in residential exposure to drug traffickers in childhood.

ABSTRACT. This paper evaluates the influence of the maximum allowed floor area ratio (FAR) on the spatial size of cities. We built a novel database on building height restrictions for the 325 largest Brazilian cities and combined it with recent satellite data. Our estimations show that, as predicted by theory, tighter constraints lead to more urban sprawl, and this result is robust to several specifications, including the framework proposed by Cinelli and Hazlett (2020). Using the share of homeowners among high-income households as an instrumental variable for FAR stringency, we find that the decrease of one standard deviation in the maximum allowed FAR increases the spatial area of a city by 12.4% (or 8km2). Additionally, increasing the stringency of maximum FAR generates an annual cost of about US$ 2.36 million per year per average city.

ABSTRACT. The Brazilian Amazon is marked by attempts at infrastructure-driven development. The construction of the Belo Monte dam, the third-largest in the world, brought chaotical and rapid urbanization to surrounding cities. This paper answered whether the Belo Monte dam impacted the level of violent crime in the region after Altamira was ranked as the most violent city in Brazil in 2015. Following a difference-in-difference approach, I explore the timing of the Belo Monte dam construction and the distance from the construction site to identify the causal effect of unplanned urbanization on homicide rate. In two exogenous shocks, the beginning (2011) and the end of the construction (2015), I estimated a significant rise in the homicide rate in closer cities. The results are driven by criminal activity, with drug trafficking being one of the channels behind the rising homicide rate. The homicide victims are mainly the young male population causing a significant loss of human capital. Increased drug trafficking and rising homicide rate even after the construction indicate that the Belo Monte dam may have a long-term effect on the violence level in the region. Violence imposes high social costs and may jeopardize future growth in the Amazon.

ABSTRACT. This paper assesses the effects of adverse economic shocks on children’s mental health. We rely on the Brazilian High Risk Cohort Study for Mental Conditions, which provides an unprecedented array of data on psychopathology, life events, family medical history as well as parental behavior and polygenic scores for mental disorders over a 10 year period. Our empirical strategy exploits parental job loss events over time in a differences-in-differences framework. We document that parental job loss has strong and persistent negative effects on parental income and household assets. We then show that parental job loss significantly worsens children’s mental health and that this result is robust to different specifications, placebo tests and choices of measurement scales. Turning to potential mechanisms, we document significant effects on children’s exposure to abuse and neglect. Yet, these effects dissipate in later follow-ups, as do the effects on mental health outcomes. Effects do not vary with polygenic risk scores for mental disorders, suggesting that negative effects of economic distress on children’s mental health might be triggered by environmental factors to a greater extent.

ABSTRACT. Using a uniquely rich set of data sources spanning more than 3,000 Brazilian municipalities over a horizon of 25 years, we investigate whether and how infant mortality responds to a permanent shock generated by a trade liberalization reform. We exploit variation in import tariff reductions, together with differences in the baseline industry composition across locations, for identification. We estimate a robust decline in infant mortality in areas with greater exposure to the tariff cuts. In our exploration of mechanisms, we find the most support for the hypothesis that worse labor market opportunities make it less costly undertake health-improving behaviors that are time intensive. Consistent with this hypothesis, we observe a significant decline in female employment rates and an increase in the use of basic health services among women of childbearing ages and infants. We also document that the rollout of a community-based intervention that brings basic health services to the home in a flexible fashion lowers the impacts of the trade shock on infant mortality, providing further evidence in favor of the parental time mechanism. The findings of this paper illustrate that trade policy can have important implications for the household production of child health status and returns of public health policies in a relatively high-mortality context.

ABSTRACT. Despite a globally improved access to primary healthcare, many developing and emerging economies still suffer from a geographically highly imbalanced healthcare provision. With this background, this paper exploits the sudden exit of 8,316 Cuban doctors from Brazilian municipalities. The Cuban doctors formed an integral part of Brazil’s large-scale “More Doctors Program” which is aimed at increasing the number and coverage of public primary health care, especially in vulnerable areas. Anticipating a rift with Brazil’s incoming president, Jair Bolsonaro, most Cuban doctors had to exit the program and Brazil in November 2018, thereby leaving many municipalities deficient in primary health care supply. Leveraging this exodus as an exogenous shock, this paper employs a Difference-in-Difference event study technique to show that there are significant, substantive, and persistent decreases in the number of NCD-related primary care consultations. The effect is limited to doctoral duties, there is no discernible effect on duties performed by nurses and other healthcare workers, and to consultations on NCDs and related risk factors. This paper contributes to the literature by highlighting the essential role of continuity in care in successfully fighting the health and economic burdens of NCDs in low- and middle-income countries.

