Acknowledgements
Abstract
Income inequality in access to tertiary education in Brazil has reached its lowest recorded level, despite expansion through a predominantly private system in which access to most seats depends on cappacity to afford tuition. What explains this change in inequalit? How did state policies alter the terms of access, and through what political processes were they enacted? This dissertation addresses these questions in three parts. Part I brings into dialogue two complementary traditions: social stratification research, which has documented individual-level educational inequality while leaving political processes largely in the background, and comparative political economy, which has theorized education as a redistributive problem without measuring access inequality at the individual level. It reconstructs their premises and identifies the conceptual space opened by their joint reading for the two subsequent empirical parts.
Part II adjudicates among competing explanations. It argues that continuation decisions reflect household income constraints, educational costs, and expected returns, rather than saturation or competition dynamics. Using harmonized PNAD and PNAD Contínua surveys (1992–2025) and inference to the best explanation (IBE), it identifies a four-phase trajectory: a 1990s plateau, a sharp decline in inequality between 2005 and 2015, stagnation from 2015 to 2021, and renewed decline thereafter. This trajectory tracks two channels: affordability, as real incomes rose among lower-income households while tuition fell relative to income; and de-commodification, as targeted policies reduced the dependence of access on family income. Free public secondary education was a necessary precondition for both. The evidence indicates that affordability was the larger contributor to the decline; saturation and competition accounts lack the premises their mechanisms require in Brazil.
Part III uses IBE process tracing with legislative records, official documents, and a systematically collected press corpus. It finds that the relevant partisan cleavage concerns subsidization and its redistributive design, rather than enrollment expansion; for-profit educational actors and social movements decisively shaped policy in opposing directions; and path dependence better explains persistent private-sector dominance than partisan ideological preference. Brazil is thus a diagnostic case that refines theories developed primarily for advanced OECD democracies: reducing inequality in educational access requires both redistributive income policy and redistributive educational instruments.