The first Brazilian Stata Users Group Meeting takes place on Friday, December 2, 2016, at the Faculdade de Economia e Administração da Universidade de São Paulo.
The meeting will provide Stata users the opportunity to exchange ideas, experiences, and information on new applications of the software. Everybody who is interested in using Stata is welcome.
Invited speakers include Naercio Aquino Menezes Filho (FEA-USP e Insper), José Feres (IPEA e FGV-EPGE), and Bill Rising (StataCorp).
Income shocks, early child development, and long-term human capital outcomes in Brazil
Invited speaker: Naercio Aquino Menezes Filho
FEA-USP e Insper
Economic evaluation of a housing project: diff-diff and matching using Stata
Abstract: In recent years, the importance of economic evaluation in social areas has been growing. The challenge has been to evaluate the effectiveness of an action that may have an explicit goal but an impact on different dimensions in the life of a family or person. Thus this article aims, through the use of Stata software, to show the applicability and potential of the instruments of economic evaluation, particularly the diff-diff method using matching, to understand the results of an urban project in a low-income community in Brazil. The project intends to eliminate housing precariousness that could result in health and safety problems for the residents. We constructed several indicators using information at the baseline of the project and ex-post data collected seven years after the beginning of the project. The indicators reflected different dimensions of impact: improvement in the home environment, in health, and in the perceptions of residents on safe housing. With the construction of different control groups, considering effects of contamination, we observed a statistically significant impact on some indicators, despite their low magnitude.
Danielle Carusi Machado
Faculdade de Economia da UFF
Income-related inequalities in utilization of health services among private health insurance beneficiaries in Brazil
Background: Brazil's 1988 health reform resulted in a two-tier system represented by those with dual coverage—public and private—and those who rely solely on the public insurance—SUS. Several studies show that private insurance increases health service utilization levels, but can private coverage increase utilization regardless of the individuals' income?
Methods: This study uses 1998 and 2008 PNAD data, assessing utilization across income and calculating concentration indexes as summary measures of inequality. Dependent variables' distributions across income are standardized by need using the indirect method. Concentration curves compare inequality over time, and decomposition analysis identifies inequality contributors. All the analyses, including data preparation, are done using Stata 13.
Results: Physician services present low inequalities. SUS hospitalizations are rare but concentrated on the poor. Poor individuals are less likely to use private hospitals than the rich. In 1998, this was not the case, suggesting that insurers have developed mechanisms to deter hospital utilization for the poor.
Conclusion: The Brazilian Private Insurance Regulatory Agency (ANS) should monitor utilization across income. Making ambulatory and inpatient only coverage available would increase options for consumers that could sort more appropriate coverage according to their capacity to pay and their health needs.
Asymmetric price transmission in the Brazilian rice market: A review of methodologies
Abstract: The following paper presents and briefly discusses some arguments on the price asymmetry literature by applying the most used methodologies to the case of the rice market in the Brazilian economy. We assess the econometric models by the aftermath of each of them to that specific case. The results using the error correction model (ECT) indicated that there is some asymmetry, but it is not possible to infer that this asymmetry remains in the long run. Previous methodologies also showed that upward and downward price movements between these markets occurred in a different level and speed; the variables are significantly different for both rising and falling price phases, which demonstrates that the more sophisticated models can indeed forecast results accurately, shedding light on the output of prior models.
Jacques Henrique Dias
José Luis Parré
Waldemiro Alcântara da Silva Neto
Zero-inflated models in Stata
Abstract: Count models represent an important part of an econometrician's toolkit. Stata 14 presents a command suite aimed at estimating count models, with an emphasis on its main distributions (Poisson and Gamma–Poisson). However, not much has been made of a related application of such models in Stata, namely, zero-inflated models (ZIM). In this tutorial, we present a detailed guideline related to ZIM estimation, with an application to a dataset based on corruption practices among United Nations' diplomats. In doing so, we want to illustrate pitfalls related to analyses that ignore the presence of excessive zeros in the dependent variable. The results reported in this tutorial are important for emphasizing potential uses of this new class of models in distinct areas of applied econometric research. In our view, both researchers and practitioners could obtain substantial gains and insight by incorporating models with excess zeros in the dependent variable in their toolkits.
