SBE39: 39TH MEETING OF THE BRAZILIAN ECONOMETRIC SOCIETY
PROGRAM FOR WEDNESDAY, DECEMBER 13TH
Days:
previous day
next day
all days

View: session overviewtalk overview

08:30-10:15 Session 5A: Health
Chair:
Bladimir Carrillo (UFV, Brazil)
Location: Mezanino Bossa Nova
08:30
Pedro Américo (PUC-Rio, Brazil)
Rudi Rocha (UFRJ, Brazil)
Prescription Drug Cost-Sharing and Health Outcomes: Evidence from a Developing Country

ABSTRACT. This paper evaluates the health effects of a novel and large-scale cost-sharing system of prescription drugs introduced in Brazil, the Aqui Tem Farmácia Popular program (ATFP). In the ATFP system, the government establishes a reference price for the generic version of each listed medicine. Patients pay for the difference between the reference price and the retail price, generally resulting in substantially lower prices for the patient at the pharmacy counter. We evaluate the effects of ATFP on hospitalization and mortality rates by circulatory conditions and diabetes for individuals aged 40 years or more. We find that ATFP is significantly associated with a reduction in mortality for the most acute circulatory conditions, such as ischemia and cerebrovascular diseases; and with a reduction in hospitalization rates for the most chronic conditions, such as hypertension and diabetes. In order to assess equity and efficiency, we also examine whether ATFP utilization and its effects on health outcomes vary with socioeconomic status, and provide a cost-benefit analysis in terms of averted deaths and hospitalization costs.

09:00
Leticia Nunes (FGV/EPGE, Brazil)
Francisco Costa (FGV/EPGE, Brazil)
Fabio Miessi Sanches (PUC-Rio, Brazil)
Unveiling Physicians' Geographical Misallocation: Evidence from a Developing Country
SPEAKER: Leticia Nunes

ABSTRACT. The maldistribution of health professionals in rural and remote areas constitutes a significant constraint to the delivery of basic healthcare. This paper exploit a comprehensive dataset of all generalist physicians graduated in Brazil between 2001 and 2013 to shed new light on the factors influencing their decision of practice location in a developing context. Our strategy consists of using a discrete choice model with a random parameters specification to estimate physicians' demand for working regions right after graduation. We find that wages, though important, are not key on physicians' choices. In contrast, other local characteristics such as the infrastructure of health facilities and the region's quality of life play a very relevant role. Physicians graduated from lower ranked universities tend to migrate to areas with smaller labor market competition, measured by the density of doctors working in a region. Counterfactual simulations suggest that the construction of medical schools in under-served areas, may considerably improve the geographical distribution of physicians and that affirmative actions also have positive effect.

09:30
Bladimir Carrillo (UFV, Brazil)
Jose Feres (IPEA, Brazil)
More Doctors, Better Health? Evidence from a Physician Distribution Policy

ABSTRACT. In 2013, the Brazilian government implemented one of the largest physician distribution programs on record. Using a difference-in-difference framework, we document that the number of physicians increased by 17 percent in treated areas, with effects that are substantially larger in magnitude for family doctors. This expansion increased doctor visits by 4.3 percent and prenatal care by physicians by 10 percent. Yet despite these improvements in physician supply and utilization of doctors, we find little evidence that the program led to better infant health, measured by low birth weight, prematurity and infant mortality. 

08:30-10:15 Session 5B: Labor I
Chair:
Leonardo Monasterio (UCB e IPEA, Brazil)
Location: Jade
08:30
Stephen O'Connell (Massachusetts Institute of Technology, United States)
Lucas Mation (IPEA, Brazil)
Mark Dutz (World Bank, United States)
Can business input improve the effectiveness of worker training? Evidence from Brazil's Pronatec-MDIC
SPEAKER: Mark Dutz

ABSTRACT. We evaluate the employment effects of a publicly-run national technical vocational education training program in Brazil that explicitly takes input from firms in determining the location, scale, and skill content of courses offered. Using exogenous course capacity restrictions, we find that those completing the course following receipt of a course offer see an 8.6 percent increase in employment over the year following course completion. These effects come from previously unemployed trainees who find employment at non-requesting firms. The demand-driven program's effects are larger and statistically distinguishable from those of a broader and institutionally-similar publicly-administered skills training program run at the same time that did not take input from firms. We find the demand-driven program better aligned skill training with future aggregate occupational employment growth -- suggesting the input from firms captured meaningful information about growth in skill demand. Courses offered in occupations that grew more over the year following requests exhibited larger employment effects, explaining the effectiveness of the demand-driven model.

09:00
Carlos Alberto Belchior (PUC-Rio, Brazil)
Wages and tasks of outsourced workers: the case of call centers in Brazil

ABSTRACT. The development of new work arrangements has been an important new feature of the labor market. In this study, we attempt to estimate the impact of domestic outsourcing on the outsourced workers’ wages. Previous studies have identified negative impacts of outsourcing on wages ranging from 3% to 15%. We argue that these previous studies have potentially biased estimates since they rely on the orthogonality of the outsourcing of activities in relation to the tasks performed by the workers. We use the Brazilian regulation of outsourced hiring, which exogenously restrained certain companies of outsourcing their call centers, as a natural experiment to estimate the outsourced wage gap – controlling for different tasks. Our results point to much higher negative impacts of outsourcing on wages – ranging from 31% to 38% - and are robust to a series of checks.