ABSTRACT. This paper investigates the educational impacts of a policy-driven change in health services available to public elementary school students in Brazil. We study a nationwide program designed to induce activities of primary health care professionals at schools -ranging from anthropometric measurement, nutritional and ophthalmological services to coordinated efforts to identify and fight endemic diseases- and to refer children to other professionals of the public health care network. Exploring variation in the timing of participation induced by rules that prioritized some municipalities first, we show that the program had negative impacts in retention and early withdrawal rates. An analysis of potential health mediators points to an important role for the components of the program associated with local endemic disease control. Our results contribute to the literature on the impacts of access to health services on human capital accumulation after early childhood, and suggest that programs that explore schools to target health-related conditions can be effective to improve educational outcomes.

ABSTRACT. Do generations affected by natural disasters during the critical years of adolescence and early adulthood form different preferences towards the environment than generations who are not? Consistent with the theories of social psychology, we show that an environmental shock experienced during the impressionable years (18-25 years old) helps shape positive environmental preferences. Individuals tend not to change beliefs in response to natural disasters experienced in other age ranges. Using information from the General Social Survey and World Values Survey, we exploit yearly natural disasters variation both within the US and across countries to identify these effects.

ABSTRACT. Renewable energies have become central for global sustainable development. Due to the great potential for exploring solar energy in Brazil, the states have adopted tax incentives to push for solar market development within the country. In this paper, we aim to estimate the effect of a state tax incentive on the solar photovoltaic (PV) system installation rate. We exploit a novel Brazilian administrative dataset of a small-scale distributed generation system to construct a monthly panel of municipalities from 2014 to 2019. We then use the policy staggered adoption from April 2015 to June 2018 and the recent developments of the staggered differences-in-differences literature to assess the causal effects of the policy. We find a positive impact of the state tax incentives on PV adoption. The policy has created 14% of the total installation after treatment, which translates into 8 GWh energy savings in five years. These results imply that regional tax design may help governments meet their climate goals.

ABSTRACT. This paper studies how production responses from agricultural commodity booms affect greenhouse gas emissions, the primary cause of climate change. Using novel data and a shift-share strategy, we show that Brazilian localities more exposed to booms exhibit substantially increased deforestation and agricultural fires, leading to higher emissions. The effects are significantly larger in Brazil’s Cerrado than in other biomes. Commodity booms also lead to production responses toward lower emissions, such as higher output per area. Taking into account higher- and lower-emission production responses, high-exposure localities show an increase in net emissions. Moreover, our findings highlight that positive economic shocks may have unintended consequences, as high-exposure localities present lower adherence to an emission-curbing policy.

ABSTRACT. We propose a no-arbitrage framework related to stochastic discount factors (or pricing kernels) that takes seriously the consequences of no-arbitrage in asset pricing. First, we derive a no-arbitrage one-factor model for the logarithm of asset returns, where the single factor is the logarithm of a valid stochastic discount factor, containing all the pervasive elements of (log) asset returns. Second, based on this one-factor model, we derive a consistent estimator of a valid SDF in a panel-data framework, when the number of assets and of time periods increase without bounds. Identification of a valid SDF is based on economic theory -- no-arbitrage, the asset-pricing equation. The asymptotic character of this no-arbitrage SDF estimator is opposed to standard small-sample alternatives where it is hard to interpret empirical results since these often change when different groups of assets are used in estimation. In theory, asymptotic estimates are immune to this problem. Based on a consistent estimator for a valid SDF, we first investigate which type of utility function best fits U.S. data among popular preference specifications in the literature: the constant-relative-risk-aversion (CRRA) coefficient utility function; the external habit utility function; and the Kreps-Porteus specification. Second, using estimation results, we present a no-arbitrage simulation study assessing how close our consistent SDF estimator is to actual SDF for medium-size panel-data samples. Finally, we estimate a multi-effect linear regression model that allows for parameter heterogeneity in the intercept and in its slope that is consistent with our derived one-factor model. Using regression results we assess how well this heterogeneous one-factor model fits the cross-section and time-series data of assets returns.

