Robert L. Clark[*]
Determining the financial status of the Old Age Survivors and Disability Insurance program (OASDI or Social Security) is complex and difficult. Understanding the current and future financial status of Social Security, however, is necessary for modifying and reforming national retirement policies. A number of methods illustrate the current and projected future actuarial balance of Social Security. In his interesting article,[1] Professor Jackson describes some of the methods, selects his favorite, and argues that it is the best method of presenting the financial status of Social Security. He makes a series of important observations, but also ignores some key issues associated with unbiased reporting of the financial status of Social Security. This Comment begins by describing the range of possible financial measures of Social Security’s health and examining the role of each measure. The next Part presents a basic critique of Jackson’s article and its analysis of the 2003 Trustees Report of the OASDI Trust Funds.
Policymakers, the press, and the public need to know the current and projected future financial status of Social Security. Optimal retirement policies should be based on quality research, unbiased projections, and sound accounting practices. It is extremely important to know:
(1) whether the Social Security system is actuarially solvent using current benefit formulas and tax rates and what solvency means,
*** Top of Page 162 ***
(2) what level of taxation is needed to pay promised benefits if the basic structure of Social Security is maintained, or, alternatively, how much benefits would need to be reduced so current tax rates would provide sufficient revenues,
(3) whether the trust funds are increasing or decreasing in value and when these funds will be completely depleted, and
(4) what is the magnitude of implied liabilities accrued by workers to date and how much additional money would be needed to pay off all promised benefits if OASDI were terminated today.
Traditionally, the Trustees of OASDI have focused on determining the seventy-five-year actuarial balance of the system and the size and trend of assets in the Trust Funds.[2] Jackson asserts that accrual accounting of promised benefits based on participation to date is the best method of summarizing the current financial status of Social Security.[3] In essence, he argues that the best accounting method for OASDI is based on the termination of Social Security and the needed revenues to fund (over and above existing assets held in the Trust Funds) the present value of all promised benefits. Another method of determining the present value of Social Security net liabilities is to assume that OASDI will continue to exist, to determine the present value of paying all present and future benefits, and finally, to compare this to the existing Trust Funds and the present value of all future scheduled tax collections. This method assumes Social Security will be maintained and tries to determine the magnitude of any projected shortfall in revenues.
Jackson aims his criticism at the 2003 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (hereinafter, “2003 Trustees Report”) and its measures of financial status.[4] Early drafts of his paper were written prior to the publication of the report. The report incorporates numerous changes in the evaluation and measurement of the financial status of Social Security.[5] The 2003 Trustees Report is a significant improvement over previous reports, and as such, sharply reduces the relevance of many of
*** Top of Page 163 ***
Jackson’s criticisms. Jackson’s paper would have more relevant policy implications if it acknowledged the material presented in the report and recognized the important changes it has made in accounting for Social Security. In particular, he should compare his preferred method of accrual accounting to the present-value calculations in the report rather than devote so much attention to the methods that dominated the pre-2003 reports.
Prior to 2003, the Trustees Reports focused almost exclusively on three concepts of the financial status of Social Security: the actuarial balance, the year that the Trust Funds will be depleted, and the change in the size of the Trust Funds. The actuarial balance concept is a useful indicator of the financial status of an ongoing Social Security system. It indicates the amount payroll taxes would need to be raised today (and continue at this higher rate over the next seventy-five years) to fund the system during this period. The pre-2003 Trustees Reports gave this measure by far the most attention. The 2002 Trustees Report indicated that the actuarial balance over the seventy-five-year projection period was –1.87% of taxable payroll.[6] This measure shows Congress how much additional revenue as a percent of payroll is needed over a seventy-five-year period if benefits are held constant. The 2003 Report showed an actuarial balance of –1.92% of taxable payroll, indicating that the financial status of the system had deteriorated slightly.[7]
The year that the Trust Funds have zero assets is another important measure of the financial status of OASDI. Law requires that Social Security benefits be paid from revenues generated by the payroll tax or from monies from the Trust Funds.[8] If the balance in the Trust Funds is zero and payroll tax revenues are insufficient then full benefits cannot be paid. The imminent exhaustion of the Trust Funds was a looming crisis in the early 1980s.[9] The immediate prospect of insufficient funds stimulated Congress to enact substantial changes in OASDI proposed by the Greenspan Commission in 1983.[10] Thus, the date of exhaustion of all assets in the Trust Funds has important policy implications and may affect reform decisions. The 2003 Trustees Report estimates that the OASDI Trust Funds will be exhausted in 2042.[11]
The annual change in the value of the Trust Funds also provides important information that indicates whether the Funds are increasing or
*** Top of Page 164 ***
decreasing in assets. In 2002, the Trust Funds added $165.4 billion in assets.[12] If viewed in isolation, this increase in value does give the impression of a system that is financially secure. Jackson is concerned especially that this measure can be misleading.[13] He is correct on this point, and thus it is imperative that other measures be included in any evaluation of the financial status of Social Security.