09:30
Raphael Corbi (FEA/USP, Brazil)
Tiago Ferraz (FEA/USP, Brazil)
Rainfall, Internal Migration and Local Labor Markets in Brazil
SPEAKER: Tiago Ferraz

ABSTRACT. We investigate the labor market impacts of internal migration in Brazil by following a two-step approach. First, we explore the variation of out-migration flows from the semiarid driven by deviations from historical rainfall averages. Second, we distribute the predicted out-migration flow based on the pre-existing support network in each destination based on the migrant's region of origin. Our results indicate that increasing in-migration rate by 1p.p. (one standard deviation) reduces overall native wages in 0.6% mostly in the formal sector and has no effect on overall employment. These results mask considerable heterogeneity. Migration hampers the prospects of low-skilled natives while improving employability of the high-skilled. On the other hand, the negative effect on wages is concentrated on high-skilled formal workers. This can be reconciled with the view that low-skilled migrants are substitutes for low-skilled natives and complements for high-skilled natives. Moreover, it is consistent with a model in which (more regulated) formal labor markets have higher wage rigidity compared to the more competitive informal sector.

10:00
Philipp Ehrl (UCB, Brazil)
Leonardo Monasterio (UCB e IPEA, Brazil)
Inherited cultural diversity and wages in Brazil

ABSTRACT. This paper estimates the long-term impact of immigration to Rio Grande do Sul/Brazil on contemporary wages. Based on a unique micro-data panel that includes the names of workers, we apply machine learning algorithms to classify surnames and infer each workers’ ancestry in order to calculate the inherited cultural diversity in the workforce by municipality. We address the endogeneity of cultural diversity by using three sets of instruments: distance to settlements created by the government for non-Iberian immigrants between 1824-1918, share of street names with foreign surnames and share of foreigners in 1920. Our IV-estimations prove robust to human capital differences, institutions, geography, the spatial sorting of workers based on intrinsic abilities and the diffusion of knowledge through imports. The results clearly indicate that an increase in diversity – exclusively transmitted through the share of workers with non-Iberian ancestry – leads to a positive wage externality.

08:30-10:15 Session 5C: Finance I
Chair:
Eduardo Zilberman (PUC-Rio, Brazil)
Location: Topázio
08:30
Fernando Chague (FGV/EESP, Brazil)
Rodrigo De-Losso (FEA/USP, Brazil)
Bruno Giovannetti (FGV/ESSP, Brazil)
Individual Investors and the Price Tag Illusion

ABSTRACT. Individuals take their experiences as consumers to the stock market. In particular, they believe a stock to be “on Sale” when its price falls, just like when they see a T-shirt with a lower price tag. We call this bias “Price Tag Illusion” (PTI). The PTI explains the well-documented perverse contrarian behavior of individuals. Consistent with the PTI hypothesis, we show that individuals buy a stock when its price falls without analyzing anything else and that they become much more contrarian in the stock market when real Sales are announced in the goods market. We also show that individuals contrarian behavior has three characteristics which are typical to behavioral biases: it leads to losses, it is somewhat stable, and it diminishes with experience.

09:00
Joelson Sampaio (FGV/EESP, Brazil)
Antonio Carvalho (FGV/EAESP, Brazil)
Roberto Pinheiro (Federal Reserve at Cleveland, United States)
Dotcom Price Spiral

ABSTRACT. We show that during the bubble implied growth rates coming from the underpricing of IPO market explains short term returns on the NASDAQ composite index. This result remains even if we replace actual underprice for others different instruments for underpricing that are based on predetermined variables and not correlated to market returns. We also do placebo tests to assess the relation between underpricing and NASDAQ returns over other periods besides the bubble. We show that growth proxies that are not contaminated by the booms and busts of the stock market as a whole are uncorrelated with the returns on the NASDAQ composite index in periods outside the bubble.

09:30
Eduardo Zilberman (PUC-Rio, Brazil)
Carlos Carvalho (BCB e PUC-Rio, Brazil)
Daniel Cordeiro (XP, Brazil)
Ruy Ribeiro (PUC-Rio, Brazil)
Gambling, Risk Appetite and Asset Pricing

ABSTRACT. We show that time variation in gambling activity in the US provides relevant information for asset pricing. Adding a conditioning variable based on gambling activity improves the fit of both CAPM and consumption CAPM. In particular, it performs better than using Lettau and Ludvigson (2001)’s consumption-wealth ratio (cay) as conditioning variable. Moreover, this measure helps forecasting excess returns at longer horizons, providing information that is not already contained in either price-dividend ratio nor consumption-wealth ratio. We also find that our gambling variable has better out-of-sample predictive performance than all other alternative variables.

08:30-10:15 Session 5D: Macroeconomics I
Chair:
Napoleão Silva (IPEA, Brazil)
Location: Ametista
08:30
Wagner Gaglianone (BCB, Brazil)
Raffaella Giacomini (University College London, UK)
João Victor Issler (FGV/EPGE, Brazil)
Vasiliki Skreta (University College London, UK)
Incentive-driven Inattention

ABSTRACT. We build a dynamic rational inattention model where agents have a budget of attention or cognitive resources to devote to updating expectations. Both the amount of attention and the expectations’ accuracy are endogenous and linked to the cost and benefit of updating, which can vary across agents and over time. The model captures novel stylized facts from a panel of professional forecasters–the Central Bank of Brazil’s Focus Survey–that is unique among surveys in that agents choose when to update and there is a recurring contest that ranks agents based on their accuracy. We find that the incentives linked to the contest are the primary drivers of updates and accuracy gains. We structurally estimate the model to uncover the deep parameters characterizing the cost-benefit distribution across agents and how this changes around the contest. The model fits the data well and the estimates allow us to perform counterfactual exercises to understand the value of the contest and to investigate alternative survey designs.

08:55
Marco Bonomo (Insper, Brazil)
Carlos Carvalho (BCB e PUC-Rio, Brazil)
Rodolfo Rigato (PUC-Rio, Brazil)
Aggregate versus Sectoral Shocks: What Drives Price Setting Dynamics?