ABSTRACT. Firms' currency exposure may result in financial distress and trigger macroeconomic instability. Such exposure can be hedged using currency over-the-counter derivatives. We investigate whether and how lending relationships affect the access to these derivatives using novel loan and derivatives microdata. We document that firms are more likely to buy derivatives from one of their lenders than from a non-lending bank. We also find that prices are lower for derivatives provided by the main lender. These results are stronger among small firms. Our findings are consistent with lending relationships mitigating information asymmetries and derivatives reducing a bank's loan portfolio risk.

ABSTRACT. We provide evidence of how technological progress affects greenhouse gas emissions. Using a dynamic difference-in-differences design, we show that producers in Brazilian localities with high suitability for genetically engineered seeds increase crop output by substituting away from higher-emission activities (such as livestock breeding). This increase in crop output is not accompanied by deforestation and fires---two major greenhouse gas emitters. Further, our findings indicate that the agriculture sector emits greenhouse gases more efficiently after the land-use change, generating more output for a given emission level. Our evidence suggests that technological innovations may increase production and simultaneously offset emissions.

ABSTRACT. This paper treats Brazil as a case study to document the causal effects of health spending on infant mortality. By leveraging the variation in health spending prompted by Brazil’s 29th Constitutional Amendment of 2000, we are able to document not only the effects of health spending on infant mortality, but also the links in the chain connecting spending to health outcomes. We show that (a) a constitutional amendment mandating minimum health spending effectively changes spending patterns both for municipalities which were previously below spending floors as well as those that were above spending floors; (b) these increases in health spending translate into greater primary care coverage, higher supply of hospitals and low-skilled professionals, and; (c) that spending increases in low-spending municipalities bring about moderate reductions in infant mortality within 24 hours and due to perinatal conditions, as well as long term reductions in total infant mortality and infant mortality amenable to primary care (among other causes). Our results contribute to the literature on the impacts of health spending by providing one of the first well-identified causal parameters of the relationship between spending and health outcomes

ABSTRACT. This paper investigates the relationship between international trade and asymmetrical labor income risk. Using the case study of Brazil, we inspect how an increase in import penetration following the China shock impacted the distribution of idiosyncratic earnings changes across the country’s local labor markets, depending on the initial sectoral composition of each region. We find that an increase in import penetration leads to a more disperse and negatively skewed distribution and that these effects can partially be explained by an increase in the volatility of hours worked following job and industry transitions. Moreover, the effect on dispersion grows larger as the lags between periods increase, suggesting a rise in the permanent risk. Through the lens of an incomplete market model, an unborn individual would be willing to forgo up to 4.4% of consumption to avoid the riskier labor market. The welfare cost is half if the higher-order risk is ignored.

ABSTRACT. This paper revisits the competition-innovation debate in light of the recent empirical evidence on the effects of increased exposure to China product competition. From the empirical perspective the evidence is mixed. Faced with fiercer competition, firms in European countries innovate more whereas for the US the effects are negative. In theory, two competing forces are in place. On one hand, more intense competition decreases the profit stream by decreasing markups, the standard Schumpeterian effect. On the other hand, competition may increase the firm's incentives to gain a technological lead over its competitor increasing the firm's ability to charge higher markups, the escape competition effect. The extent to which one of the forces dominates will depend on the technological distance between competitors. We argue that changes in the distinct initial level of exposure to foreign competition between Europe and the U.S. can account for part of the responses empirically observed. We build a model of step-by-step innovation carried by incumbents that are subject to an entry shock that replaces the follower with a new, one step ahead competitor. We calibrate the model to the U.S. and Europe, pre and post China's WTO accession. The results suggest a stronger negative effect on innovation for the U.S. relative to Europe.

ABSTRACT. This paper investigates the performance of forecasting aggregate inflation from the aggregation of disaggregated forecasts considering a large number of predictors and machine learning methods. The benchmarks are survey-based forecasts and traditional time series models forecasting the aggregate directly. Considering the Brazilian inflation, we show the benefits of owning both disaggregated data and machine learning techniques in the forecaster's toolbox. Besides exploring the models in isolation, we propose three model combinations that contribute to obtaining more accurate forecasts: (i) average between model-based forecasts and inflation expectation of the Focus survey available when the econometrician computes your forecast, (ii) distinct methods forecasting different disaggregates to compute the aggregation, and (iii) choice over time between all approached models. The model selections of the two last combinations are based on the minor out-of-sample squared error in a previous period.

