The Case for Extending Social Security to Government Employees. Comment

Abstract

Mickey Levy has contributed very worthwhile paper on the desirability of covering federal, state, and local government employees under Social Security in the Journal .' I am in complete agreement with his general viewpoint (and also with his views on the somewhat unrelated matter of the undesirability of financing the Social Security program in part from general revenues). However, I believe that he has placed too much emphasis on the windfalls that govemment employees have received from short periods of Social Security coverage after or during their government service in order to obtain the minimum Social Security benefit. The value of that benefit will diminish over time, because the 1977 Act froze it at $122 per month at the initial time of entering the roll. But even in the past, the minimum benefit was not the most important element in windfall benefits, but rather the availability of heavily weighted benefit for short service at moderate or high earnings. It did not, in the past, require much earnings to obtain more than the minimum benefit, and this will be even more the case in the future. In the absence of universal coverage, these windfall benefits can be ameliorated, at least for future service, by change in the benefit computation made by Social Security (for several methods of accomplishing this, see Robert J. Myers, An Alternative Approach to Universal Social Security Coverage, Transactions, Society of Actuaries, Vol. 31, 1979). Levy suggests method of transition for covering existing government employees under Social Security, by using a weighted average of predesignated rates of return that would have occurred under each pension program , based on the proportion of their contributions to each system. This might be fine in theory, but it is utterly impossible to do in practice. Rates of return are by no means precise elements, but rather, they differ widely depending upon the assumptions made. Finally, Levy states (in footnote 15) that the cost to the federal government of the Civil Service Retirement system is about 171/2 percent of payroll. Actually, the situation is quite different from this. The latest actuarial valuation showed that the level cost to the government was about 33 percent of payroll. However, if CSR were financed by the same standards applicable to private pension plans under ERISA, the annual employer cost in 1980 would be 79.8 percent of payroll (see Federal Pension Programs, Working Papers, President's Commission on Pension Policy, January 1981).