Although each of these items is important to understanding the financial status of Social Security, considered alone they present an incomplete and somewhat misleading picture. The actuarial balance concept focuses attention almost exclusively on the need for tax increases. While this is one possible response to the current underfunding of OASDI, policy makers might decide to reduce future benefits instead of raising taxes. Additionally, the seventy-five-year actuarial balance concept does not provide any information on either the amount of money needed if the system were to be terminated and all accrued liabilities paid or the amount of money needed from general revenue funds to cover the shortfall in revenues. Neither does the seventy-five-year actuarial balance provide the necessary information to assess the merits of reform proposals that require upfront transitional costs and contain positive revenue effects outside the projection period.
It is against this backdrop that Jackson developed his analysis. Many of his criticisms apply most directly to the pre-2003 Trustees Reports. In particular, he argues that describing changes in the Trust Funds presents a misleading picture when there are short-term annual improvements but large long-term annual deficits.[14] Additionally, he criticizes the limitation of a fixed-term projection period and identifies the need for an indicator of the present value of outstanding liabilities.[15] The world of Social Security accounting changed significantly in the last year, and unfortunately, Jackson focuses too much attention on past practices without giving sufficient attention to the new accounting methods adopted by the Trustees in the 2003 report.
Jackson argues that accrual accounting is the best and most informative method of illustrating the financial status of Social Security. Simply stated, under the accrual accounting method proposed by Jackson the current financial status of OASDI is the present value of all accrued benefits by workers and retirees minus the assets in the trust funds. This is the amount of additional money that would be needed if Social Security were
*** Top of Page 165 ***
terminated today, no future taxes were paid, no additional benefits were earned, and all promised benefits based on past taxes were paid. It is also the maximum transition cost of changing to a fully funded system or the adoption of a system of individual accounts in which the government transfers sufficient assets to cover all accrued benefits earned to date.
I agree with Jackson that the accrual accounting measure of benefit liability is an important and useful indicator of the financial status of Social Security. Annual changes in this measure are another significant indicator of the financial status of the system. The fundamental question concerning accounting for Social Security is whether this is the only measure that should be reported or is even the best measure of the financial status of OASDI.
Jackson bases his conclusion on the premise that all promised benefits will be paid. He contends that there is a common belief among workers and their families that they accrue an entitlement—politically, if not constitutionally—to receive Social Security benefits at promised levels sometime after entering the workforce and prior to reaching retirement age.[16] While he acknowledges that retirement benefits are not legally guaranteed, he argues that they are politically guaranteed and thus should be viewed as a certain obligation or debt of the federal government.
Jackson completely ignores the fact that the last two major reforms of OASDI in 1977 and 1983 dramatically cut retirement benefits. In 1977, Congress altered the indexing formula used to calculate retirement bene-fits.[17] Though one can argue that this legislation was merely correcting a mistake introduced into the system by the 1972 amendments, the result was very real to millions of older Americans—their retirement benefits were substantially lower than they expected and much lower than those of persons who began receiving benefits only a few years earlier.[18] The impact of the 1977 legislation on the so-called “notch babies” and all future cohorts was a major reduction in benefits.[19] This experience directly contradicts the fundamental principle on which Jackson bases his argument that accrual accounting is the best method of assessing the financial status of Social Security.