ABSTRACT. L-moments are linear combinations of order statistics that provide robust alternatives to standard moments. The estimation of parametric density models by matching sample L-moments is known to outperform maximum likelihood estimation in small samples from several distributions. Recently, it has been further shown that, by varying the number of L-moments with sample size and weighting these acoordingly, one is able to construct an estimator that outperforms MLE in small samples, and yet does not underperform asymptotically. Methods to automatically select the number of L-moments have also been developed. Given their good statistical properties and computational simplicity, it is expected that extending L-moment-based approaches to estimation to semi- and nonparametric settings may be able to produce computationally convenient estimators with good statistical properties. In this paper, we undertake this task by extending the L-moment approach to the estimation of semiparametric models of treatment effects in randomised trials. Recently, Athey (2021) introduced semiparametric models as a convenient tool for the analysis of experiments with heavy-tailed data. We show that, in their setting, a ``plug-in'' L-moment estimator produces an efficient estimator without requiring further corrections. For flexible parametrisations of treatment effects, our estimator is also computationally acttrative, as it can be obtained by solving a quadratic program. We also discuss how to perform specification testing and moment selection. As an application, we apply our methods to a randomised experiment conducted by a large car-hailing service in São Paulo, which randomised discounts to some of its active users. The goal of the experiment was to understand whether short-run changes in the prices of car rides starting or ending at train stations could lead to long-run changes in the demand for public transportation due to learning effects. For the largest discounts randomised, we observe large effects on bimodal rides during the weeks the discount was in place and no effects thereafter. As for unimodal (car-only) rides, we observe a negative effect after the discount is over, which persists over more than a month. We introduce a simple learning model that is able to rationalise these results.

ABSTRACT. health externalities price stabilization

ABSTRACT. We investigate the transmission of a credit crunch to the labor market, studying the effects of the 2008 international financial crisis in Brazil, where the crisis was arguably exogenous to banks and non-financial firms. Building upon a unique data set that connects banks to firms and firms to workers, we are able to investigate the transmission chain from banks to workers. We construct exogenous variation in credit supply by exploiting firms’ relationships with banks before the crisis and how severe the banks’ credit supply was affected. Our results show that the credit crunch had an economic and statistically significant effect on firms’ employment level and their wage bill. Employment adjustment occurred though an increase in layoffs rather than through a reduction in hirings or an increase in quits. Firms with working capital loans and loans maturing during the peak of the crisis experienced a more severe credit reduction. Additionally, skilled workers had a lower proba- bility of being fired as a result of their employee’s credit restriction than unskilled workers.

ABSTRACT. Which banks did COVID-19 affect the most in credit markets: traditional or more digitalized? We answer this question by analyzing prices, marginal costs, and market power at the bank branch level using microdata from Brazil. We found no effect on prices. Marginal costs increased at branches in localities most affected by COVID-19, reducing market power. These branches experienced lending reductions and could not adjust costs in the short term. However, branches of more digitalized banks mitigated these effects due to their cost and lending flexibility: they could expand credit to other locations less affected by COVID-19, improving their market power.

ABSTRACT. We use a change in regulation that improved financial institutions’ information about borrowers with good credit repayment history as a natural experiment for investigating the causal effect of reducing informational asymmetries on credit outcomes. Brazilian SCR contains data for all loans with amounts above a certain threshold. The SCR information contains individual information derived from financial institutions’ loans and is shared with all lending institutions. We exploit a reduction in the SCR individual debt threshold that made information about a group of borrowers suddenly available to all lending institutions. We investigate whether this improved positive information benefited households with loan amounts between the new and the previous thresholds. According to our estimation, after the threshold change, the cost of credit for those borrowers decreased by 33.5 p.p., the size of new loans increased while the average maturity was not changed. We also document that the results were driven by smaller banks. Our results indicate that information asymmetry is an important determinant of lending rates. Thus, measures to improve positive information access should increase the efficiency of credit markets, benefiting on-time borrowers.

ABSTRACT. Financially constrained firms that face moral hazard concerns may not have sufficient pledgeable income or liquidity to invest in projects with positive net present value. We hypothesize that internal capital markets can help minimize this effect through within-group lending. We investigate whether and how financially unconstrained firms within corporate groups raise funds from external capital markets and, in turn, use their internal capital markets to alleviate their affiliates' financial constraints. To examine the effects of internal capital markets on investments and other firm outcomes, we use a rich 15-year dataset of all the limited liability companies in Sweden. We contribute to the literature by analyzing how this channel allows corporate group firms to mitigate financial constraints through internalizing external capital markets within a corporate group in the financial and legal setting of a developed country. Internal capital markets give corporate groups significant advantages over firms that do not belong to a group. Our findings suggest that financially constrained group firms use funds from intragroup loans to invest more, increase their cash cushion, and decrease their external debt. We obtain these findings in a developed country with strong creditor rights and deep external capital markets.

