This paper analyses the existence of a gender bias in the parents' decisions when they spend money on the education of their children. A sample of families is obtained from a Household Budget Survey (POF) collected in 2002-2003, by the Brazilian Institute of Geography and Statistic (IBGE). Tobit models are used to estimate Engel curves for the family education expenses and expenditures with educational items, as a function of socioeconomic and demographic characteristics of the households. Wald tests are used to check the equality of the parameters estimated with the number of enrolled sons and daughters. Such tests do not indicate discrimination against daughters regarding educational fees.
Gender bias; Educational expenditure; Engel curve; Tobit censored regression
