ABSTRACT
A common criticism of pay-as-you-go systems of pensions is that such systems are, in the long-run, more costly than fully-funded pension systems. A basic fallacy in this argument is the failure to distinguish between projected costs and ultimate costs of a pension system. In the case of a fully-funded system, the two costs would be the same only if the assumptions used in the cost projection materialized; but in reality they never do, and, as our analysis in Part 1 showed, this means that some of the basic objectives of a dynamic pension will not be met.