ABSTRACT. We document that oil-producing areas of Brazil experienced increases in crime during the period of increased economic growth driven by the offshore oil boom of the 2000s, at the same time as income and local economic activity grew. We argue that this episode challenges the prevalent understanding that the impact of income shocks on crime is driven primarily by the legal status, or the contestability, of the market in question. Offshore oil production, refining, and distribution in Brazil are entirely concentrated in large firms and there is no scope for income contestability in these markets. We show that accelerated urban growth, not accompanied by increases in the provision of public goods, and possibly also expanded demand for illegal goods, are likely contributing factors to the observed increase in crime.

ABSTRACT. We use novel quasi-experimental variation to (1) test whether firm-specific demand shocks impact wages and (2) to disentangle between predictions coming from wage bargaining and firm upward labor supply. We use a unique institutional feature of public procurement auctions in Brazil: the moment in which the auction ends is random. Under this setting, for auctions in which firms are constantly outbidding each other by incremental amounts, winner and runner-up are as good as randomly assigned. In addition, auction duration affects contract value for a given quantity demanded. We use these two sources of exogenous variation to disentangle the effect on wages that comes from changes in firm size (firm upward labor supply) and the part that comes from changes in contract value holding size constant (bargaining).

ABSTRACT. The paper studies the causal effects of a sizable paid maternity leave extension (from four to six months) on the behavior of workers and firms. Our empirical strategy exploits the precise leavetaking dates and the staggered implementation of the policy across firms, using monthly information from matched employer-employee data covering the universe of the Brazilian formal labor market. Our results show that only 40% of eligible women take extended maternity leave and employment eects are modest and temporarily conned during the months around the leave. We observe both women's and firms' strategic responses to leave extensions: some women voluntarily quit their jobs after the extended leave period is over, while others are red when returning to their jobs after a leave. We oer a comprehensive view on the interplay among labor markets, firms' and workers' characteristics driving the take-up rates of paid leave extensions and its impacts. Besides the regressive distributional effects of the policy and evidence of learning from peers for less-educated workers, the findings also indicate that take-up and labor market effects of leave extensions depend on multiple factors, such as: job security, worker's ability, firm's premium, labor market frictions, and workers' substitutability/complementarity within the firms.

ABSTRACT. This paper studies the role of labor courts in determining labor market outcomes in the Brazilian economy. First, by exploring the fact that judges are assigned randomly to cases and using the universe of labor lawsuits filed in the country’s largest labor court from 2008 to 2013, we show that small firms that draw a more pro-worker judge hire less, experience greater financial distress and exhibit lower survival rates. Second, we develop and calibrate a search-matching model in which laid-off workers decide whether to take firms to court or not. The model is then used to conduct counterfactual exercises simulating the changes brought by a large labor reform in 2017 that transferred part of the legal costs from firms to workers if plaintiff’s case is dismissed. Our model replicates well a set of features of the Brazilian labor market. The counterfactual analysis suggests that this cost-shifting policy implied significant positive effects on employment and aggregate output.

ABSTRACT. Although an extensive and influential literature examines the effects of labor-displacing technologies on labor markets, very little is known about whether and how individuals alter their human capital investments in response to these technological innovations. This paper provides detailed empirical evidence on this question by examining the consequences of the unprecedented advance in robotics technology in the United States. Our research design exploits variation in the penetration of robots across locations and in the timing of exposure across birth cohorts in a cross-cohort identification strategy. Our results show that cohorts differentially exposed to robots before or during the typical college-going ages are significantly more likely to complete a Bachelor’s degree and experience an increase (or a smaller decline) in their labor market earnings. Empirical tests suggest that changes in the college premium and opportunity costs are the key mechanisms generating these effects. We then estimate a structural model of human capital investments to evaluate mechanisms and the importance of these effects for the earnings inequality. Mapping this model to the data, we find that the skill premium is the single most important component of our results, accounting for approximately two-thirds of the overall effect. Further simulations from the estimated model indicate that the effect of robots on earnings inequality declines substantially over time as younger worker generations with different educational choices enter the economy. These findings have important implications for the role of skill investments for the adjustment of the economy to technology in models of skill-biased technological change.