The Century Foundation
41 East 70th Street, New York, NY 10021
820 First Street, NE, Suite 510, Washington, DC 20002
Ph: 202-408-1080 Fax: 202-408-1080
Response to the Co-Chairs of the Commission
Henry J. Aaron, Alan S. Blinder, Alicia H. Munnell, and Peter R. Orszag
July 24, 2001
PDF of report
If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.
Yesterday, we issued a report analyzing various arguments presented in the draft interim report of the President's Commission to Strengthen Social Security.(1) In response to our analysis, the co-chairs of the Commission issued a statement claiming to answer our objections to the draft interim report. Unfortunately, the co-chairs' statement merely repeats many of the claims from the draft interim report itself. The statement thus shares the fundamental shortcoming of the draft interim report, in that various of its assertions are either unfounded or phrased in a way that could create misleading impressions. This note responds to the points raised in the co-chairs' statement.
Comparison Between 1983 and 2016
The co-chairs apparently view the projected condition of Social Security for 2016 as equivalent to Social Security's condition in 1983. They argue that "the crisis that we have described in 2016 is one in which the program's revenues from payroll and benefit taxation will be inadequate to cover annual program costs. The last year in which this was true was 1983."
This statement is problematic. The reason that Social Security faced a "crisis" in 1983 was that it was on the verge of being unable to pay full benefits. Its Trust Fund had reached alarmingly low levels and was continuing to decline. Indeed, to avoid exhaustion, the Trust Fund for the Old-Age and Survivors Insurance component of Social Security was forced to borrow from the Disability Insurance and Medicare Hospital Insurance Trust Funds in 1982. The situation is dramatically different in 2016: The Social Security Trust Fund will amount to more than $5 trillion in that year, and Social Security will be running a surplus of more than $250 billion (because total revenue, including interest on the Trust Fund's assets, is projected to exceed total costs in that year).
The co-chairs seem to be attempting to redefine the nature of the crisis in 1983 as being one in which annual program costs exceeded revenues from payroll and benefit taxation (i.e., one in which costs were higher than revenue if interest on the Trust Fund's assets is excluded). But that situation occurred in every year between 1973 and 1983; do the co-chairs believe that Social Security was in crisis in every year between 1973 and 1983?
In our analysis, we criticized a passage in the draft interim report that suggested that Social Security benefits end with the death of the beneficiary. In response to that suggestion, we noted that Social Security provides important benefits to heirs through survivors' benefits. In fact, $75 billion approximately 20 percent of total Social Security benefits were paid in 1999 to survivors of deceased workers. We also noted that these Social Security survivors' benefits are superior to the benefits that would be available under some systems of individual accounts, because Social Security survivors' benefits are fully protected against inflation and private annuities are not.
In their statement, the co-chairs argue that we unfairly quote from their introductory letter in the draft interim report while ignoring references to survivors' benefits in several places in the rest of the report. If the passage in question does not accurately reflect the views of the co-chairs, we would welcome a correction in the final version of the report.
The co-chairs' statement emphasizes the importance of "permanent wealth creation that can be passed from one generation to the next." As we explained in our analysis, using individual accounts to make bequests would inevitably reduce the monthly retirement checks that retirees would receive. You can't use the same dollars twice.
National Saving and the Social Security Trust Funds
The co-chairs repeat the assertion implicit in the draft interim report that individual accounts financed from existing Social Security revenue would add to national saving. But diverting money that is accumulating as part of Social Security reserves and is being used to pay down the national debt and that thereby increases national saving into individual accounts does not increase saving. The co-chairs are effectively arguing that if Social Security revenue were not diverted into individual accounts, it would not be saved but instead would be used to finance other programs.
This argument is entirely inconsistent with the strong bipartisan support for a lock-box for the Social Security surplus, which is designed to ensure that such surpluses are devoted to public debt reduction and therefore add to national saving. The co-chairs do not explain how diverting revenue from Social Security into individual accounts would raise national saving in the presence of the Congressionally approved lock-box for the Social Security surplus. Do they deny the existence of the lock-box or somehow doubt its efficacy? If so, they should make explicit their basis for assuming that the Social Security surplus will not contribute to national saving.
The co-chairs also argue that, historically, the partial funding of Social Security did not contribute to greater national saving because the government as a whole ran a budget deficit in the 1980s and early 1990s. This assertion ignores the fact that reducing a budget deficit adds as much to national saving as increasing a surplus. In other words, if the non-Social Security portion of the budget had a deficit of $300 billion in a given year, and Social Security ran a $100 billion surplus, the net deficit would be $100 billion smaller and national saving $100 billion higher than otherwise. The only way in which Social Security surpluses would fail to increase government saving is if Congress decided to increase spending or reduce taxes in the non-Social Security part of the budget because of the surplus in Social Security. There is little evidence that such offsets have occurred to any significant degree since the mid-1980s.
As a result, the co-chairs greatly underestimate the contribution of Social Security to national saving. But they also overestimate the contribution of individual accounts. As we argued in our analysis, the most plausible behavioral response to individual accounts - that individuals would reduce some of their other saving in response to the creation of individual accounts would cause a decline in national saving. In other words, the co-chairs have not explained why national saving should even remain constant given the diversion of Social Security revenue into individual accounts, let along why it should rise, especially in the presence of the lock-box on Social Security's cash-flow surpluses.
