»  Home »  Stata Conferences »  2019 Brazil

The 2019 Brazilian Stata Conference was held on 5 December at the University of São Paulo.

Proceedings

9:00–9:45 Water and birth outcomes: Lessons from a policy intervention in a climate-vulnerable area Abstract: This presentation studies how in-utero exposure to a large water harvesting program affects birth outcomes. We assess the effects of the Cisterns Program, which built about 1,000,000 cisterns in Brazil's poorest and driest region to
... (Read more)
promote small-scale decentralized rainfall harvesting and storage. Our empirical strategy compares the outcomes of women exposed to cisterns in different stages of their pregnancy. The results show that access to cisterns during early pregnancy increases birth weight, particularly for more educated women. Our findings suggest that policies for adaptation and reduction of vulnerability may bring about positive effects on an important predictor of future individual outcomes.
(Read less)

Additional information:
Brazil19_da_Mata.pdf

Daniel da Mata
Escola de Economia de São Paulo da Fundação Getulio Vargas
9:45–10:10 A command to study the impact of different correction methods on cancer mortality trends and magnitude Abstract: Dealing with ill-defined causes of death in vital registration is important to estimate mortality trends in ecological studies. In Brazil, authors have opted for different approaches to deal with ill-defined causes of death. The new
... (Read more)
charon command permits one to calculate mortality trends for 41 cancer types in 4 different geographical levels while allowing different types of correction for ill-defined causes of death. The command was designed to compare the impact of different distribution methods on cancer mortality and magnitude. It can be applied differently to investigate cancer mortality using the microdata provided by the Brazilian Ministry of Health through the Mortality Information System.
(Read less)

Additional information:
Brazil19_Bigoni.pptx

Alessandro Bigoni
University of Sao Pãulo
10:10–10:35 Timing restriction and information provision effects on birth type choice Abstract: Scheduled Cesarean sections (CSs) before 39 weeks of gestation are common in Brazil. They are frequently associated with respiratory and other adverse neonatal outcomes. In 2016, Brazil implemented a policy prohibiting (CSs)
... (Read more)
before the 39th week of pregnancy by the mother's request. This research measures this law's effects on the scheduled CSs ratio and natural delivery (ND) ratio. A differences-in-differences methodology is applied using DiD and DiDiD regressions to quantify the policy effects. The policy decreased the scheduled CSs ratio up to 1.7 percentage points for births at the 37th week and 1.95 percentage points at the 38th week of gestation, equivalent to a 5% decrease in the number of scheduled CSs. However, the policy increased the scheduled CS ratio up to 1.1 percentage points for births at the 39th week, equivalent to a 3.6% increase in the number of scheduled CSs. Therefore, it is concluded that the policy induced women to postpone scheduled CSs from the 37th and 38th weeks to the 39th gestational week.
(Read less)

Additional information:
Brazil19_Braga.pdf

Lucas Braga
Dados de Consultoria
11:05–11:50 Lasso for prediction and inference after model selection Abstract: Lasso is a popular machine-learning technique used for model selection and prediction. Although lasso has been used widely for prediction, using the model selected by lasso for estimation must be done carefully. The standard errors
... (Read more)
that come from taking the model selected by lasso as the true model and then fitting a regression are incorrect. I will talk about lasso for prediction and will show you how to avoid the perils of using lasso after model selection to fit models.
(Read less)

Additional information:
Brazil19_Pinzón.pdf

Enrique Pinzón
StataCorp
11:50–12:15 Diferencial de salários público-privado e ciclo econômico no Brazil Abstract: Este trabalho tem como objetivo analisar o comportamento do prêmio salarial pago ao funcionalismo público estadual brasileiro, relativamente ao setor privado. Esse diferencial é positivo e corroborado pela literatura, bem como pelas
... (Read more)
equações mincerianas aqui estimadas utilizando os microdados da PNAD anual de 2002 a 2015. Uma vez quantificados os diferenciais por ano e por UF, eles foram adotados como variável dependente em uma abordagem de painel para verificar a hipótese de um comportamento cíclico, fazendo-se uso de covariadas como a taxa de variação do PIB real, a taxa de desemprego, indicadores de finanças públicas e participação dos funcionários públicos na força de trabalho. Para a estimação do modelo de painel o estimador de Arellano-Bond mostrou-se o mais adequado, dado o comportamento auto-regressivo do prêmio salarial, tendo sido adotadas as contribuições de David Roodman (2006) através do pacote xtabond2. Os resultados mostram que o diferencial de salários apresenta dois comportamentos, um pro e outro contracíclico, o primeiro atrelado aos reajustes acima da média obtidos pelo setor público em anos de crescimento econômico, e o segundo associado à rigidez de demissões no setor público em períodos de aumento do nível de desemprego da economia.
(Read less)