ABSTRACT. We estimate a structural multi-sector price setting model in order to distinguish the roles of aggregate and sector-specic shocks in price setting dynamics. Previous estimates of the size of these uctuations, as in Bonomo, Carvalho, Garcia and Malta (2016), have relied on models which feature a single economic sector. Identication of the size of aggregate disturbances, however, requires a multi-sector framework, as the omission of sectoral heterogeneity is likely to generate an upward bias in the estimated size of aggregate shocks. The model features price setter rms that are subject to two distinct frictions: besides facing menu costs, rms have only partial information about their optimal prices. Estimation is carried out by Simulated Method of Moments. Three questions are addressed using the results obtained: assessing the relative importance of aggregate and sectoral innovations for business cycle dynamics, quantifying the cross-sectoral dierence of responses of output and in ation to monetary policy shocks and, last, comparing the eects of such shocks in a single sector against a multi-sector economy.

09:20
Bernardo S. C. Ribeiro (PUC-Rio, Brazil)
Structural transformation in Brazil: the 1980s and the importance of an open economy framework

ABSTRACT. During the period 1970-2015 in Brazil, as the income level grew, manufacture share on total value added has tracked a decreasing trend, while services followed the opposite direction. This structural transformation panorama, however, was not stable throughout the entire period. We document how the 1980s were characterized by a faster speed in such process and by a distinction in data movements when constant and current prices are concerned. The paper argues that external unbalances and the country capacity to import could have played a role in the process and, with a two-sector and two-country model, presents qualitative evidence on that. In a static comparative simulation, in which we lower the country capacity to import with a distortion term that raises import prices, the model replicate the movements in the data. With constant price time series, a closer qualitatively picture to 1980s scenario was achieved with a combination of worsening capacity to import and slowdown in manufacture growth rate. Based on these simulations, the paper supports the growing importance placed on open economy models when structural transformation analysis is concerned.

09:45
Napoleão Silva (IPEA, Brazil)
Eduardo Zilberman (PUC-Rio, Brazil)
Restrições financeiras e o PIB per capita no Brasil

ABSTRACT. As restrições financeiras sobre as firmas brasileiras são muito elevadas em relação às economias avançadas. No Brasil, 59% das firmas têm acesso a um empréstimo bancário ou uma linha de crédito. Nos países desenvolvidos o percentual médio é de 95%. Os requerimentos de colaterais para empréstimos são bem maiores no Brasil (95% do valor do empréstimo) quando comparados à média dos países desenvolvidos (50% do valor do empréstimo). O “spread” da taxa de juros no Brasil é muito elevado em relação à média dos países desenvolvidos (12% no Brasil e 3% nos países desenvolvidos). Neste contexto, o objetivo do nosso trabalho é avaliar o impacto de reduções em três diferentes restrições financeiras sobre o PIB per capita no Brasil. Em termos mais específicos, nossa meta principal é avaliar qual das três fricções financeiras (custo de participação no mercado de crédito, limite de endividamento ou custo de monitoramento) é a mais importante para afetar o PIB per capita no Brasil. Para realizar nosso objetivo utilizamos a estrutura teórica desenvolvida em Dabla-Norris e outros (2015). Esta estrutura consiste de uma versão do modelo de crescimento neoclássico com agentes heterogêneos e três fricções financeiras. O modelo é calibrado para a economia brasileira em 2009 e fazemos exercícios de simulação onde avaliamos os impactos de reduções nas fricções financeiras. Neste texto realizamos três exercícios. No primeiro exercício, a redução do custo de participação no mercado de crédito (que eleva o acesso financeiro para o nível dos países desenvolvidos) gera um aumento no PIB per capita de 3,6%. No segundo exercício, a redução do custo de monitoramento (que eleva a eficiência do sistema para o nível dos países desenvolvidos) gera uma elevação no PIB per capita de 1,7%. Por último, no terceiro exercício avaliamos um relaxamento nas restrições de endividamento que seria obtido se os colaterais como proporção dos empréstimos no Brasil fossem iguais à média dos países desenvolvidos. Os resultados mostram que a redução dos colaterais no Brasil elevaria o PIB per capita em 12%. Neste contexto, a restrição financeira com maior impacto sobre o PIB per capita, no caso brasileiro, é o limite de endividamento. Neste sentido, políticas voltadas para a redução das restrições de endividamento terão um impacto maior sobre o PIB do que políticas que visem reduzir o spread ou elevar a participação no mercado de crédito.

08:30-10:15 Session 5E: Minicurso SBE - Aula 1
Location: Água Marinha
08:30
Amanda Arabage (FGV/EESP Clear, Brazil)
Juliana Camargo (FGV/EESP Clear, Brazil)
Introdução à Avaliação Quantitativa de Impacto

ABSTRACT. O minicurso "Introdução à Avaliação Quantitativa de Impacto" tem como objetivo apresentar as finalidades dessa abordagem de avaliação e introduzir os principais conceitos relacionados, como causalidade, contrafactual e definições de grupos de tratamento e controle. Serão abordados o método experimental e os principais métodos não-experimentais, destacando a aplicabilidade, vantagens e limitações de cada um.

10:30-12:15 Session 6A: Sessão Especial Banco Mundial

Productivity and inclusion in Brazil

Chair:
Mark Dutz (World Bank, United States)
Location: Jade
10:30
Antonio Nucifora (World Bank, Brazil)
Rong Qian (World Bank, United States)
Jorge Thompson Araujo (World Bank, United States)
Productivity performance in Brazil
11:05
Pedro Olinto (World Bank, Brazil)
Mariana Vijil (World Bank, United States)
Vivian de Fatima Amorim (World Bank, Brazil)
Mark Dutz (World Bank, United States)
The distributional effects of trade policy in Brazil
SPEAKER: Pedro Olinto
11:40
Eduardo Ribeiro (UFRJ, Brazil)
Steen Byskov (World Bank, United States)
Pietro Calice (World Bank, United States)
Subsidized Capital Goods Credit, Financial Dependency and Misallocation in Manufacturing and Services
10:30-12:15 Session 6B: Sessão Especial Prêmio SBE Finanças
Chair:
Bruno Giovannetti (FGV/ESSP, Brazil)
Discussants:
Fernando Chague (FGV/EESP, Brazil)
Bruno Giovannetti (FGV/ESSP, Brazil)
Ruy Ribeiro (PUC-Rio, Brazil)
Location: Topázio
10:30
Thiago Silva (BCB, Brazil)
Benjamin Tabak (UCB, Brazil)
Michel Alexandre (BCB, Brazil)
Systemic Risk in Financial Systems: a feedback approach