The 1983 Social Security amendments postponed the cost of living adjustment for current retirees, and, more importantly, increased the
*** Top of Page 166 ***
normal retirement age for future retirees from sixty-five to sixty-seven.[20] Increases in the normal retirement age and the accompanying decreases in the proportion of the primary insurance amount (PIA) received at all ages is realistically equivalent to a reduction in retirement benefits. For example, individuals born in 1943 who start their retirement benefits at age sixty-five will receive only 93.3% of their PIA instead of the 100.0% of PIA that earlier cohorts of retirees received. Thus, these individuals who were aged forty at the time of the 1983 amendments had a permanent reduction in the future benefits. Once again, history does not conform to the basic tenet of Jackson’s argument.
Projected revenues based on current law are insufficient to pay all promised benefits. Past modification of Social Security suggest that any future reform of OASDI will likely include a reduction in promised benefits. Jackson ignores historical experience when he concludes otherwise. He also ignores the recommendations of a series of presidential and congressional commissions, including the 1994–96 Social Security Advisory Council and the Bush Commission to Strengthen Social Security.[21] All of these review panels made recommendations for reductions in promised benefits as part of Social Security reform. In this way, Jackson’s assertion that all accrued benefits will be paid is contradicted also by most reform proposals. Once the premise that all promised benefits will be paid is removed, Jackson’s argument that the accrual accounting measure is the best indicator of the financial status of OASDI is substantially weakened. Thus, accrual accounting should be considered a useful and important piece of information that should be reported along with other measures to provide a more complete picture of the financial status of OASDI rather than the best measure of the financial status of Social Security.
The primary objective of the annual Trustees Report should be to present the best, most unbiased estimates of the financial status of OASDI. This goal demands projections of future revenues based on the current payroll taxes and future expenditures based on the current benefit formulas. Such projections must be long term and based on assumptions that reflect the best available information and research. Projections indicate annual differences between expenditures and revenues. After making
*** Top of Page 167 ***
the best and most unbiased projections, the next step is to determine how to summarize these results and present the data for evaluation. Prior to the 2003 Trustees Report, the Trustees reported the seventy-five-year actuarial balance and the year of depletion of the Trust Funds.
Jackson’s strongest arguments for the merits of accrual accounting are based on a comparison to the methods used in the pre-2003 Trustees Reports. For example, he writes, “the trustees’ reports are, in essence, statements of annual cash flows, comparing the system’s annual cash receipts to its yearly payments for benefits and administrative expenses.”[22] Additionally, he contends, “as long as inflow was adequate to meet outflow, the system was considered to be in balance.”[23] These statements leave the impression that annual revenues and expenditures are compared and that OASDI is declared to be on sound financial footing if there are surpluses in the current year. This has not been the case for many years. All recent Trustees Reports have considered a seventy-five-year horizon of revenues and expenditures. Prior to 2003 Trustees Report, the reports presented the actuarial balance as the increase in payroll taxes needed to pay all promised benefits over this entire period. Thus, even though there are current annual “surpluses,” the system was not considered to be in actuarial balance because of long-run deficits.
These measures presented a limited and somewhat biased assessment of the financial status of Social Security. In contrast, the 2003 Trustees Report presents a much more balanced assessment of the financial status of Social Security. First, throughout the report, the Trustees indicate the amount benefits would need to be reduced if revenues are not increased.[24] From this data the reader is able to consider the two basic options confronting policy makers (increasing taxes or reducing benefits). Policymakers and the public need to know the OASDI deficit in terms of both higher taxes and lower benefits. Together these measures present the range of possible policies that can be considered. Moreover, the explicit possibility of benefit reductions is acknowledged in this latest report.[25] Current Social Security accounting thus goes beyond the assumption adopted by Jackson.