Additional information:
Brazil19_Martins.pdf

Fernando Fabian de Freitas Martins
Daniela Vaz
UNIFESP
12:15–12:40 Mulheres no mercado de trabalho: Uma análise sobre o prêmia salarial urbano no Brazil Abstract: O objetivo é avaliar o mercado de trabalho feminino nas Regiões Metropolitanas (MA) brasileiras quanto ao prêmio salarial urbano (UWP). Tal objetivo é, em si mesmo, uma contribuição relevante para a literatura que majoritariamente
... (Read more)
analisa o UWP para os homens. Agrega-se outras contribuições à análise do UWP observando-o nos diferentes grupos de trabalhadores (formais-informais e com diferentes posições no domicílio) conforme o tamanho da MA. São utilizados os dados da Pesquisa Nacional por Amostra de Domicílios Contínua (PNADC) revelando que o UWP das mulheres supera o dos homens (com exceção de mulheres que são chefes do domicílio) com pouca variaçõo entre as MAs, enquanto os homens possuem menor UWP quanto maior for a MA. Existe um padrão similar de UWP entre homens e mulheres entre setores, porém o prêmio dos homens segue o padrão dos trabalhadores informais, o que não ocorre no grupo de mulheres. Regressões quantílicas mostram que o UWP tem diferente magnitude, ao longo da distribuição, e trajetória entre homens e mulheres, principalmente nas Medium e Large MAs. Assim, o UWP pode estar subestimado quando se considera apenas o grupo de homens, mas é superestimado ao não considerar a distribuição (resultado consistente independentemente do tamanho da MA).
(Read less)

Additional information:
Brazil19_de_Almeida.pdf

Eloiza Regina Ferreira de Almeida
Veneziano de Castro Araújo
Solange Ledi Gonçalves
Universidade Federal de São Paulo (UNIFESP)
2:05–2:50 The role of teachers' expectations on the human capital formation technology Abstract: Recent literature on economics on education shows that parents' beliefs impact children’s educational outcomes (Cunha et al. [2018]), Boneva and Rauh [2018]). Teachers, on the other hand, also play a fundamental role on children's
... (Read more)
education, and it is expected that their beliefs also have an impact on child development. This presentation aims to understand the role of teachers' beliefs about the importance on the non-cognitive skills on their task allocation, and the impact on teachers’ tasks allocation on student’s outcomes. We collected data on 168 third- and fourth-grade teachers and about 3,500 students at 84 municipal schools in Rio de Janeiro. Following a human capital formation model that incorporates teachers’ beliefs in a simple way, we created an instrument to measure teachers' investments in their students and their beliefs on the importance of noncognitive skills. To overcome endogeneity issues, we conducted a randomized control trial in which half of teachers received text messages with pieces of evidence on the importance of socioemotional skills. We are able to show that such intervention had a positive and significant effect on teachers' socioemotional belief, which led teachers to increase their socioemotional investments by about 6%, compared with the teachers in the control group. In addition, treatment has a direct positive impact on students’ executive function and emotion recognition skills. This result has a clear implication for public policy because we show that a simple and inexpensive informational intervention can improve students' skills without the need for long training or directly interfering with school routine.
(Read less)

Additional information:
Brazil19_Pinto.pdf

Cristine Pinto
Escola de Economia de São Paulo da Fundação Getulio Vargas
2:50–3:15 Gravity effects of culture, institutions and religion distances on internal migration in Brazil Abstract: This presentation evaluates the dynamics of the influx of mass internal migration in Brazilian culture, institutions, and religion. This issue is getting more attention in the literature. We conducted an empirical analysis of the role of cultural,
... (Read more)
institutional, and religion distance in migration inflow of the Brazilian population using traditional gravity models in international economics. Origin and destination cities are characterized by intrinsic differences in culture, institutions, and religion. We confirm some expectations about the explanatory power of variables like personal trust, community trust, market orientation, collectivism, personal collectivism, freedom politics, corruption, hard work and thrift, intention of migrating to another country, uncertainty avoidance, and religious diversity as attraction factors of people for choosing a county over another because our sample is restricted to Brazilian migrants. For this, we employ the data of Censo 2010, considering that the original place is the birth place of the individuals. About the methodological issue, we consider that it is necessary to employ fixed effects to account for the individual variability of municipalities. Furthermore, we estimate with Poisson pseudo maximum likelihood (PPML) accounting for the fixed effects. For this, we apply the new community-contributed ppmlhdfe command.
(Read less)
Daisy Assmann Lima
Brazilian Public Defense Union (DPU)
Philipp Ehrl
Universidade Católica de Brasília (UCB)
3:15–3:40 An instrumental-variables estimation of the impacts of fiscal incentives to Brazilian movies Abstract: The Brazilian motion picture industry is highly dependent on governmental incentives. The main public policy instrument in the last decades was given by the fiscal deduction to the investment and sponsorship of national movies.
... (Read more)
Between 1995 and 2016, 930 movies from a total of 1,370 national movies exhibited in Brazil were financed through fiscal incentive mechanisms, where this financial promotion reached R$ 5.3 billion. Given the scarcity of resources allocated to culture (0.3% of public expenditure in 2010), it is even more relevant to evaluate the efficacy of public policy over cinema (approximately 3% of public expenditure on culture in 2010). A separate public policy analysis has indicated that the main declared objective was increasing Brazilian consumption of national movies, which guided the definition of a proper empirical strategy to the impact evaluation. Then, given the novel database constructed, the instrumental-variables estimation strategy revealed a low box office elasticity with respect to fiscal incentive, which was equal to 0.05. Together with the government that cuts some of its tax revenues, the investor and the sponsor are crucial to the functioning of the fiscal incentive mechanism. Databases concerning those supporters were used to explore the exogenous variations in its incomes to conceive exogenous variations in the fiscal incentive provided for national movies.
(Read less)

