To What Extent are Public Pensions Pareto-Improving: On the Interaction of Means Tested Basic Income and Public Pensions

Abstract

If there is a means tested basic income for old age, households will tend to reduce precautionary savings to an inefficiently low level. This might serve as a justification for a public pension system. In a representative agent framework, indeed, the introduction of a compulsory pension system is shown to be Pareto improving. This analysis is extended to two income types where compulsory savings are found to be Pareto improving only up to a point. Increases in contribution rates beyond that point simply result in increasingly regressive (implicit) taxation, potentially eliminating all redistribution via the means tested basic income. Using these results in a pay-as-you-go framework, we show that an unfunded pensions system (with intragenerational fairness) plays a role similar to compulsory savings in preventing the savings moral hazard and could have the same adverse effects on redistribution if it is too large. If the population is aging, however, an unfunded system with a constant contribution rate is found to become less effective at preventing the savings moral hazard. In this case, the introduction of a funded system of the right size is needed to restore Pareto efficiency.