Social Security
Reasons for altruism?

In many countries social security accounts for a substantial proportion of the government budget, but the number of recipients of social security benefits at any given time is smaller than the number of contributors.
In Discussion Paper No. 394, Research Fellow Guido Tabellini presents a simple theoretical model that seeks to explain why a large majority of citizens are prepared to support a system that redistributes to a minority, and how the size of social security provision is determined.
A social security programme redistributes both across and within generations, and the contributions to the system are linked to wage income, but the benefits are not, so the system also redistributes from high to low income households. Hence poor workers/taxpayers may be expected to favour the programme, since the gain to their retired parents is larger than the cost to themselves.
Tabellini presents a simple political-economic equilibrium model in which the social security programme is chosen in each period by a majority vote of all citizens currently alive, who are divided into two generations: `old' and `young'. Thus the social security legislation can be repealed in any period, and the link between current contributions and future benefits is broken. In any period, a vote on the social security programme determines how much to transfer from the current young to the current old, with no repercussions on future legislation. The model allows a degree of reciprocal altruism within families, so that parents care about their children, and children care about their parents. This altruism is relatively weak, however, so that no private transfers occur in equilibrium, and altruism plays a role only in the political equilibrium. Young voters trade off the tax burden they incur in financing the programme against the benefits received by their parents.
Tabellini's main analytical result is that, if there is a sufficient degree of inequality in labour income, a social security programme will receive the support of a majority of the voters in equilibrium. His analysis also implies that the extent of social security provision will increase both with pre-tax income inequality and with the proportion of `old' people in the population. In the empirical part of the paper he compares the size of the social security programmes in 63 developing and industrialized countries, and finds that cross-country differences in the size of social security are explained well by the inequality of pre-tax income and the age composition of the population, as predicted by the theory. This finding is robust to alternative specifications and to the possibility of measurement error in the explanatory variables.

A Positive Theory of Social Security
Guido Tabellini

Discussion Paper No. 394, April 1990 (IM)