Research

Working papers

How should pension benefits be linked to pension contributions? (Job Market Paper)

I compare pay-as-you-go pension arrangements that link a person's pension benefits to her labor earnings history (notional defined contribution (NDC) and defined benefit systems (DB)) with those that do not (flat benefit systems (FL)) in terms of long run macroeconomic outcomes and welfare. I show that, because under the FL system pension benefits do not reflect previous earnings and, hence, the labor supplied by a person and the earnings shocks received while working, this system promotes the lowest labor supply, but also the lowest consumption inequality and generally leads to the highest capital accumulation. The NDC and DB systems are similar in terms of macroeconomic outcomes. Using a model calibrated on the US economy, I obtain that welfare is the highest in the steady state with the FL pension system: the better insurance provided by the system and the higher capital accumulation dominate the higher labor supply distortions. Considering a higher pension contribution rate - such as the one of the Italian pension system - the FL system still provides the highest welfare. However, when shutting down the impact of higher capital accumulation on welfare by assuming a partial equilibrium setting, the NDC system provides the highest welfare.

Job loss risk and retirement (with Iulian Ciobica)

In a life cycle model with permanent earnings losses following a job loss shock and endogeneous retirement, we show that higher job loss risk can lower the retirement age of workers. We prove that the interaction between the job loss shock and the contribution to a pay-as-you-go pension system is crucial in obtaining a negative relation between job loss risk and the retirement age of a worker. The negative impact of job loss risk on retirement age is sustained by cross-country data. We document a statistically significant negative relationship between the incidence of long-term unemployment, on the one hand, and the participation rate of people aged 55-64 years and the effective retirement age, on the other hand.

The political arrangement of pay-as-you-go pension systems in the presence of financial and demographic shocks (with Ward Romp)

We analyze the outcome of voting over the contribution to a pay-as-you-go (PAYG) pension system in the presence of financial and demographic shocks. The impact of shocks on pension contributions and benefits replicates major developments of pension systems around the world. A decrease in the return on capital increases contributions and benefits, while a decrease in the population growth rate increases contributions but it lowers benefits. The first and second moments of the demographic shock matter: in the case of a lower mean or a higher variance, a smaller average contribution results from the vote. In contrast to this, the Ramsey planner sets a higher average contribution when the population growth rate has a lower mean or a higher variance.