Brazil Trade Determinants

Project Summary

This dissertation—completed at the University of York—investigates the determinants of Brazil’s bilateral trade flows with 13 partner countries using an extended gravity model. I applied OLS, Fixed Effects & Random Effects regressions to a 1990–2020 panel dataset, achieving up to 94.7% explanatory power.

Challenge — Approach — Results

• Challenge: Identify economic & geographic drivers of Brazil’s trade, focusing on Mercosur and FTAs.
• Tools: Excel (data prep), Stata (econometric modelling), R (data validation), gravity model framework, panel data techniques.
• Approach: Cleaned & log-transformed trade, GDP, and endowment data; estimated OLS, FE, and RE models; ran robustness checks.
• Results: Distance negatively impacts trade; GDP & population are strong positives; FTAs & exchange-rate terms show model-specific effects; models explain up to 94.7% of trade variation.

Key Tables

Table 1: Descriptive Statistics

Table 2: OLS vs FE vs RE Comparison