Luiz Paulo Fávero
SEM in health sciences: Combining multiscale questionnaires and checking the performance of GOF parameters up to the minimally acceptable sample size
Abstract: In health sciences, relevant issues are handled with complex questionnaires, such as quality of life (QOL) and stress evaluation instruments. These questionnaires oftentimes present dozens of indicators under Likert scales. However, Likert scales can be challenging to curb with an overarching "regression" approach. What is more, ordinal in principle, they usually present a skewed distribution even after the algebraic transformation in 20-point or 100-point scales. The panoply of scales leads to a plethora of criteria of normality. To approach several questionnaires at once and, at the same time, to provide reliable measures of association among them, the analysis may rely on the standardization of the coefficients. We present a strategy to work with complex stress and QOL questionnaires assembled into an overarching model. For this case study, two latent variables were created as reflective "exogenous" indicators. The number of loadings was decreased by parceling items by similarity and treating these parceled constructs as endogenous variables. After performing the modeling in the total sample (around 600 individuals), GOF tests of the original model were compared with random subsamples. With this strategy, the model's reliability as well as the performance of GOF parameters under progressively smaller sample sizes was analyzed.
Marcos Antonio Almeida Santos
Universidade Tiradentes and Harvard T. H. Chan School of Public Health
A platform for batch analysis of national health surveys using Stata
Abstract: The International Center for Equity in Health (ICEH) excels at providing highly reliable, standardized, disaggregated data for several maternal, newborn, and child health indicators and at being able to quickly process new indicators and deliver these results for over 300 national household surveys representing around 100 countries. For the past five years, a platform for analyzing such surveys was developed using Stata, making use of its powerful programming language. The code must be flexible enough to allow for variations in the variables, names, and categories. We use a dataset for selecting the surveys to be analyzed and to assist the platform with nonstandard information, such as country-specific water sources and sanitation facilities. This information is extracted by a subroutine that sweeps the raw datasets and exports them to a spreadsheet for manual confirmation. At first, the platform generates an intermediate dataset with individual level data for each health indicator as a variable. Then, it estimates coverage levels for each of the indicators for population subgroups, taking into account the survey sample design and saving additional statistics such as standard errors, confidence intervals, and p-values. The platform regularly calculates hundreds of indicators from hundreds of surveys.
Leonardo Zanini Ferreira
Determinants of environmental innovation and its impacts on the productivity of the Brazilian manufacturing sector"
Invited speaker: José Gustavo Feres
IPEA e FGV-EPGE
Stata for Big Data and Data Science — Overview and Prospects
Debora Kusmierska Bilard
Consultora em inteligência de mercado, economia e finanças
In-kind transfers: Distortion effects and impacts on household consumption
Bruno Toni Palialol
Abstract: Stata 14 introduced a suite of commands for performing Bayesian analysis directly in Stata. This presentation will be a hands‐on demonstration of these commands that will illustrate the full path of a Bayesian analysis from specification of prior distributions through use of MCMC chains to estimate the model to assessment of the mixing and convergence of the MCMC chain. Along the way, the flexibility of Bayesian models as well as the ways in which the analysis and interpretation differ from their frequentist counterparts will be illustrated.
Wishes and grumbles
Faculdade de Economia e
Administração da Universidade de São
Av. Prof. Luciano Gualberto
908 - Butantã
São Paulo - SP
The meeting is free, though registration is required. Participants must travel at their own expense.
You may also register by email to Timberlake Analytics Brasil or by telephone to +55 11 4680-3503. Include your name, email, institution, department, and phone number.
Faculdade de Economia e Administração da Universidade de São Paulo
Faculdade de Economia e Administração da Universidade de São Paulo
Faculdade de Medicina da Universidade de São Paulo
The logistics organizer for the 2016 Brazilian Stata Users Group meeting is Timberlake Analytics Brasil, the distributor of Stata in Brazil.
For more information on the 2016 Stata Users Group meeting, visit the official meeting page.
View the proceedings of previous Stata Users Group meetings.