The Century Foundation
41 East 70th Street, New York, NY 10021
(212) 535-4441

www.tcf.org

CBPP's logo
820 First Street, NE, Suite 510, Washington, DC 20002
Ph: 202-408-1080 Fax: 202-408-1080
www.cbpp.org

 

Revised July 15, 2002

 A RESPONSE TO THE EXECUTIVE DIRECTOR OF THE
PRESIDENT’S COMMISSION TO STRENGTHEN SOCIAL SECURITY

Peter A. Diamond and Peter R. Orszag[1]

 

PDF of response

Related Reports
Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security
PDF of technical 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.

In response to our analysis of the plans proposed by the President’s Commission to Strengthen Social Security, Charles Blahous (the Executive Director of the Commission and a member of the White House staff) has circulated a response.[2]  This document examines and attempts to clarify the sources of disagreement between the Executive Director and us so that interested observers can better understand the debate.

 

The Executive Director’s memo makes four fundamental assertions: (1) that our paper inappropriately focuses on the standard measure used to assess Social Security solvency — the ability of the traditional Social Security program to finance its benefits over the next 75 years; (2) that, contrary to one of our key conclusions, the Commission proposals we analyzed dramatically reduce Social Security’s reliance on general revenue transfers from the rest of the budget to a reasonable and affordable level; (3) that the figures from the actuaries’ analysis of the Commission’s plans that we used to measure the expected retirement income from individual accounts underestimate the value of that income; and (4) that our comparison of benefit levels under the Commission proposals to the benefits scheduled under the current benefit formula is inappropriate because the benefits scheduled under the current formula can not be financed out of current-law revenue.

This note responds to each of those assertions.  It is important to emphasize at the outset that the sources of disagreement do not generally arise from disputes over data.  Both our analysis and the Executive Director’s response are based on the same set of actuarial analyses produced by the highly respected Office of the Chief Actuary at the Social Security Administration.  The disagreements typically involve the interpretation and use of the figures produced by the actuaries, not those figures themselves.

This note makes the following observations:

1.   The use of the standard measure of Social Security solvency — its ability to finance traditional benefits over the next 75 years — in evaluating reform plans is entirely appropriate.  It is the method that the Office of the Chief Actuary itself used to evaluate the Commission’s plans and uses to evaluate all other reform plans as well.  (To be sure, the standard method of evaluating Social Security solvency is a vital part, but not the only component, of a full analysis of the effects of a reform plan.  The fact that supplementary measures can be useful in evaluating a plan does not undermine the crucial role and importance of the standard measure.)

2.   The methodology used by the Commission’s Executive Director to support his claim that the Commission proposals result in a dramatic reduction in general revenue requirements over the next 75 years is inconsistent with the standard analytical practice of both actuaries and economists, inconsistent with the methodology used by the Office of the Chief Actuary to evaluate this issue, and even inconsistent with the Commission’s own instructions on how to conduct such analyses;

3.   The figures produced by the actuaries that we used to estimate the expected retirement income produced by the individual accounts are the most appropriate figures to use.  These figures are more likely to overstate than to understate the annuitized income from the individual accounts that would be received during retirement, because the figures reflect an assumption (contrary to the rhetoric of the Commission) that the entire account balance would be converted into an annuity, with none of it set aside for bequests to heirs.[3]  If part of the account were instead set aside for bequests, the annuitized benefits that could be financed from the accounts would be lower than shown in our paper, and the reduction in combined retirement income relative to the benefits scheduled under the current benefit formula would be larger.

4.   All aspects of a Social Security plan should be compared against a single baseline, to avoid the confusion that arises from selective use of different baselines to evaluate different implications of a single plan.  The Executive Director’s memo, perhaps unwittingly, highlights this danger.  On one page in the memo, for example, the Executive Director uses “current law” to refer to the benefits scheduled under the current benefit structure (in a chart comparing financing issues).  On the very next page, he uses “current law” to refer to the benefits that would be paid if Social Security’s long-term financing shortfall were closed solely through benefit reductions after the Social Security Trust Fund was exhausted (in a table comparing benefit levels).  In each case, the definition of “current law” is the one that makes the Commission plans appear more attractive.  Even sophisticated analysts of Social Security are likely to be confused by such selective use of multiple baselines.  To avoid such confusion, a single baseline should be used to evaluate all the implications of a reform plan.  Despite its flaws, the baseline reflecting the benefits scheduled to be paid under the current benefit formula is preferable to the other potential baselines for this purpose.[4]

