Any long-run contract will involve risks. Under the current Social Security system, there are risks that arise from the fact that the system is not in balance¯it will run out of money. Will taxes be raised? If so, on whom? Will benefits be reduced? Whose benefits? Under a fully funded, defined benefit system of the sort I support, there will also be risks. If the economy does not perform as expected, the return on the collective investment of reserves comes up short, or the demographic assumptions prove to be wrong, adjustments will have to be made to keep the defined benefit system afloat. But those adjustments will be worked out collectively through our political system by our elected representatives. As a result, the adjustments, most likely, will phased in gradually and shared broadly by tax payers and beneficiaries¯that is what has happened in the past.
Under a system of individual accounts, the adjustments will fall on the individuals. If financial markets perform poorly, unanticipated inflation erodes the value of ones account balance, or the worker or the worker¢s spouse lives an unexpectedly long life, the individual could end up with inadequate retirement income. These risks can be reduced by restricting account investments to balanced index funds of stocks and bonds (as the plan put forward by Ways and Means Committee Chairman Bill Archer and Subcommittee Chairman Clay Shaw would do) and by requiring that account balances be used to purchase inflation indexed joint and survivor annuities of at least a minimum size. But the risks can not be reduced to a level that I find acceptable for the program that is intended to provide the foundation for retirement income. (It is worth noting that people accept the uncertainty inherent in arrangements like IRAs and 401(k)s, in part, because they also expect to receive an inflation protected basic pension from Social Security.)
If history is any guide, individual accounts, even if invested in the same assets, will generate highly variable pensions for cohorts only a few years different in age. Dr. Gary Burtless, my colleague at Brookings, has calculated the variation in replacement rates (pension as a percent of earnings) that different cohorts of male workers would have received had they contributed 6 percent of their earnings to a total stock market index fund from the time they were age 20 until they reached age 62. He assumed that these accounts would not bear any administrative costs or pay taxes, that all the earnings would be reinvested, and that at age 62 the account balances could be converted without cost into annuities at prevailing interest rates. Under these assumptions, the replacement rate for those turning 62 in 1969-70 would have been close to 100 percent; for those turning 62 in the mid 1970s only 40 percent. Variations of this magnitude would be neither desirable nor politically sustainable. While this is an admittedly extreme example, it illustrates the nature of the risk inherent in a system of individual accounts.