Second, in 2003, the trustees for the first time present the present value of the seventy-five-year actuarial balance. This value indicates the amount of money that is needed today to provide sufficient funds to pay all promised benefits if the payroll tax and the benefit formula are unchanged. This concept provides an assessment of the additional funds needed today to finance the current system over the next seventy-five years. Jackson certainly overstates the facts by arguing that the reports do not
*** Top of Page 168 ***
recognize benefit obligations until the year of payment.[26] Since the long-range projections are based on benefit obligations and expected tax revenues, Jackson is incorrect. As reported in the 2003 Trustees Report, the present value of the seventy-five-year shortfall is $3.5 trillion.[27] This means that if $3.5 trillion were immediately transferred to the OASDI Trust Funds, all promised benefits could be paid with currently scheduled taxes.
Third, Jackson criticizes the use of the seventy-five-year projection of unfunded liabilities and points out the well-known “cliff” problem associated with projections of any specific length of time.[28] For years, analysts have recognized and discussed the implications of a fixed-term projection period.[29] Jackson correctly notes that an infusion of $3.5 trillion of new money to put the system in balance for seventy-five years would only temporarily solve the financial problem of insufficient revenues as a new deficit would emerge with each passing year.[30] He overlooks, however, the fact that in the 2003 Trustees Report the Trustees also presented an infinite time horizon estimate of the deficit. The infinite time horizon does not suffer from the “cliff” problem. The report estimates that the present value of the deficit associated with retaining the current system for an infinite time horizon is $10.5 trillion.[31]
These latter two concepts, the seventy-five-year actuarial balance and the infinite time horizon, are measures of the unfunded obligations of OASDI conditional on the system continuing to operate. They are important counterparts to the accrual accounting debt proposed by Jackson that illustrate the unfunded obligation conditional on the termination of the system. How should we decide which method should be the primary focus of attention? Jackson weakens his case by making a series of misstatements or overstatements concerning the current Trustees Report and the release of financial information for the system. Today, it would seem more likely that Social Security will be retained in some form, and so we should focus on the best methods of evaluating the financial status of the on-going system rather than focusing on the liabilities associated with terminating Social Security.
Beside his emphasis on the best measure of financial accounting for Social Security, Jackson is concerned with how the information is used by policy makers and the press. This is an important consideration, but even transparent accounting can be abused. Jackson describes a possible
*** Top of Page 169 ***
policy response to addressing the financial deficit using the current methods: the one-time transfer of sufficient monies to “solve” or eliminate the actuarial deficit.[32] He then concludes, “one of the unfortunate consequences of focusing on long-range actuarial deficits is that it encourages irresponsible recommendations of this sort.”[33]
First, I am not sure what is irresponsible about providing additional money to cover a deficit. Second, Jackson’s own proposal would create exactly the same incentive. Using accrual accounting, policymakers would be shown the amount of “debt” the system faces. There would still not be a tendency to try to solve the debt overhang with additional monies from the general fund, such as the Clinton proposals to allocate a portion of the projected federal surplus to OASDI.[34] If we relied entirely on accrual accounting as Jackson suggests, the underfunding of Social Security could be fully resolved by the federal government’s issuing new government securities and transferring them to the OASDI Trust Funds. This solution would eliminate the Social Security unfunded obligation. In Jackson’s analysis, it would have no effect on the government’s financial status, because the Social Security debt is merely being replaced by government bonds and Jackson implicitly assigns the same standing to both types of government obligations.
In today’s reform debates, transitional costs are well known. Any proposal for termination of the current OASDI program and replacing it with an individual account system must address the issue of transitional costs. This issue was made explicit by one of the proposals in the 1994–1996 Social Security Advisory Council.[35] Transitional costs and future revenue gains were also an important part of the discussion surrounding the recommendations of the Commission to Strengthen Social Security.[36] Jackson seems to have missed these debates as he writes, “In discussions of reform proposal, particularly those involving the partial privatization of Social Security benefits, analysts sometimes overlook the accrued claims of current workers and retirees.”[37] While it is undoubtedly true that some people overlook these costs, no reform proposals will receive serious consideration without directly addressing these transition costs, and all recent proposals by leading advocates of reform have included explicit methods of dealing with the transition costs.