Additional information:
Brazil19_Pereda.pdf

Paula Carvalho Pereda
Matheus Rosso
University of São Paulo
3:55–4:20 Are cryptocurrencies suitable for portfolio diversification? Cross-country evidence Abstract: In this presentation, we analyze whether a representative investor, who holds a well-diversified portfolio of stocks, may benefit from investing in cryptocurrencies. Our analysis encompasses several capital markets and the four most liquid
... (Read more)
cryptocurrencies. Using country-specific stock market indices and risk-free rates, results indicate that, for most countries, cryptocurrencies fit in the tangent portfolio (maximum Sharpe ratio), but not—or very little—in the minimum variance portfolio (MVP). Cryptocurrencies' returns are riskier, but their co-movements with global stock indices are close to zero, on average. Globally, we find that the optimal holding of cryptocurrencies is represented only by Bitcoin (BTC) and at a 4.6% share of the portfolio (when not considering short sales). Importantly, optimization exercises using rolling windows reveal that the optimal weight of cryptocurrencies on the tangent portfolio is extremely both sensitive to the period being analyzed and the premises on expected returns.
(Read less)

Additional information:
Brazil19_Colombo.pptx

Jéfferson A. Colombo
Escola de Economia de São Paulo da Fundação Getulio Vargas
4:20–4:45 Giving with one hand and taking away with the other Abstract: This presentation analyzes the impact of fiscal policy on income inequality for the Brazilian states over the period 2004–2014. Using the GMM system, we find robust evidence that, while increases in total expenditure, spending on social
... (Read more)
assistance and security, and expenditure on infrastructure reduce inequality, total and capital revenues and expenditures related to public debt increase inequality. Thus, the net effect of fiscal policy on inequality may not be positive, mainly because the tax system is still very regressive and many of the resources are spent on debt. Therefore, the discussion of how to determine and qualify social spending levels in Brazil, which are necessary to fight poverty and reduce social and income inequalities, is directly related to the revision of the tax collection system.
(Read less)

Additional information:
Brazil19_Fernandes.pdf

Carlândia Brito Santos Fernandes
Marina Silva da Cunha
Marcos Roberto Vasconcelos
State University of Maringá
4:45–5:10 Trillion RGPS Show? Quantifying the fiscal and distributional aspects of Bolsonaro's pension reform Abstract: This presentation aims to calculate the fiscal and distributive impacts on the Brazilian National Pension Scheme (RGPS), generated from the original version of the Constitutional Amendment 6 (PEC 6/2019) (the Bolsonaro government's
... (Read more)
pension reform). This is done with a micro-simulation model that calculates the contributions to the pension system as well as old-age and survivors' benefits over a 30-year period. The approval of the PEC would reduce expenses from BRL 13.42 trillion to BRL 10.59 trillion in the three decades analyzed. In the first 10 years, the net result would be BRL 952 billion. Net pension liabilities would fall from BRL 5.99 trillion to BRL 3.43 trillion, a 42.74% reduction. The average replacement rate would fall from 73.99% to 67.65%. The required contribution rate would reduce significantly from 42.70% to 32.87%. The average internal rate of return would fall from 2.37% to -1.00%. These results show that the reform would greatly reduce but not eliminate the RGPS deficit. There is a remarkable reduction in all pension indicators, particularly for retirement by age. The exception is the retirement by time of contribution for women.
(Read less)

Additional information:
Brazil19_Afonso.pptx

Luús Eduardo Afonso
João Vinícius de França Carvalho
Universidade de São Paulo
5:10–6:00 Wishes and grumbles
Enrique Pinzón
StataCorp

Organizers

Scientific committee

Organizing committee

Registration

Registration is now closed.

Visit the official meeting page for more information.

Venue

Logistics organizer

The logistics organizer for the 2019 Brazilian Stata Conference is Timberlake Analytics Brazil, the distributor of Stata in Brazil.

View the proceedings of previous Stata Conferences.


Stata

Shop

Support

Company


The Stata Blog: Not Elsewhere Classified Find us on Facebook Follow us on Twitter LinkedIn YouTube Instagram
© Copyright 1996–2019 StataCorp LLC   •   Terms of use   •   Privacy   •   Contact us