Essays on pension reform, its impact on growth and its redistributive effects.

Summary This thesis empirically analyzes the macroeconomic effects of different PAYG pension reforms and their redistributive consequences. In a general equilibrium model with overlapping generations, the criterion of the effect on growth makes it possible to rule out the choice of an increase in social contributions as the exclusive lever of the reform. However, it does not allow us to distinguish, particularly in France, between scenarios as different as an increase in the effective retirement age linked to life expectancy or a decrease in the replacement rate with an unchanged retirement age, each of which would make it possible to support the future average annual growth rate of GDP per capita by about 0.2% compared to the scenario of an increase in compulsory contributions. Taking into account the redistributive consequences becomes decisive for the choice of reform. A freeze in the retirement age fuels an intergenerational conflict between working people, whereas an increase in the retirement age distributes the welfare cost of the reform more evenly among working people. A utilitarian planner with a moderate aversion to intergenerational inequality thus favors the effective retirement age as the main lever of a reform. Reforms with an increase in the retirement age also moderate the redistribution effects according to income level compared to reforms with an unchanged retirement age. Finally, taking into account the effects of reforms on the welfare of low-income earners (incomplete careers, eligibility for the minimum old-age pension) may influence the choice of adjustment parameter (social security contribution rate or replacement rate) accompanying an increase in the retirement age. Overall, this thesis suggests that France should increase the retirement age by a little more than one year per decade, with a contained decline in the average replacement rate of 0.5 percentage point per year over the next twenty years.