Comments and Questions

This paper introduces a model of a rather complex social security system. The government has three policy instruments at its discretion: the contribution rate, the benefit level and a parameter with which it can decide whether contributions are funneled into a PAYGO or a funded scheme (or a mixture of both). The papers investigates how changes of the three parameters affect the two skill groups (the high-skilled and the low-skilled) under consideration. In a first step, this is done for a setting with inelastic labor supply, then elastic labor supply is assumed. The paper appears to be more work-in-progress than a full-fledged, polished paper yet. In the beginning I got somewhat misled by the author. It appeared to me that the paper would deal with the transition from a PAYGO to a funded system through (parametric) pension reforms. However, the (anyway superficial) literature review does not mention a single relevant paper from this field (e.g., Aaron 1966, Fenge 1995, Sinn 2000) and later in the paper nothing like this is discussed. The abstract suggests further a welfare-theoretic analysis, possibly dealing with intergenerational justice issues (reminiscent of the Auerbach-Kotlikoff intergenerational accounting literature). Ultimately, the paper provides insights on how two different groups in society will be affected by a pension reform, but since the groups are affected differently, one would like to know something about the total effect or about which type of reform is being selected in a majority vote (in the specific context of this paper, the literature by, e.g., Casamatta et al. 2000a,b, Conde-Ruiz/Profeta 2007, Krieger/Lange 2012, could be helpful). But nothing is said about all this. To me, this is the main weakness of the paper. It does not have a clear research focus that relates to the literature. My second major concern relates to the relevance of the model. Of course, one could argue that the proposed model is a generalization of a whole class of social-security models (the author does not claim this, though), but when working on a topic like this, what I would like to see that the model corresponds closely to the real world. However, the author does not make any efforts to explain where in the real world we observe the described pension system. For instance, supplemental benefits are typically financed by taxes, while basic pensions come from the pension system itself. That makes the supplemental benefits a policy variable and not a residual variable in the model. This implies in the current model that the lack of an upper bound (or an objective or justification of the supplement), a majority may use this variable strategically to redistribute money in its favor. This reasoning raises yet another problematic aspect of the model: it is not a model about pensions, but about some arbitrary redistributive system. Except for section 4.1, the relation of r and n does not play any role here. That is, intergenerational redistribution is virtually absent in the model (interestingly, the term ‘dynamic efficiency/inefficiency’ or Aaron 1966 is not mentioned). Almost everything is about intragenerational redistribution (that is why having a look at Casamatta et al. 2000 might be insightful). There is also no intertemporal budget constraint in the model; only by assumption shifting the financial burden of today’s generous benefits into the future is excluded, although it would generate a win-win situation for both today’s high-skilled and low-skilled workers.