Small assumptions (can) have a large bearing: evaluating pension system reforms with OLG models

The objective of this paper is to inquire the consequences of some simplifying assumptions typically made in the overlapping generations (OLG) models of pension systems and pension system reforms. This literature is largely driven by policy motivations and the alternative modelling choices are rarely discussed. On the other hand, the complexity of general equilibrium OLG modelling necessitates some simplifications in the model. We run a series of experiments in which the same reform in the same economy is modelled with six different sets of assumptions concerning the functional form of the utility function, time inconsistency, bequests' redistribution, labour supply decisions and internalization of the linkage between social security contributions and benefits in these decisions as well as public spending. We find that these assumptions significantly affect both the size and the sign of the macroeconomic and welfare measures of the policy effects with the order of magnitude comparable to the reform itself.