Before exploring these four issues in more detail, it is worth noting that the Executive Director never responded to two of our principal findings: that the individual accounts under the Commission’s principal plans (Models 2 and 3) are subsidized and therefore permanently worsen the financial condition of the S`ocial Security Trust Fund; and that without applying their reductions in traditional benefits to the disabled and children of deceased workers, those plans would not achieve actuarial balance (without even larger general revenue transfers than they already assume).  It is also worth noting that our analysis examined the effects of stock market risk on the retirement income generated by the individual accounts under the Commission’s plans.  The Executive Director sidesteps this issue: He entirely ignores the risks associated with the stock market.  By paying no attention to the risks connected to the stock market, the Executive Director’s memo makes individual accounts appear more attractive than they truly are.

 

Issue #1: Use of the Conventional Measure of Social Security Solvency

The initial sections of the Executive Director’s memo appear to criticize our use in assessing reform plans of the standard measure of Social Security solvency — the ability of the traditional program to finance its benefits over the next 75 years.  This criticism is unwarranted for two reasons:

The Executive Director appears to argue that actuarial balance within the traditional Social Security system itself is not an important issue and that, in assessing Social Security’s finances, the assets in individual accounts should be added to the Social Security Trust Fund’s reserves.  He fails to mention that however valuable the balances in the accounts may be to individuals, individual accounts do not themselves improve the ability of the Social Security system to finance its traditional benefits.  Individual accounts affect Social Security solvency only to the extent that they are linked in some way to benefit reductions within the traditional system.  The individual accounts under the Commission’s plans are tied to traditional benefit reductions,[5] and our paper takes such traditional benefit reductions into account in measuring the effects of the Commission plans on Social Security solvency.

To be sure, as suggested by our discussion below of the dangers associated with achieving actuarial balance in Social Security through large and unspecified general revenue transfers from the rest of the budget, a full analysis of the effects of a reform plan involves more than just the actuarial balance measure.  The standard method of evaluating Social Security solvency is a vital part, but not the only component, of a full analysis of the effects of a reform plan.  The fact that supplementary measures can be useful in evaluating a plan does not undermine the crucial role and importance of the standard measure.

Actuarial Balance within the traditional Social Security System versus Budget Effects

After criticizing us for applying the standard actuarial methodology used by the Office of the Chief Actuary, the Executive Director proceeds to mischaracterize one of our key conclusions: He appears to criticize us for encouraging unspecified general revenue transfers as a “solution” to Social Security’s long-term deficit, when our paper makes exactly the opposite argument.

Mr. Blahous spends an entire page arguing that actuarial balance within Social Security does not necessarily produce the resources necessary to finance Social Security benefits in the future.  In particular, he argues, transfers to the Social Security Trust Fund could eliminate the measured actuarial imbalance within Social Security but this would not improve the government’s ability to finance Social Security benefits (in the absence of other budgetary changes).  That is exactly the point we made in the Conclusion of our paper, however.  As we wrote:

“A claim of long-term balance that is heavily dependent on substantial, unspecified general revenue transfers, however, raises questions of credibility….Indeed, Congress could erase the long-term deficit in Social Security without any other changes simply by legislating that the Trust Fund would be able to draw upon general revenue as needed to finance scheduled benefits.  If no other budget changes were made, such legislation would raise serious questions about how the general revenue transfers could be financed when the need arrived, and in fact, in its Interim Report, the Commission underscored such questions when discussing — and disparaging — the idea of financing scheduled benefits by transferring funds from the rest of the budget….Given the current budget outlook, simply assuming the availability of such large transfers is highly problematic and could be regarded as fiscally reckless.”[7]

In other words, we agree that it is important to distinguish between two key issues:

That distinction is precisely why we are concerned about the size of the revenue transfers under the Commission’s plans.  As we discuss in the next section, the Commission’s plans require substantial amounts of general revenue transfers to restore solvency to Social Security.  Furthermore, the Commission failed to specify where these transfers would come from — that is, how the rest of the budget would be able to finance the transfers.