ABSTRACT. We develop an innovative framework to estimate systemic risk that accounts for feedback effects between the real and financial sectors. We model the feedback effects through successive deterioration of borrowers' creditworthiness and illiquidity spreading, thus giving rise to a micro-level financial accelerator between firms and banks. We demonstrate that the model converges to a unique fixed point and the key role that centrality plays in shaping the level of amplification of shocks. We also provide a mathematical framework to explain systemic risk variations in time as a function of the network characteristics of economic agents. Finally, we supply empirical evidence on the economic significance of the feedback effects on comprehensive loan-level data of the Brazilian credit register. Our results corroborate the importance of incorporating new contagion channels besides the traditional interbank market in systemic risk models.

11:15
Kym Ardison (FGV/EPGE, Brazil)
Caio Almeida (FGV/EPGE, Brazil)
René Garcia (University of Montreal, Canada)
Nonparametric Assessment of Hedge Fund Performance
SPEAKER: Kym Ardison

ABSTRACT. This paper proposes a novel class of non-parametric, discount based, performance measures that incorporate non-arbitrage restrictions as well as higher order moments information. We show theoretically the relationship between the proposed class and two well know alternatives: the Jensen's Alpha and the Hansen and Jagannathan estimator. We provide a full set of asymptotic results that allow one to test both for the relevance of the underlying factors in the stochastic discount factor determination as well as the performance measure. Empirically we apply our methodology to a large panel of individual hedge fund returns. Our results reveal sizable dierences across estimators, indicating that performance measurement in incomplete markets might dier across investors.

10:30-12:15 Session 6C: Economics and Politics I
Chair:
Raphael Corbi (FEA/USP, Brazil)
Location: Água Marinha
10:30
Valdemar Pinho Neto (FGV/EPGE, Brazil)
Cecilia Machado (FGV/EPGE, Brazil)
Conditional Cash Transfers and Voting Behavior in Brazil

ABSTRACT. This paper studies the effects of a large Conditional Cash Transfer program, the Brazilian \textit{Bolsa Fam\'ilia} (BF), on voting behavior. Using administrative data on beneficiaries and on where they vote, we explore random variation of the proportion of BF beneficiaries at the polling station level, which generally has fewer than 400 registered voters. The results indicate that the cash transfers positively affect the turnout in the presidential elections and mainly increase the support for the incumbent party that implemented the policy. We find that CCTs can affect voting through two mechanisms: more voters turning out to vote (mobilizing effect) and voters shifting their preferences between candidates (or invalid vote). In addition, the electoral bonus to the political party that implemented the policy is led by those who were still in the program at the moment of the election, relative to those who left the program.

10:55
Alexsandros Fraga (Bocconi University, Brazil)
Raphael Corbi (FEA/USP, Brazil)
Luis Meloni (FGV/EESP, Brazil)
Edited democracy? Evidence from the 1989 Brazilian Presidential Election
SPEAKER: Luis Meloni

ABSTRACT. In this paper we investigate if biased media coverage of electoral campaigns can affect electoral choices by looking at a famous and publicly acknowledged case of biased news coverage - the coverage of Rede Globo, the main television broadcaster in Brazil, of the second-round of the 1989 Brazilian presidential election. In 1989, the first democratic presidential election in Brazil was decided in a run-off between Lula da Silva, a programmatic left-wing candidate, and Fernando Collor, a young political outsider. Two days before the election, when the opinion polls indicated a technical tie, Rede Globo, the most popular television channel in Brazil, aired a montage of the final debate that was harmful to Lula during its prime time newscast. Books and documentaries long speculated that the broadcast of this montage of the final debate played a crucial role in deciding the election. The aim of this paper is to bring this hypothesis to the data and to test if access to Globo's news coverage during the second-round of the 1989 presidential election affected 1989 presidential election's outcomes. Our main finding is that the availability of Globo's coverage had a statistically significant negative effect on Lula's vote-share in the second-round of 1989 election. In our favorite specification, the availability of Globo's signal decreased Lula's vote-share by 2.23 percent points, which is equivalent to 1.47 million votes. This effect was not enough to decide the election but considerably decreased its competitiveness: the wining margin of Collor would be 73.6 percent smaller in the absence of the Globo's coverage.

11:20
Mariana Davi (Banco Cooperativo Sicredi, Brazil)
Marcelo Portugal (UFRGS, Brazil)
Politics and finance: a study on the impact of campaign donations on Brazilian firms
SPEAKER: Mariana Davi

ABSTRACT. This paper aims to identify potential benefits obtained by companies for their contributions to political campaigns. We used an extensive database with information on donations to House, Senate, and Presidency candidates in the 2006 and 2010 elections. The variables of interest analyzed were the cumulative abnormal return by the time the results of each election became known and the return on equity in the year following the elections. Panel regressions were estimated as ordinary least squares (OLS), and fixed effects of year and industry were included. The results indicate that not only does the market anticipate future benefits for companies that contributed to campaigns - which is reflected in positive cumulative abnormal returns at the announcement of the election results - but these companies also have higher returns on equity than those that were not involved in the political process. In addition, donations to winning candidates generate higher returns than donations to losing candidates, which supports the return of favors hypothesis. Similarly, contributions to candidates affiliated with the president’s coalition's also had a higher impact when compared to donations to political opponents.

