Abstract
<jats:p>We estimate the long-term intergenerational mobility (IM) in Brazil, by mesoregion, utiliz ing the methodology developed by Güell, Mora and Telmer (2015) and Güell et al. (2018). Benefiting from the unique features of Brazil, such as data availability on income, edu cation, full names, our study employed a machine learning algorithm to estimate sur name-based ancestry. This allows us to bypass limitations of previous studies and provide an initial exploration of the link between intergenerational mobility (IM) and inequality in a middle-income, Latin American country. We also examined how IM spatially aligns with historical factors and current socioeconomic development indicators in Brazil. Our find ings connect past land inequality and slavery to lower long-term mobility, which is linked to lower current income per capita and educational attainment. Unlike other IM studies in Brazil, we found no clear geographic patterns. Additionally, there is no distinct relationship between inequality and social mobility, or the Gatsby Curve, at the mesoregional level.</jats:p>