The concept of transition from one retirement plan to another raises additional questions. Specifically, one needs to consider which workers
*** Top of Page 170 ***
would be required to shift to the new plan, who would have the option of changing plans, and who would remain in the old plan. Jackson discusses this issue and concludes that somehow accrual accounting would make the answers to these questions easier to ascertain.[38] In fact, virtually all reform proposals begin by stating that current retirees will not be affected.[39]
Future retirement policy must be based on a clear understanding of the cost of maintaining the current Social Security program and the cost of modifying or eliminating OASDI. The issue is what information the public and Congress need to develop an optimal retirement policy for the 21st century. In his thoughtful article, Jackson proposes that the primary method of presenting the financial status of Social Security be based on accrual accounting of existing liabilities and assets. His position is based on two basic principles: (1) participants in Social Security have an earned right to benefits based on taxes paid to date and (2) measurement of net liabilities is best done on a termination basis.[40]
Recognition of accrued liabilities is a useful and important aspect of evaluating the financial status of OASDI. It is not, however, necessarily the best method nor a method that is sufficient alone. First, promised benefits are not guaranteed. They have been reduced in the past and are very likely to be modified as part of any reform of Social Security. Second, OASDI is likely to be retained in some form. The current accounting approach is based on existing law and so to assess the financial status of Social Security under this assumption we should consider the projections of promised benefits along with scheduled taxes.
Every four years, the Social Security Advisory Board appoints a technical panel to review the assumptions and methods used in the annual Trustees Reports. I chaired the 2003 Panel that carefully reviewed the 2002 and 2003 reports. The Panel strongly endorsed the new methods for reporting the financial status of OASDI.[41] In particular, the Panel believed the inclusion of the present-value estimates and the addition of indications of the benefit reductions needed to bring the system into actuarial balance greatly improved the reports.[42] Moreover, the Panel believed that appropriate indicators of the financial status of OASDI as a continuing
*** Top of Page 171 ***
program were superior to accrual accounting methods based on the termination liability.[43]
Since Jackson began his paper, the OASDI Trustees have made significant changes in their annual report. If Jackson were starting his paper today, it most likely would have a much different emphasis. The case for accrual accounting must be made in comparison to the new measures presented in the 2003 Trustees Report that are based on the continued operation of OASDI. Based on the changes in the report, the central issue is why accrual accounting should be preferred to the use of the present value of projected revenues minus expenditures of an ongoing retirement program. Unfortunately, Jackson spends too much time attacking the shortcomings of other accounting methods.
[*] Professor, College of
Management, North Carolina State University. Chairman of the 2003 Technical
Panel on Assumptions and Methods that reviewed and evaluated the long- range
projections of the Trustees of Old Age Survivors and Disability Insurance as
presented in their annual reports. Ph.D., Duke University, 1974; M.A., Duke
University, 1972; B.A., Millsaps College, 1971.
[1] See Howell E. Jackson, Accounting
for Social Security and Its Reform, 41 Harv. J. on Legis.59
(2004).
[2] Compare 2002Bd. of Trs. of the
Fed. Old-Age And Survivors Ins. and Disability Ins. Trust Funds Ann. Rep.
[hereinafter 2002 Trustees Report], with 2003 Bd. of Trs. of the Fed.
Old-Age and Survivors Ins. Disability and Ins. Trust Funds Ann. Rep.
[hereinafter 2003 Trustees Report].
[3] See Jackson, supra note 1, at Part II.C.5.
[4] See id. at Part I.A.
[5] See 2003 Trustees Report, supra
note 2, at 42, 72–74. The report
presents estimates of the present value of the seventy-five-year open-group
unfunded obligations, the present value of the closed-group unfunded
obligations, and the present value of the infinite-time period open group
unfunded obligations. See 2003 Trustees Report, supra note 2, at 61–63. In addition, the report
indicates the amount that benefits would need to be reduced for expenditures to
equal revenues obtained from currently scheduled payroll tax rates. See
2003 Trustees Report, supra note 2,
at 3, 8.
[6] See 2002Trustees Report,
supra note 2, at 3.