The Executive Director underscores his concern that transfers “create a tremendous inducement for federal policy makers to behave irresponsibly, to ‘save’ Social Security simply by issuing ‘transfers’ of credits from the general fund”  and that such an approach would be “irresponsible in the extreme.”  We fervently agree.  Given current budget projections, no responsible policy analyst should advocate simply crediting the Trust Fund with transfers from the rest of the budget and doing nothing else to eliminate the long-term deficit in Social Security.  Indeed, without any other changes in the rest of the budget, anything even resembling such an approach would be fiscally irresponsible in our view.

The Executive Director, however, somehow fails to apply the same standard to the Commission’s own key proposals — Models 2 and 3.  As our paper demonstrates and as the next section discusses, Models 2 and 3 are themselves heavily dependent on large transfers to “save” Social Security.  The general revenue transfers that would be required under Models 2 and 3 if all eligible workers participated in the individual accounts would amount to between two-thirds and four-fifths of what would be required to pay scheduled benefits simply by transferring funds from the rest of the budget to Social Security.  Both we and the Executive Director agree that using general revenue to eliminate the entire actuarial imbalance in Social Security would be grossly irresponsible in the current budgetary context.  It is difficult to see why assuming general revenue transfers that are between two-thirds and four-fifths as much over the next 75 years would not also be irresponsible.  Apparently without recognizing it, the Executive Director has basically labeled the Commission’s own plans as fiscally irresponsible.

 

Issue #2: Degree of Reliance on General Revenue Transfers

The Executive Director asserts that Models 2 and 3 dramatically reduce the need for general revenue transfers to Social Security over the next 75 years.  But as our paper demonstrates, the Commission’s two principal proposals — Models 2 and 3 — require between two-thirds and four-fifths as much general revenue over the next 75 years as simply using general revenue to pay scheduled benefits while making no other changes.  (These figures assume that all eligible workers choose to participate in the accounts; the range depends on whether the disabled are held harmless from the benefit reductions under the Commission plans.)

The Executive Director and some Members of the Commission have argued that the general revenue required by Models 2 and 3 is much smaller, relative to what is required to finance scheduled benefits, than we have shown it to be.  For example, the Executive Director claims that the current system “would require more than three times the amount of general revenues required by Model 2!” [italics in original].  In other words, he claims that the general revenue transfers required under Model 2 would amount to less than one-third of the general revenue transfers that would be needed to finance scheduled benefits (with no other changes), a much smaller fraction than our analysis finds it to be.  (Note: The Executive Director’s figures assume the disabled are not protected at all from the benefit reductions to which they would be subject under Model 2.  Under that assumption, our paper demonstrates that Model 2 would require two-thirds as much general revenue as simply financing scheduled benefits from the rest of the budget over the next 75 years.  That is twice the share the Executive Director acknowledges.)

How can we claim that Model 2 would require two-thirds as much general revenue as paying scheduled benefits under the current system, whereas the Executive Director claims that the general revenues required would be less than one-third?

Three factors explain the difference:

Since this third factor — the method used to total figures over a 75-year period — is so important, it may helpful to explain the difference using a simple example.  Suppose the general revenue transfers required were $100 (in inflation-adjusted terms) in each year for the next 75 years.  The present-value approach, which is the accepted practice and the one we use, measures the amount today that, with interest, would exactly pay off the $100 in each year in the future.  Under the assumptions used to evaluate the Commission plans, the present value of $100 per year for the next 75 years is $2,970.  In other words, the interest and principal on $2,970 today would be sufficient to pay $100 per year for the next 75 years.

The Executive Director’s approach, by contrast, would simply add up the $100 per year over the next 75 years and arrive at a total for the 75 years of $7,500.  The problem with this approach is that $7,500 today is worth much more than $100 per year for the next 75 years (because one could earn interest on the $7,500 in the meanwhile).  Fundamentally, the present-value approach recognizes that an inflation-adjusted dollar today is worth more than an inflation-adjusted dollar in forty years, because today’s dollar could earn interest over the 40-year period.  The Executive Director’s approach simply assumes an inflation-adjusted dollar in 2042 is equivalent to an inflation-adjusted dollar in 2002 and disregards the interest that could be earned in the intervening period.  In effect, the present-value approach recognizes that the government must pay interest on its borrowing and therefore treats a dollar paid today as more expensive than a dollar paid in the future.  The Executive Director’s approach treats a dollar of cost to the government as having the same value whether it is paid today or in the future.

