Méthodes et usages des comptes générationnels : un regard décalé

This paper presents a critical evaluation of generational accounting (denoted CG in French), as it has been initiated by Kotlikoff and his colleagues. Kotlikoff's great error, which has been the cause of multiple misunderstandings, is to have used CG results both in order to estimate the long term unbalance of current public policy and to measure resulting actual inequalities between generations, thus confusing indicators of sustainability and measures of generational equity. We show that these two objectives cannot be simultaneously fulfilled : the first one, which is the purpose of CG, requires a thought experiment made in a static and purely accounting framework, and uses a cost or a cash-received basis (for the government) ; the second one needs a more ambitious dynamic framework, using a utility-based approach (for the consumer). However, both types of information appear needed and complementary for governement decisions. In that perspective, Kotlikoff's great achievement comes from the fact that CG, properly amended, appears indeed the appropriate tool in order to evaluate the overall sustainability of public transfer policies while incorporating governement commitments involving young and future generations. The method could still be improved in three principal directions : (i) treat educational expenses as individual investments in beneficiary's human capital ; (ii) simulate current social legislation instead of extrapolating per capital transfers by age ; (iii) rely for the future on a structural trend of current policies, while "eliminating" most of their short-term or cohort specific elements.