11:45
Raphael Corbi (FEA/USP, Brazil)
Emotional Voters
SPEAKER: Raphael Corbi

ABSTRACT. This paper analyzes the impact of emotional shocks induced by soccer game outcomes on voting for the status-quo in Brazilian elections. Conditioning on the pregame betting markets implied probabilities of each match outcome, I am able to interpret the estimate of actual soccer results as a causal effect on voting behavior. The main results indicate that a 10 p.p. increase in the share of people in a municipality experiencing negative emotional shocks in the days preceding an election decreases voting for the incumbent by 1.1 p.p. In contrast, positive emotional shocks do not influence voting in most cases. This asymmetric effect is stronger for games with higher stadium attendance, for local teams and for more intense emotional shocks. Similar findings arise when I focus on gubernatorial and presidential elections. These patterns cannot be explained by turnout and disappear when I focus on close electoral races. Furthermore I show that a similar close election pattern is also found in a different setting using data from Spain (Bagues and Esteve-Volart, 2016) and provide extra evidence from Google searches that individuals actively seek more information about candidates the closer the election. Overall these results are consistent with a behavioral political economy model with emotional voters whose voting behavior deviate from the forecasts of rational theory and such deviation decreases with how close elections are, as close electoral races tend to make information about candidates more salient in the media hence lowering the attention cost to picking the best candidate.

10:30-12:15 Session 6D: Human Capital
Chair:
João Victor Issler (FGV/EPGE, Brazil)
Location: Ametista
10:30
Luiz Brotherhood (FGV/EPGE, Brazil)
Bruno Ricardo Delalibera (FGV/EPGE, Brazil)
Do Developing Countries Overinvest in Public College?

ABSTRACT. In this paper we study the macroeconomic consequences of reallocating public expenditures from college education to basic education in a developing country. We construct a general equilibrium model that features heterogeneous agents, credit restrictions, public and private basic schooling and college. We present calibration results of a simpler version of the model to Brazilian data, and describe several counterfactual analysis that we plan to execute in the near future.

11:00
Bruno Delalibera (FGV/EPGE, Brazil)
Fernando Barros Jr (FUCAPE, Brazil)
The importance of teachers' human capital in an economy with market frictions

ABSTRACT. We identify a negative relation between teachers' salaries and income in Brazilian data. Then, we propose a general equilibrium model where workers decision are distorted due to market frictions and teachers' quality is computed endogenously and it is an input for the formation of human capital. The model is calibrated to the Brazil for 2013 and matches the data and closely reproduces the share of workers and the average wage for each state and occupation. Our benchmark economy suggests that there is a misallocation of workers in the brazilian economy and a reallocation of better workers to teachers' occupation could increase GDP due to a multiplicative effect of teachers' human capital. We also find that there is a strong outside option to teachers' career in more developed regions, which contribute for higher talented workers choose teachers' career in less developed regions.

11:30
Luiz Brotherhood (FGV/EPGE, Brazil)
João Victor Issler (FGV/EPGE, Brazil)
Measuring Human Capital Using Labor Market Data: An Application to the Study of Cross-Country Economic Growth

ABSTRACT. In this paper, we propose a novel estimate for human capital and use it in estimating cross-country production functions in a panel-data setting. We then compare our results with the ones previously obtained to emphasize the neglected importance that human capital has received so far in this context. Here, human capital is the expected present-discounted value of future labor income. We show how this can be easily implemented in empirical exercises, despite the fact that labor income is a non-stationary integrated process. Our human capital proxy has two interesting characteristics: first, it captures the well-known trend observed on the number of years of schooling, present in every country. Second, its cross-sectional variation (across countries) is much grater than that observed for the number of years of schooling, which we credit to the fact that it captures cross-country differences in the quality of human capital. Indeed, the coefficient of variation (unit free) of our proxy of human capital is more than 7 times that provided by the Penn World Tables -- Human Capital Index When we run panel-data regressions estimating the aggregate production function, we find a human-capital elasticity in production much larger than previous research has found. Our estimates are in the range 0.43-0.92, and very significant as well. These are larger than those obtained by Mankiw, Romer and Weil (1992), in the range 0.28-0.37. Comparisons with estimates using the human-capital proxy provided by PWT, 8.1, the so-called Human Capital Index, show that the latter is about 1/7 of the upper bound of our estimates, and about 1/4 of its lower bound. In growth-accounting exercises, we compared our results with two main papers in the literature -- Caselli (2005) and Schoellman (2012). We found a much more prominent role for human capital than previous research has found. We credit the success of our proxy, to the fact that market prices capture well the quality of human capital, something the previous literature has not been able to do.

10:30-12:15 Session 6E: Labor II
Chair:
Flávia Chein (UFJF, Brazil)
Location: Mezanino Bossa Nova
10:30
Juliana Camargo (FGV/EESP, Brazil)
Lycia Lima (FGV/EAESP, Brazil)
André Portela Souza (FGV/EESP, Brazil)
Flávio Riva (FGV/EESP, Brazil)
Vocational education and socioeconomic outcomes: experimental evidence from Brazil

ABSTRACT. The aim of this study is to provide rigorous evidence of the impact of a brazilian vocational education and training program on its beneficiaries. For this, we took advantage of the selection criteria for the program, in which the beneficiaries were selected through a lottery . Besides, our unique dataset allow us to evaluate the effect of the intervention not only on the economic dimension, but also on human capital, socio-emotional, crime and risky behavior outcomes. This is a valuable contribution since there is a lack of evidence in the literature about the impact of VET programs on these outcomes. Our estimations were based on ITT and LATE strategies. Overall, our results indicated that, at least in the short-run, there was no effect of the program on conventional labor market outcomes, such as employment and formal work probability and wages. Also, we found no effect in socio-emotional, crime and risky behavior dimensions. However, our results pointed towards a possible heterogeneity of the program, since we found positive and significant effects of the intervention on women, specially in labor market (formal work probability and wage) and socio-emotional (extraversion) outcomes.