It is our view, the view of the Social Security actuaries, and the view of most economists that the present-value approach is the only acceptable approach when analyzing very long periods of time.  Indeed, the Social Security actuaries present figures for each of the 75 years on a constant dollar basis but present the figures for the 75 years as a whole only on a present-value basis.[9]  Even more strikingly, the Commission’s Report itself states: “When measuring amounts that must be summed over a long-term time horizon, such as 75 years, the Commission report will use present values, discounted at the Treasury yield rate.”[10]  That is precisely the approach we adopt and the one the Executive Director rejects.  The Executive Director’s approach violates not only the accepted methodology for calculations involving 75 years but also the Commission’s own statement on this issue.

The table below demonstrates the crucial impact of this assumption.[11]  The first number in bold shows a 32 percent figure, which is consistent with the Executive Director’s assertion that the general revenue transfers required under Model 2 would be less than one-third as large as those required to finance scheduled benefits over the next 75 years with no other changes.  The difference between that figure and our figure (that the general revenue required under Model 2 would be two-thirds as large as the general revenue that would be needed to finance scheduled benefits over the next 75 years with no other changes, as shown in bold in the bottom right of the table) arises from two sources:

The table makes clear the crucial importance of using a present-value approach in interpreting the impact of the Commission’s proposals on general revenue requirements.  The inclusion of the constant dollar totals in the table is meant only to emphasize this point, not to suggest in any way that totaling constant-dollar figures over 75 years is appropriate or meaningful.

Gross Cash Flow Requirements from General Revenue Over the Next 75 years[12]

 

Model 2

Scheduled benefits

Model 2 as % of scheduled benefits

 

With two-thirds participation in the accounts

With full participation in the accounts

(with no other changes in the Social Security system)

With two-thirds participation in the accounts

With full participation in the accounts

Constant 2001 dollars

$7.0 trillion

$8.2 trillion

$22.0 trillion

32%

37%

Present value

$2.8 trillion

$3.4 trillion

$5.1 trillion

54%

66%

Note: Gross cash flow excludes the value of the existing Social Security Trust Fund and any cash-flow surpluses in Social Security.  In other words, it counts only cash-flow deficits in Social Security.  Under our preferred measure of general revenue transfers, all of the figures in the table would be lower.  Also of note: The figures presented above assume that the disabled and children of deceased workers are not protected from significant benefit reductions under Model 2.  If the disabled and children of deceased workers were protected from such reductions, the general revenue requirements under Model 2 would be greater than those shown here.

The bottom line is that Models 2 and 3 are heavily dependent on substantial transfers from the rest of the budget, and the size of those transfers betrays the claims of fiscal responsibility often made by the Commission’s members and staff.  The required transfers are only somewhat lower (between one-fifth and one-third lower) than would be required to finance all scheduled benefits out of general revenue over the next 75 years with no other changes in the Social Security system.  Since almost all budget analysts agree that the latter option — simply financing scheduled benefits out of general revenue transfers, without identifying a source for such transfers — would be unsustainable and unwise, the substantial reliance on general revenue in the Commission plans also appears to be unsustainable and unwise.

 

Issue #3: Evaluating the Retirement Income Produced by Individual Accounts

The Executive Director’s third assertion involves the expected retirement income produced by an individual account.  His disagreement with us involves how an accumulated balance in an account upon retirement is assumed to be transformed into expected retirement income. 

The retirement income figures presented by the Executive Director rest upon two assumptions: that the entire account balance would be transformed into an annuity (that is, a payment per month as long as the annuitant is alive); and that the annuity would take a particular form under which the retiree would continue to be exposed to financial market risks after retirement.  We explore the implications of each of these assumptions below, and show that both assumptions tend to overstate the amount of annuitized income in retirement that would be produced by an individual account.

Annuitization

The figures cited by the Executive Director (as well as the ones in our report) are predicated on an assumption that the entire balance in an individual account is transformed into a life annuity upon retirement.  It is important to realize that this assumption increases the annuitized income available in retirement, but only because none of the account is saved for bequests after retirement.