Finally, the co-chairs confuse the issue by citing a variety of non-partisan experts on a tangentially related issue. The non-partisan experts are cited as supporting the statement that "assets to the Social Security Trust Fund are obligations of the federal Treasury to an exactly offsetting degree." That statement is true, but irrelevant to the question of whether Social Security surpluses contribute to national saving. In other words, the statement is merely an accounting identity that is true regardless of whether Social Security surpluses, recorded by the Trust Fund, raise national saving and therefore better prepare the nation and the Federal government for our future obligations. The truth of the statement thus does not offer any support for the co-chairs' argument that Social Security surpluses fail to contribute to national saving.
The draft interim report claims that recent studies show that Social Security accomplishes little redistribution to lower earners. In our analysis, we criticized the commission's presentation on this important issue as being biased and incomplete. In particular, the draft interim report ignores at least two recent studies showing that Social Security is becoming more progressive over time, it fails to compare Social Security on this dimension to a likely individual account system, and it presents a misleading picture of Social Security's effects on minorities and women.
The co-chairs respond to this criticism by arguing that, "All of the research on redistribution cited by the Commission is footnoted in the report." Our criticism, however, was not that the draft interim report failed to provide footnotes regarding which papers had been cited. It was rather that the report ignores at least two recent papers from respected scholars showing different conclusions. It presents a misleading picture of the academic research on the issue.
In addition, the relevant question is not the absolute amount of redistribution accomplished by Social Security, but rather how much is accomplished by Social Security relative to a system of individual accounts. Such comparisons, which are missing from the draft interim report and the co-chairs' statement, are crucial to evaluating the likely effects of individual accounts. Such analysis would show that individual account proposals generally are less progressive than Social Security.
The draft interim report argues that "African Americans receive nearly $21,000 less on a lifetime basis from Social Security's retirement program than whites with similar income and marital status." In our analysis, we argued that the statement is misleading because it compares an African-American and white American at the same earnings level, even though African-Americans tend, on average, to have lower earnings and therefore to benefit disproportionately from Social Security's progressive benefit formula. The co-chairs respond by arguing that the purpose "was to take two individuals in the same income bracket, with the same work history, and show the discrepancy that results due to the different life expectancy of whites and African Americans. Failing to control for income would only obscure this discrepancy."
This example provides a good illustration of items in the draft interim report that could easily create misimpressions among the broader public. African-Americans differ, on average, from whites in two relevant ways: They have shorter life expectancies and lower earnings.(2) Highlighting an example in which one relevant attribute, but not the other, is included is misleading. We could just as easily, for example, construct an example in which we compare an African-American and a white American who die at the same age (thus excluding the effects of the lower life expectancies among African-Americans). Since the African-American, on average, would have had lower earnings, and since Social Security's benefit formula is progressive, such a comparison would show that the African-American gains substantially more from Social Security than a white with the same life expectancy. Such an example would be misleading, however, because it would exclude one relevant factor from the calculation. This type of misleading approach is what the draft interim report employs.
If the Commission wished to examine the individual effects of life expectancy and lower earnings, it could have presented an example in which each effect was explored separately and then an overall result was provided. This is precisely what most academic studies do. Indeed, the very study cited on this matter by the Commission finds that when both life expectancies and earnings are taken into account, the average rate of return on Social Security is modestly higher for African-Americans than for whites and much higher for Hispanics than for whites. The draft Commission report omits this finding.
Finally, it is worth noting that an individual account system in which accumulated balances were transformed into an annuity upon retirement would produce precisely the same effect the Commission report notes for Social Security for those at the same income level, smaller lifetime benefits for those with shorter life expectancies. (Requiring annuities under individual accounts is necessary to ensure that retirees do not outlive their savings, which is why most individual account plans include such a requirement. Social Security also provides an annuity, but it includes a feature not currently available on private annuities: it fully protects benefits from inflation.) Thus, under individual accounts, African-Americans at the same level of income as whites would receive lower lifetime benefits from such accounts, on average, because they would receive their retirement income for fewer years. Furthermore, African-Americans, on average, would not be as well off under most individual account plans as under Social Security. The reason is that under most such plans, the lower average earnings for African-Americans would not result in relatively higher annual benefits, as occurs under Social Security's progressive benefit formula. On average, therefore, African-Americans would receive a smaller annual benefit for the same number of years as under Social Security (which would be fewer years, on average, than a white with the same level of income). The average lifetime benefit would thus be smaller than under Social Security.
The fundamental problem here, too, is that the commission document fails to undertake the crucial comparison: How do the results under Social Security compare to those under individual accounts?
The co-chairs' response to our analysis does not address the questions we raised about the draft interim report. The criticisms we issued in our analysis stand.
1. Henry J. Aaron, Alan S. Blinder, Alicia H. Munnell, and Peter R. Orszag, "Perspectives on the Draft Interim Report of the President's Commission to Strengthen Social Security,"Center on Budget and Policy Priorities and the Century Foundation, July 23, 2001.
2. African-Americans also have higher disability rates than whites, which produces disproportionate benefits for African-Americans under the Disability Insurance program. Those benefits are not included in the example cited by the Commission.
Search | Publication Library | Job Opportunities | Internship Information
To join the Center's e-mail notification list, ask questions,
or send comments, write to [email protected]
Center on Budget and Policy Priorities
820 First Street, NE, Suite 510
Washington, DC 20002
Ph: (202) 408-1080
Fax: (202) 408-1056