11:00
Rafael Cayres (BNDES, Brazil)
Gustavo Gonzaga (PUC-Rio, Brazil)
Juliano Assunção (PUC-Rio, Brazil)
Tenure dependent Firing costs and turnover in Brazil

ABSTRACT. This paper maps turnover incentives present in Brazilian labor market institutions and quantifies their effect. These comprise both intended features such as advance notice and firing taxes, and unintended ones such as collusion between workers and firms to share unemployment insurance, and avoidance of mandatory benefits by firms. We use an endogenous turnover model based on learning about match quality, which produces rich firing hazard dynamics. The model is estimated by maximum likelihood using a dataset comprising accurate tenure information on all formal job contracts. The results indicate that, together with learning, the incentives considered can explain the empirical hazard very precisely, suggesting that they do play a role on the distribution of layoffs over the employment duration. However, the effects on the overall turnover, productivity and efficiency are small. We conclude that the distortions of firms' firing decisions alone are insufficient for generating a significative impact on turnover or productivity.

11:30
Cleiton Franco (UNEMAT, Brazil)
Breno Sampaio (UFPE, Brazil)
Gustavo Ramos Sampaio (UFPE, Brazil)
Paulo Henrique Vaz (UFPE, Brazil)
Tax Compliance Costs and Employment in SMEs: Evidence from Brazil

ABSTRACT. Small and medium-sized enterprises play an important role on economic development, they employ a great share of the labor force and may even trigger increases of aggregate productivity. In this paper we evaluate how their growth is affected by reductions on tax compliance costs. More specifically, we infer the causal effect of simplifying bureaucracy and reducing the tax burden on firms employment decisions in Brazil, one of the worst countries in the world to operate in terms of business taxes. Our identification strategy exploits the discontinuity among firms next to a eligibility threshold associated with the Simples Nacional program, which is meant to reduce compliance costs. Using firm-level data from the Annual Industrial Survey (PIA), we provide compiling evidence that companies participating in the program reduced their industrial operating cost by 23%, increased job creation in 21.5% and payroll in 25.18%.

12:00
Marcelo Arbex (University of Windsor, Canada)
Flávia Chein (UFJF, Brazil)
Isabela Furtado (FGV/EESP, Brazil)
Enlinson Mattos (FGV/EESP, Brazil)
Publicly Provided Private Goods and Informal Labor Supply
SPEAKER: Flávia Chein

ABSTRACT. In this paper we study how households spend their time in the formal and informal sectors and evaluate how publicly provided goods with and without market substitutes affect their time allocation. A simple static public provision model motivates our analysis. Households consume a normal private good and a quasi-private (education) good. Household needs some public utility services to consume the private good and supply labor to formal and informal sectors. Using data from the PNAD (National Household Sample Survey) for the period 2007-2015 we construct indexes of access to three groups of publicly provided goods: (I) basic infrastructure or public utility services, (II) basic education and (III) higher education. Our logit results show a positive effect of access to public education (basic and higher) on the probability of evasion. Differently from public utility services, that affect negatively the probability of evasion, the consumption of these goods present substitute in the private sector. We observe a stronger effect of access to publicly provided basic education comparing with higher education. This result may be related to the different quality of publicly provision of higher and basic education. In the first case, there are several public institutions that provide undergraduate and graduate courses of high quality, whereas in the basic education, on average, the quality of publicly provision is very low and worse than the private one. Tobit results suggest a positive and significant effect only in the case of publicly provided education, i.e., an increase in the access or use of publicly provided education increases the supply of informal labor hours.

13:45-15:30 Session 7A: Sessão Especial Itaú Social

Mensurando a Efetividade dos Professores: Observação na Sala de Aula

Chair:
Naercio Menezes-Filho (Insper e FEA/USP, Brazil)
Location: Jade
13:45
David Blazar (University of Maryland, United States)
Matthew A. Kraft (Brown University, United States)
Teacher and Teaching Effects on Students' Attitudes and Behaviors
SPEAKER: David Blazar

ABSTRACT. Research has focused predominantly on how teachers affect students’ achievement on standardized tests despite evidence that a broad range of attitudes and behaviors are equally important to their long-term success. We find that upper-elementary teachers have large effects on self-reported measures of students’ self-efficacy in math, and happiness and behavior in class. Students’ attitudes and behaviors are predicted by teaching practices most proximal to these measures, including teachers’ emotional support and classroom organization. However, teachers who are effective at improving test scores often are not equally effective at improving students’ attitudes and behaviors. These findings lend empirical evidence to well-established theory on the multidimensional nature of teaching and the need to identify strategies for improving the full range of teachers’ skills.

14:30
Andrew Bacher-Hicks (Harvard University, United States)
Mark Chin (Harvard University, United States)
Thomas J. Kane (Harvard University, United States)
Douglas O. Staiger (Dartmouth College, United States)
Validating Components of Teacher Effectiveness: A Random Assignment Study of Value-Added, Classroom Observations, and Student Surveys

ABSTRACT. Policy changes from the past decade have resulted in a growing interest in identifying effective teachers and their characteristics. This study is the third study to use data from a randomized experiment to test the validity of measures of teacher effectiveness. The authors collected effectiveness measures across three school years from three broad areas: value-added, classroom observation, and student surveys. In the first two years, they collected non-experimental estimates of these measures and, in the third year, they designed a randomized experiment to test the validity of these estimates. Using these data, they answer two questions: (1) Does a combination of these three distinct non-experimental measures identify teachers who, on average, produce higher student achievement gains following random assignment?; and (2) Does the magnitude of the gains correspond with what we would have predicted based on their non-experimental estimates of effectiveness? The analysis sample consisted of 66 fourth- and fifth-grade teachers from four large East coast school districts in the 2010-2011 through 2012-2013 school years. To answer the research questions, the authors first constructed the best linear combination of non-experimental student test score, survey, and classroom observation data from the first two years of the study (2010-11 and 2011-12) to predict teachers' average contribution to student growth on state standardized math tests another year. They used these predicted outcomes as their non-experimental estimates of teachers' contributions to student test score growth in 2012-13. Then, they examined actual student growth in 2012-13 (the third year of the study in which re randomly assigned students to teachers) and compared their non-experimental prediction of growth to actual growth. Tables are appended.