To provide a payment to heirs out of a certain accumulated account balance requires that some of the account be saved to finance the bequest.  Each dollar that a pensioner sets aside to bequeath to heirs means a dollar less to support consumption in retirement, because the pool of funds available to finance retirement benefits is reduced.  Since the same dollars can be used for only one purpose, the iron laws of finance necessitate such an outcome.

In promoting its proposals, the Commission has often highlighted the potential of payments to heirs from the individual accounts.  But when presenting estimates of the retirement benefits that individual accounts would provide, the Commission assumed that retired workers would make no bequests.[13]  The Commission essentially tried to have it both ways (and its combined claims essentially counted the same dollars twice).

As we explained in our paper, the assumption of full annuitization raises the retirement benefit levels that an individual account system would generate.  Nevertheless, we adopted the assumption of full annuitization.  If one instead assumes that the individual accounts would provide some bequests to heirs after retirement, the income from traditional Social Security benefits plus the annuitized income from the individual accounts would be lower than is shown in our paper.  If anything, therefore, our figures overestimate the retirement income produced by a given accumulated balance in an individual account.  The Executive Director’s estimates likely overstate annuitized retirement income for the same reason.

Variable annuities

Although both our figures and the figures cited by the Executive Director assume full annuitization, there is a difference between our figures and his on the level of retirement income that individual accounts could generate.  Although the Office of the Chief Actuary includes benefits calculated both ways in its analysis of the Commission plans, our approach is consistent with the preferred methodology of the actuaries; the approach adopted by the Executive Director is not.

The figures the Executive Director presents are based upon an assumption that retirees would continue to be exposed to stock market risks after retirement.  That is, the figures are based on a variable annuity, where the amount paid per year depends on stock market returns.  Under this approach, the monthly income that retirees have to live on would fluctuate from year to year, rising or falling depending on the performance of the stock market.  This assumption raises the expected annuity payment per year, but primarily because of the additional financial risk that the retiree is incurring.  Investing in the stock market carries risks, and investors therefore demand higher expected returns to compensate them for the additional risk they face.

Our figures, by contrast, are based on an inflation-adjusted annuity that does not fluctuate from year to year and does not expose the retiree to financial market risk.  This approach is justified because the annuity provided by the traditional component of Social Security is paid in this way.  To ensure comparability, it is appropriate to assume a type of annuity that has the same structure as traditional benefits.[14]

Furthermore, it is important to recognize that our approach, not the Executive Director’s, is the one preferred by the Office of the Chief Actuary at the Social Security Administration.  As the actuaries wrote in their analysis of the Commission plans:

“investment in a variable annuity would lead to substantial variation in annual increases in annuity amounts.  Increases in annual payments for an annuity at the rate of the increase in the CPI could not be assured.  In fact, in years when the variable annuity portfolio substantially underperformed the expected return, benefit payments from the annuity could even be lower than in the prior year.  Because of this uncertainty, we believe that variable life annuities would be selected by relatively few individuals.  Thus, we put primary emphasis on estimates reflecting distributions with CPI-indexed life annuities.”[15] [emphasis added]

In sum, our approach to evaluating the expected retirement income from a given accumulated balance in an individual account is the same as the preferred approach adopted by the Office of the Chief Actuary.  The approach adopted by the Executive Director is not.

Finally, it is worth noting that the Executive Director’s memo entirely ignores stock market risk before retirement.  (The discussion above regarding variable annuities involves stock market risk after retirement.)  By paying no attention to the riskiness of the stock market while an account is being accumulated, the Executive Director’s memo again makes individual accounts appear more attractive than they truly are.  Our paper, by contrast, presents figures both with and without an adjustment for stock market risk.  Our risk-adjusted figures use a simplified but broadly used approach that the Office of Management and Budget recently used in an analogous situation.

 

Issue #4: Evaluating all Aspects of a Reform against a Single Baseline

To avoid confusion, it is important to evaluate the expected benefit, financing, and other aspects of a reform plan against a single baseline.  In his analysis, by contrast, the Executive Director uses multiple baselines to evaluate different aspects of the Commission plans.  Interestingly, the Executive Director’s memo itself underscores our argument.  In particular, within the span of one page, the memo shifts from one baseline to another, without informing the reader of the shift, in a manner that in each case makes the Commission plans appear more attractive.

