Mitchell, Olivia S. - Pension Research Council

Olivia S. Mitchell [1]

Returns and Administration Costs under Alternative Social Security Reforms

It is important to be clear about three dimensions of choice in the discussion over US social security reform. The first dimension pertains to finding: prefunding social security requires increasing taxes or cutting benefits to shrink the system's current $9 trillion in unfunded promises (amounting to a social security debt of $60,000 per US worker). The second dimension involves investment portfolios: diversifying social security involves investing all or part of the retirement system's assets - totaling 1.5 years of benefit payments - in a broad range of investments including corporate stocks and bonds. (Currently, all social security funds are invested in Treasury bills.) And the third dimension pertains to system type. One could remain with the current national defined-benefit system, or one could convert to individual defined-contribution accounts. All reform plans pick points along these three dimensions: for instance, the recent Social Security Advisory Council put forth two plans (the Personal Security Account and the Individual Account) requiring funded defined-contribution individual accounts with diversified assets, while the Maintain Benefits plan kept the current federally-held defined-benefit system but raised prefunding along with Trust Fund investment in equities.
The issues now polarizing the US Social Security reform debate can be seen in the light of these choices that must be made, regarding rates of return and administration costs under a reformed system. As we show, these issues are more complex than has been recognized to date.
Investment Returns in a Diversified System [2]
Some who favor diversification of social security assets offer as supporting evidence the high stock market returns of the last 20 years, and they conclude that Baby Boomers and their children could do better by investing in stocks than by remaining under the current system. But the argument that workers could earn a higher "rate of return" from investing their payroll taxes in the stock market is false for two reasons. First, returns on stocks vary more than bond returns, making investors demand a risk premium to hold these more uncertain assets. Proper risk adjustment of stock returns would render these more similar to returns on less risky assets. Second, today's workers have inherited $9 trillion in unfunded social security promises, as noted above. If this debt is to be honored, some 3% of current workers' pay would have to be allocated to cover retirees' benefits - leaving only three-quarters of the social security payroll tax to be invested in the stock market. (In fact, the potential amount is even less than that since social security disability insurance benefits exceed earmarked payroll taxes as well). Evidently, the "transition cost" of old unfunded promises undermines one of the apparent advantages of allowing system diversification, whether it is through individual or government-managed accounts.
Despite this caveat, important benefits would flow from prefunding individual social security accounts that can be invested in a diversified set of assets. Specifically, about half of the US population holds no stock in its retirement portfolio. Many of these people would be better off from increased retirement saving in diversified assets, and individual accounts would help achieve this. Some gains might result if the federal government held the assets in the Trust Fund, but many have deep concerns about the government "picking stocks" and worry that the Trust Fund monies would be spent rather than saved.
Administrative Expenses of an Individual Account System [3]
Even among those who favor funded, diversified, social security accounts, people disagree about how these accounts should be managed. If the federal government invested the money centrally via a pooled Trust Fund, the additional administrative costs of money management would be modest - probably under 10 basis points per year (a basis point is one hundredth of a percent). These would be incurred on top of current social security administrative costs of about $14 per participant per year. If individual accounts are mandated instead, additional administrative costs might come to lo-50 basis points per year - on the lower end for accounts receiving contributions worth 5% of payroll, and on the higher end for accounts receiving only 2% of payroll per year. Some analysts restate these annual charges over the participant's worklife as equivalent to a fraction of the account's ultimate asset value, and they arrive at numbers ranging from 1% to 20% of the final account depending on the holding period, the rate of return, and the size of the account.
It is evident, therefore, that participants in an individual account plan must devote substantial attention to how administrative charges can affect retirement saving. In exchange for these fees, individual account participants would have the opportunity to choose their fund managers and investment portfolios, and people would reap the reward from saving and managing their own accounts. These options are unavailable under a national defined-benefit approach. The US can learn from other nations that have preceded us in moving to individual retirement saving accounts, including several sister nations in the Americas. The US economy would probably do better on administrative costs because of our larger size, our more sophisticated taxation system, our technologically advanced business and labor environment, and our more efficient and transparent capital market. In any event, additional regulatory structures would be required.
Drawing these distinctions is important since they lead to several conclusions about social security reform proposals offered in the US context:

Any reform plan boosting prefunding requires additional taxes and/or lower benefits.

Any reform permitting asset diversification into stocks will involve additional risk.

Any prefunded system will require regulation to effectively resist political interference.

An individual account system will cost more to administer but will offer additional services.

1 Mitchell is the International Foundation of Employee Benefit Plans Professor of Insurance & Risk Management and Executive Director of the Pension Research Council at the Wharton School, University of Pennsylvania. [email: mitchelo@,wharton.upenn.edu] Opinions are solely those of the author.
2 See http://prc.wharton.upenn.edu/prc/prc.html or Geanokoplos, J., O.S. Mitchell, & S. Zeldes. "Social Security Money's Worth,". In Prospects for Social Security Reform, Eds. O. Mitchell, R. Myers, H. Young. Univ. of Pennsylvania Press, 1999; and Geanokoplos, J., O.S. Mitchell, & S. Zeldes. "Would a Privatized Social Security System Really Pay a Higher Rate of Return?" In Framing the Social Security Debate. Eds. D. Arnold, M. Graetz, & A. Munnell. Brookings Institution, 1998.
3 See Mitchell, Olivia S. "Administrative Costs of Public and Private Pension Plans". In Privatizing Social Security, Ed. Martin Feldstein. University of Chicago Press, 1998.