13:45-15:30 Session 7B: Theory I
Chair:
Humberto Moreira (FGV/EPGE, Brazil)
Location: Ametista
13:45
Matheus Costa (UnB, Brazil)
Paulo Henrique Ramos (UnB, Brazil)
Gil Riella (FGV/ENAE, Brazil)
Single-crossing Choice Correspondences
SPEAKER: Matheus Costa

ABSTRACT. Classic comparative statics methods relies most commonly on functional form assumptions in order to explicitly derive relations between endogenous and exogenous variables, on the implicit function theorem when no functional form is assumed, or in duality theories when a dual functions exists. Among the several properties an agent's preferences may have there are two known for it's interesting results. First, the single-crossing property, that allows for the ordering of a population of agents with respect to a particular classification of the options available. And secondly, the single-peaked property, that rules out the occurrence of condorcet cycles when the decision system is majority vote by pairs. In this paper we characterize some sufficient and necessary conditions over choice correspondences so that they have a pseudo-rational representation satisfying these properties.

14:15
Paulo K Monteiro (FGV/EPGE, Brazil)
Jaime Orrillo (UCB, Brazil)
Rudy Rosas-Bazán (Pontifical Catholic University of Peru, Peru)
Hyperopic Topologies on $l^{\infty}$
SPEAKER: Jaime Orrillo

ABSTRACT. Myopic economic agents are well studied in economics. They are impatient. A myopic topology is a topology such that every continuous preference relation is myopic. If the space is $l^{\infty}$, the Mackey topology $\tau _{M}(l^{\infty},l^{1})$, is the largest locally convex such topology. However there is a growing interest in patient consumers. In this paper we analyze the extreme case of consumers who only value the long run. We call such a consumer hyperopic. We define hyperopic preferences and hyperopic topologies. We show the existence of the largest locally convex hyperopic topology, characterize its dual and determine its relationship with the norm dual of $l^{\infty}$.

14:45
Salvador Barberà (Universitat Autonoma de Barcelona, Spain)
Danilo Santa Cruz Coelho (IPEA, Brazil)
On the advantages and disadvantages of being the first mover under rules of k names

ABSTRACT. Rules of k names are methods that allow two potentially conflicting parties to share the power to appoint officers. One of the parties (the proposer) selects k candidates out of a larger pool, and then the other party (the chooser) selects the winner from this restricted list. We investigate conditions under which the two parties could agree ex ante on the distributions of roles, one of them preferring to be the chooser and the other preferring to be the proposer. We show that this may not always be possible, and discuss what are the relevant characteristics of the environments where agreement can be reached.

15:15
Humberto Moreira (FGV/EPGE, Brazil)
Paulo Monteiro (FGV/EPGE, Brazil)
Vinicius Carrasco (PUC-Rio, Brazil)
Vitor Farinha Luz (The University of British Columbia, Canada)
Robust mechanisms: the curvature case

ABSTRACT. This paper considers the problem of a principal (she) who faces a privately informed agent (he) and only knows one moment of the type distribution. Preferences are nonlinear in the allocation and the principal maximizes her worst-case expected profits. A robustness property of the optimal mechanism imposes restrictions on the principal’s ex-post payoff function: conditional on the allocation being non-zero, ex-post payoffs are linear in the agent’s type. The robust mechanism entails exclusion of low types, distortions at the intensive margin and efficiency at the top. We show that, under additional assumptions, distortions in the optimal mechanism are decreasing with types. This monotonicity has relevant consequences for several applications discussed. Our characterization uses an auxiliary zero-sum game played by the principal and an adversarial nature who seeks to minimize her expected payoffs which also gives us a characterization of the worst-case distribution from the principal’s perspective. Applications of our framework to insurance provision, optimal taxation, non-linear pricing and regulation are discussed.

13:45-15:30 Session 7C: Gender
Chair:
Sergio Firpo (Insper, Brazil)
Location: Topázio
13:45
Olivier Marie (Erasmus School of Economics, Netherlands)
Arnaud Chevalier (Royal Holloway, UK)
Risky Moms, Risky Kids? Mother's Fertility Decision and Criminal Propensity of Children after the Fall of the Berlin Wall
SPEAKER: Olivier Marie

ABSTRACT. We study the link between parental selection and children criminality using a natural experiment which dramatically affected fertility decisions. Following the fall of the Berlin Wall, number of birth more than halved in East Germany. Such a change in fertility decision likely altered the composition of cohorts. Indeed, we estimate that the children from these (smaller) cohorts are 40 percent more likely to commit crimes, and that their mothers had worse observable characteristics. We further investigate the role of emotional attachment and risk attitudes in the fertility-crime relationship.

14:15
Laura Schiavon (UFJF, Brazil)
Claudio Ferraz (PUC-Rio, Brazil)
Breaking the cycle: the impact of legal reforms on domestic violence

ABSTRACT. This paper assesses the impact of a legal reform to protect women against domestic violence. Maria da Penha Law (MPL) was passed in Brazil in 2006 and implemented legal mechanisms aimed at effectively protecting women, punishing agressors and changing social norms. We use a differences-in-differences framework to estimate the dynamic law effect, exploiting the high comparability between male and female household homicide rates before 2006. MPL caused a gradual relative decline in female household homicide rates. The effect was concentrated in small municipalities and was greater in those where women presented lower wages, lower divorce rates, lower levels of education and lower rates of participation in the labor market in the year 2000. MPL’s high capability of preventing domestic violence is unconditioned to the creation of special criminal judicial bodies, despite their important role.

