📈 Determinants of Infant Mortality in Minas Gerais: A Panel Data Analysis

An econometric study on the socioeconomic factors impacting infant mortality across municipalities in Minas Gerais, Brazil.

⚠️ Important Disclaimer regarding Source Code

The Stata do-files were intentionally omitted from this repository as the data cleaning and transformations were performed directly on the .dta file without a script backup at the time. However, all essential methodological steps, variable constructions, and econometric outputs are fully documented and available in the attached PDF report.

---

🎯 Project Overview

This research explores the primary factors influencing infant mortality rates in the state of Minas Gerais. By utilizing a robust dataset of municipal indicators, the study aims to identify how variables such as education, sanitation, and economic performance correlate with early childhood health outcomes.

---

🛠️ Methodology & Econometric Approach

The project employs advanced econometric techniques to handle the multidimensional nature of the data (municipalities over time).

1. Models Implemented

• Pooled OLS: Used as a baseline for linear regression analysis.
• Fixed Effects (Within Transformation): To control for time-invariant unobserved heterogeneity specific to each municipality.
• Random Effects (GLS/MQG): To account for potential serial correlation and provide efficient estimates when appropriate.

2. Statistical Validation (Tests)

To ensure the reliability of the results, the following tests were conducted:

• Breusch-Pagan Lagrange Multiplier: To choose between Pooled OLS and Random Effects.
• Hausman Test: To determine the consistency of Random Effects versus the robustness of Fixed Effects.

---

📊 Key Variables & Indicators

The analysis integrates data from multiple sources (IBGE, DATASUS, etc.), focusing on:

• Infant Mortality: Mortality rates (Dependent Variable).
• Education: Average years of schooling (Adults 25+).
• Economic: Municipal GDP and indebtedness levels.
• Social: Unemployment rates and local infrastructure.

---

📈 Summary of Findings

The results highlight that:

1. Education remains a crucial factor; higher levels of schooling for adults significantly correlate with lower infant mortality rates due to better preventive care.
2. Public Policy: The study suggests that government awareness programs and infrastructure investments are vital for reducing mortality trends in Minas Gerais.

---

👥 Authors

• Artur Damasceno
• Lauro Aguiar
• Lucas Godinho
• Vinícius Nahan

---

This is an academic and educational project developed as part of the Econometrics III coursework at Ibmec.