The aim of this project is to investigate the determinants of the pension system's coverage in Chile, Colombia and Mexico. These 3 countries deeply reformed their pensions systems, moving from a traditional PAYG system to a Fully Funded one (although there are some differences in the final arrangements across countries). As possible factors that affect participation in the pension systems we will look at the importance of the business cycle, the design and incentives of the pension systems, the relationship with the informal labour market and alternative sources of old-age funding available to individuals. The project will complement the analysis of the design of the pension systems and of aggregated time series descriptive statistics with a multivariate probabilistic model. For the latter we use panel data in 2 out of the 3 countries, which allow us to control for unobserved heterogeneity that may affect the outcome variables.