Social policy has been transformed in Mexico during the last two decades. Social insurance programmes on which welfare provision was based throughout the twentieth century have been retrenched and new social assistance programmes have been introduced and expanded. This thesis aims to reveal the redistributive potential of these social policy reforms. The research maps the changes in the welfare policy architecture and estimates their vertical and horizontal redistributive effects. The study identifies the winners and losers of the reform process, the specific changes in the structure of the architecture that benefited different population groups and characterises the welfare model that emerged from the reform process. The research consists on the hypothetical simulation of the effects of changes in the tax/benefit system. The thesis adapts the ‘model families’ method to estimate and compare the effects of taxes and benefits at the household level before and after the reforms. Findings highlight that the productivist/residual logics adopted in the design of the reforms neglects the welfare of families who do not qualify for social assistance, and results in minimal benefit levels insufficient to raise families above poverty lines. The reforms have reproduced the tax/benefit system’s fragmentation and the unequal levels of protection offered by the state, which block the formation of redistributive coalitions. The outcome is a system of low redistributive potential to reduce high levels of poverty and inequality. The thesis represents an original contribution at the empirical level, because it provides a comprehensive assessment of recent reforms in terms of the population segments favoured by them; at the methodological level, since it applies the micro-simulation logic to the study of social policy in the country; and at the theoretical level, because it characterises the welfare model that has emerged from the reform process.