[7] See 2003 Trustees Report, supra
note 2, at 2.
[8] See 42 U.S.C. § 401(a)(3)-(4)
(2000). Section 401 of the Social Security Act appropriates into the Trust Funds
amounts equal to the payroll and self-employment taxes reported to the Internal
Revenue Service. Id. at § 401(a).
[9] For a discussion of the financial status of
OASDI in the early 1980s, see generally Greenspan Comm’n, Report of the
National Commission on Social Security Reform (1983), available at http://www.ssa.gov/history/reports/gspan.html.
[10] See id. at 1.
[11] See 2003 Trustees Report,
supra note 2, at 3.
[12] See id. at 4.
[13] See Jackson, supra note 1, at introdution.
[14] See id. at Part
I.A.2.a.
[15] See id. at Part
I.A.2.a–b.
[16] See id.
[17] For a discussion of the 1977 Social
Security Amendments, see John Snee & Marry Ross, Social Security
Amendments of 1977: Legislative History and Summary of Provisions, Soc. Sec.
Bull., Mar. 1978, at 3–20; see also Larry DeWitt, Brief
History, at http://www.ssa.gov/history/briefhistory3.html
(updated Mar. 2003).
[18] See Snee & Ross, supra
note 17, at 13.
[19] See Sylvester Schieber & John
Schoven, The Real Deal. 176–82 (1999). The benefit reductions imposed by
the 1977 Amendments were referred to as “the benefit notch,” and
persons retiring who were directly affected by the reduction in benefits were
called “notch babies.”
[20] See Social Security Amendments of
1983, Pub. L. No. 98-21, 97 Stat. 65 (1983).
[21] See 1994–1996 Soc. Sec.
Advisory Council, Report On Social Security: Findings And Recommendations
25–33 (1997) [hereinafter 1994–1996 Advisory Council], available
at http://www.ssa.gov/history/reports/adcouncil/report/toc.htm;
President’s Comm’n To Strengthen Soc. Sec., Strengthening Social
Security And Creating Personal Wealth For All Americans 9–15 (2001),
available at http://www.csss.gov/reports/Final_report.pdf.
[22] Jackson, supra note 1, at introduction.
[23] Id. at Part I.
[24] See 2003 Trustees Report,
supra note 5, at 8.
[25] See id. at 3, 8.
[26] See Jackson, supra note 1, at Part I.A.2.a.
[27] See 2003 Trustees Report,
supra note 5, at 3.
[28] See Jackson, supra note 1, Part I.A.2.b.
[29] See, e.g., 1994-1996 Advisory
Council, supra note 21, at 12.
[30] See Jackson, supra note 1, at Part I.A.2.b.
[31] See 2003 Trustees Report,
supra note 5, at 61.
[32] See Jackson, supra note 1, at Part I.A.2.b.
[33] See id. at Part
I.A.2.c.
[34] See generally Douglas W. Elmendorf
et al., Fiscal Policy and Social Security Policy During the 1990s,
in Am. Econ. Pol’y in the 1990s 89, 106–09 (Jeffrey Frankel
& Peter Orszag, eds., 2002).
[35] See 1994–1996 Advisory
Council, supra note 21, at 25–33.
[36] See President’s Comm’n
To Strengthen Soc. Sec., supra note 21, at 73–93, available at http://www.csss.gov/reports/Final_report.pdf.
[37] See Jackson, supra note 1, at Part II.C.4.a.
[38] See id. at Part III.F.
[39] See, e.g., 1994–1996 Advisory
Council, supra note 21.
[40] See id. at Part I.A.2.a.
[41] See 2003 Technical Panel on
Assumptions and Methods, Rep. to the Soc. Sec. Advisory Bd. 87 (2003).
[42] See id.
[43] See id.
Copyright © 2004 by the President and Fellows of Harvard College
Harvard Journal on Legislation (JOL) - Volume 41, Number 1, Winter 2004
If you have any questions or comments about JOL, please
contact us: hlsjol[at]law[dot]harvard[dot]edu
Last modified: September 16, 2005.