P resident Bush has appealed to African-American voters by arguing that they would benefit from the replacement of part of the traditional Social Security retirement income system with private retirement accounts. The argument goes like this: The average 50-year-old African American has a life expectancy of 27.3 years, as compared to an average of 30.5 years for Caucasians. This means that African Americans receive retirement benefits for fewer years and are more likely to die before receiving any benefits at all. In a system of private retirement accounts, African Americans could bequeath their accounts to their heirs, thereby closing some of the “black-white” wealth gap and ensuring that blacks no longer get a bad deal from Social Security. There are good reasons for introducing private retirement accounts, such as increasing national savings and improving labor supply incentives. But achieving racial parity in Social Security benefits is not one of them. Blacks benefit disproportionately from other features of Social Security. Because they have below-average lifetime earnings, they are helped by the Social Security benefit formula that replaces a larger fraction of pre-retirement earnings for low earners than for high earners. For the retirement portion of Social Security, these two effects almost exactly cancel, meaning that blacks receive rates of return on their Social Security payroll tax payments at least equal to those for whites. When benefits for disability and young survivors are included, blacks clearly receive an above-average return from the current Social Security system. Because Social Security benefits are protected from inflation and last as long as you live, they are especially valuable to low-income groups that count on them for a large portion of retirement income. Thus, changing Social Security may in fact harm, rather than help, blacks. ![Figure][1] CREDIT: J. SCOTT APPLEWHITE/AP It is hard to judge how different racial and ethnic groups would fare under an approach that has not yet been fully defined, but there are reasons for worry. First, the president proposes to divert payroll tax revenue away from the current system. That raises the chance that disability and survivor benefits will be cut, disproportionately affecting blacks. Second, there might not be any redistribution from high earners to low earners in the privatized portion of the reformed system, meaning that gains to blacks from the bequeathability of accounts will be at least partially offset. Third, any sensible privatization plan will require retirees to convert their account balance into annuities upon retirement; because members of long-lived groups will receive annuity payments for more years, there will continue to be redistribution from groups with low life expectancy to groups with high life expectancy. Fourth, the introduction of market risk into Social Security will be more of a burden for low-income groups such as blacks. One could design a system with privatized accounts in which blacks do fine: by maintaining disability benefit levels, enhancing redistribution from high earners to low earners, and requiring that annuities provide payments to the retired worker or his or her heirs for a minimum of 10 years. Blacks could indeed be winners from such a reform. But these features are unrelated to privatization; all could be accomplished within the traditional Social Security system. If the real objective is to help demographic groups with short life expectancy, tinkering with Social Security rules may not be the best approach. Instead, investing in the health of low-income populations with early mortality could produce greater benefits in the short run and help to close the longevity gap in the long run. Under the present system, if black and white longevities were equal, blacks would receive about $4 billion of additional Social Security benefits per year. That's less than 2% of spending on Medicaid, the nation's medical assistance program for low-income individuals. Even if that sum were sensibly invested in expanding access to health care, improving community health centers, increasing research expenditures on racial differences in disease, and so on, it is doubtful whether longevity differences could be eliminated. That's because the differences are large, and the things that are likely to have the biggest impact—changing diet and exercise habits—are hard to alter with government policy. Thus, investing in the health of low-income populations may well be the best way to spend money on behalf of groups with short life spans. But we should not kid ourselves into believing that investments of this size will erase racial differences in health outcomes. And the president should not kid African Americans that Social Security privatization will make them richer. [1]: pending:yes