14:45
Paulo Arvate (FGV/EAESP, Brazil)
Sergio Firpo (Insper, Brazil)
Renan Pieri (Insper, Brazil)
The engagement of adolescent girls in politics when the electoral environment for women is uncertain
SPEAKER: Sergio Firpo

ABSTRACT. This paper investigates the political engagement of adolescent girls when the electoral environment for women is uncertain. We demonstrate that this environment simultaneously provokes less/more engagement in adolescent girls when women who are “role models” lose/win the election by reducing the expected general effect of empowered women on adolescent girls in politics. We use close elections in a mixed-gender race (RD estimate) as an empirical strategy for three reasons: to obtain unbiased estimated causality between the variables; to compare the robust difference in engagement of two possible electoral results (defeat/victory); and to observe if there are discontinuities provoked by the behavior of adolescent girls with regard to women’s electoral target: victory. We verified that two discontinuities exist and might be justified by the electoral disappointment/euphoria of adolescent girls with regard to the quasi-reached electoral target. In addition, guided by their value function, a growing margin of female victory/defeat provokes increasingly less engagement on the part of adolescent girls to vote in the next election. The positive influence of empowered women on the political engagement of adolescents is tiny when the result is close to the target, given that we find an increase/reduction in adolescent registrations to vote when women are elected/defeated: the difference between them is 1.0-1.6 p.p. Unlike the literature, our measure of engagement in politics is not generated from a survey (a measure of the perception of the aspirations of adolescents) but from the non-compulsory registration of adolescents in the 16-17 age-band for voting in Brazil

13:45-15:30 Session 7D: Finance II
Chair:
Flávio Ziegelmann (UFRGS, Brazil)
Location: Água Marinha
13:45
Pedro Luiz Paulino Chaim (FEA-RP/USP, Brazil)
Márcio Poletti Laurini (FEA-RP/USP, Brazil)
Foreign Exchange Expectation Errors and Filtration Enlargements

ABSTRACT. Extrapolations of future markets forward rates are a better pre- dictor of the 30-days ahead BRL-USD exchange rate than forecasts from the Focus survey of Brazilian market participants. This is puzzling because market participants observe forward rates as they submit predictions. Dandapani and Protter (2016) describe a mechanism through which new information enlarges the information set (a ltration), changing the underlying risk neutral measure and inducing a drift into the martingale process, turning the process into a strict local martingale. We argue this mechanism can explain our rational conundrum. To empirically test the plausibility of such connection we rst employ a nonparametric identication test based on Jarrow et al. (2011a), then estimate the stochastic volatility model of Andersen and Piterbarg (2007). Results suggest that Focus survey forecasts indeed display characteristics of a strict local martingale, while spot exchange rates and forward rates are con- sistent with a martingale process.

14:15
Assessing systemic risk in Brazil and its determinants with ΔCOVAR

ABSTRACT. This paper is a contribution to the literature on systemic risk and its determinants in emerging economies. Based on data for the Brazilian economy for the period from 2011 to 2015, we perform an analysis in two steps. In the first step, we measure the systemic risk in Brazil based on ΔCoVaR framework. In the second step, we present empirical evidence from a panel data analysis regarding the determinants (banking and macroeconomic variables) of the systemic risk. The findings denote that the systemic risk measured using the ΔCoVaR methodology is consistent with the main events contained in the Financial Stability Reports issued by the Central Bank of Brazil. Furthermore, the evidence indicates that the regulatory authority should monitor bank account balances and that prudential regulation policy must be coordinated with monetary policy in order to mitigate the systemic risk.

14:45
Fernando Silva (UFRGS, Brazil)
Flávio Ziegelmann (UFRGS, Brazil)
Worst Case CVaR Portfolio Optimization with Multidimensional Copulas

ABSTRACT. Using data from the S&P 500 shares from 1990 to 2015, we measure the downside market risk by Conditional Value-at-Risk (CVaR) subject to return constraints following the approach of Rockafellar and Uryasev (2000, 2002) and the extended framework of Kakouris and Rustem (2014) through the use of multidimensional mixed archimedean copulas. We implement a dynamic investing strategy where the portfolios are optimized using three different length of rolling calibration windows. The out-of-sample performance is evaluated and compared against two benchmarks; a multidimensional gaussian copula model and a constant mix portfolio. Our empirical analysis shows that the Mixed Copula-CVaR approach generates portfolios with better downside risk statistics for any rebalancing period and it is more profitable than the Gaussian Copula-CVaR and 1/N portfolios for daily and weekly rebalancing. To cope with the dimensionality problem we employ a similar approach to that used by Gatev, Goetzmann, and Rouwenhorst (2006) to select a set of assets that are the most diversified, in some sense, to the S&P 500 index in the constituent set. We find that copula-based approaches offer better hedges against losses than the 1/N portfolio. The accuracy of VaR forecasts is determined by how well they minimize a capital requirement loss function (CR). In order to mitigate data-snooping problems we apply a test for superior predictive ability (SPA) proposed by Hansen (2005) to determine which model significantly minimizes this expected loss function. The test is implemented via stationary bootstrap of Politis and Romano (1994) using the automatic block-length selection of Politis and White (2004) and 10,000 bootstrap resamples. We find that the minimum average loss of the mixed Copula-CVaR approach is smaller than the average performance of the Gaussian Copula-CVaR.

15:45-17:30 Session 8: Keynote lecture
Chair:
Vladimir Ponczek (FGV/EESP, Brazil)
Location: Jade
15:45
Flávio Cunha (Rice University, United States)
Inequality in Human Capital Investments in Early Childhood
SPEAKER: Flávio Cunha
17:45-19:30 Session 9: Minicurso Itaú - Aula 1
Location: Ametista
17:45
Sergio Almeida (FEA/USP, Brazil)
Economia Comportamental: Desenvolvimentos Teóricos e Implicações para Políticas Públicas