Inequality in Brazil seen through consumption behavior: toward better economic modeling with utility function estimation

Household consumption accounts for approximately 60 percent of Brazil's GDP. Accurate representation of this component in applied economic models requires accounting for household preferences to capture spending heterogeneity across income groups. This paper is the first to estimate parameters of a Stone-Geary utility function for Brazil, covering 31 productive sectors and ten household income deciles. Results reveal a negative relationship between income and subsistence spending share in total expenditure, especially for food and housing. Higher-income households allocate a greater proportion of their income to services, such as private education, healthcare, tourism, culture, and recreation. The estimated parameters are applied in Computable General Equilibrium (CGE) simulations to evaluate their importance. Compared to a non-parameterized scenario, the estimated model yields 30% smaller GDP impacts, 40% smaller sectoral output effects, and heterogeneous shifts in consumption patterns, including 15 times higher impacts on imports and 6 times greater impacts on insurance by top-income households.