August 1, 1999
Randall P. Mariger
Social Security privatization is now receiving much attention in policy circles. Proposed privatization plans range from modest changes in the mix of assets held by the Social Security trust fund to the phased-in replacement of all Social Security benefits and taxes with mandatory Individual Retirement Accounts (IRAs).
Much of the current interest in privatization stems from the realization that any reform of Social Security that is financially feasible and that maintains the program’s current structure would not be a good deal for current and future workers. These workers would receive benefits that represent lower than market rates of return on taxes paid. On the surface, therefore, it would seem that Social Security can be made more valuable by somehow tapping higher market rates of return. This simple reasoning suggests that privatization offers the possibility of a “free lunch” where nearly everyone is made better off.
A major objective of the paper is to demonstrate that the free lunch arguments for privatization are false. My purpose, however, is not to argue against (or for) privatization. My purpose is to help redirect the privatization debate toward what I regard as the pertinent issues. These issues concern possible political impediments to prefunding retirement incomes in government accounts--which would argue that prefunding can only take place in private accounts--and on the implications of privatization for how capital income risk is shared by individuals.
The paper begins by analyzing the current Social Security policy dilemma in the context of a simple model in which people live for two periods. This model helps elucidate an important truth: Social Security has already accumulated an implicit debt that cannot be fully revoked. Hence, any Social Security reform that is financially sound in the long run effectively levies a net tax on current and future workers to service that debt. This accounts for the below-market rates of return that Social Security can promise to current and future workers.
I then demonstrate that substituting mandatory IRAs for traditional Social Security benefits and taxes would not relieve Social Security’s implicit debt burden. This is done by showing that retirement systems based either on mandatory IRAs or on traditional publicly administered defined benefits can be designed so as to have similar implications for national saving and for the adequacy of overall retirement incomes.
The paper then turns to two issues that truly distinguish retirement systems that include mandatory IRAs from those that do not: the distribution of capital income risk, and the possibility that prefunding of retirement consumption can occur in private accounts but not in public accounts. I show that introducing mandatory IRAs would enable individuals who are currently liquidity-constrained to take on a larger share of capital income risk. With regard to prefunding, I argue on both efficiency and equity grounds that substantial prefunding is desirable, and that there is a very real possibility that prefunding can only occur in private accounts. I also point out that mandatory IRAs would most effectively prefund retirement consumption if investments were limited to one centrally managed investment fund so as to minimize administrative costs.
Finally, the paper considers arguments for investing all or part of the Social Security trust fund in equities. Again, the free lunch argument is rejected: I show that trust fund equity investments can increase the generosity of Social Security only to the extent that they reduce the returns earned on the private sector’s portfolio. Also analyzed is the possibility that trust fund equity investments would lead to a more efficient allocation of risk bearing. While this outcome cannot be ruled out, I conclude that undesirable outcomes are more likely.