On page 8, the Executive Director’s memo uses “current law” to refer to the benefits scheduled under the current benefit formula in a chart comparing financing issues.  On the very next page, the memo uses “current law” to refer to “payable benefits” in a table comparing benefit levels.  (“Payable benefits” are the benefits that could be paid if no changes were made in Social Security until the Social Security Trust Fund was exhausted, and the entire shortfall was eliminated in each year thereafter solely by reducing benefits.)  The definition of “current law” that the Executive Director uses in each case is the one that makes the Commission plans appear more attractive: Current law is defined as scheduled benefits when the Executive Director wants to make the Commission’s plans look fiscally responsible (albeit in a misleading way, since the figures he cites are not based on present value), and is defined as payable benefits when the Executive Director wants to suggest that the Commission plans would raise benefit levels.[16]

The Executive Director’s memo, perhaps unwittingly, thus makes our argument for us regarding the importance of using a single baseline to evaluate all aspects of a reform plan.  A key issue then is which baseline should be used.[17]  In our opinion, the scheduled benefit benchmark is the most appropriate one to use for this purpose, despite the projected long-term deficit associated with it.  The Executive Director criticizes this approach, since scheduled benefits exceed the revenue available to Social Security under current law (a fact we note repeatedly in our paper).  But he fails to respond to the basis we put forward in our paper for why scheduled benefits are the most appropriate benchmark:

“As is well known, the projected cost of the scheduled benefits under current law exceeds the projected revenue available to Social Security.  Nonetheless, comparing the proposed benefit levels to scheduled benefits is the clearest way of describing the proposed changes, since the workings of current law are readily understood and since this type of comparison is the standard method used to evaluate the effects of Social Security changes on Social Security benefits.  For example, both the Greenspan Commission in the 1980s and the bipartisan, Congressionally chartered Advisory Council on Social Security in the 1990s employed this approach despite projected long-term deficits in Social Security at the time.  In addition, using scheduled benefits as the benchmark in evaluating proposed benefit changes is helpful because a reduction from scheduled benefits represents a reduction in the percentage of a worker’s pre-retirement earnings that Social Security (or combined benefits from Social Security and individual accounts) will replace…. In debates over Social Security changes, it is critical to identify changes in the percentage of wages that retirement benefits would replace, since these “replacement rates” affect how people’s standards of living are altered when they retire.”[18]

The first box in our paper examines this issue and explains our reasoning in more detail.

The Executive Director also asserts that we avoid “quantifying what [the] required changes would be [to restore long-term balance to Social Security], so as not to admit that the personal accounts in the Commission proposals would be able to offer higher benefits than the solvent version of existing Social Security.”  In his haste to issue a response to our paper, the Executive Director appears to have overlooked the lengthy final box in our paper, which addresses precisely this issue and provides the comparisons he requested.  We compare Model 2 to two alternatives that achieve 75-year actuarial balance after assuming the same level of general revenue transfers as under Model 2.  Our analysis “shows that on a risk-adjusted basis, Model 2 generally produces significantly lower combined benefits over the next 75 years — that is, it results in larger benefit reductions compared to the benefits scheduled under current law — than does an alternative with the same level of general revenue transfers….The reason for the generally larger expected benefit reductions under Model 2 than under the alternatives is that Model 2 would leave the Social Security system with more assets at the end of the 75-year period.  Under Model 2, but not under the alternative, the Social Security system would remain in balance after 2076.”[19]

 

Conclusion

On each of the four major assertions put forward by the Executive Director in criticism of our paper, we find his arguments are generally either mistaken (as in misunderstanding our opposition to large general revenue transfers in the absence of other changes to the budget), inconsistent with the most appropriate analytical methodology (as in his failure to use present value in presenting total figures for 75 years), or more likely to confuse the debate than to enlighten it (as in his use of multiple baselines selectively applied to reform plans to mask their less attractive features).

Furthermore, it is worth noting that the Executive Director never responded to two of our critical arguments: that the individual accounts under Models 2 and 3 are subsidized and therefore permanently worsen the financial condition of the Social Security Trust Fund; and that Models 2 and 3 would not achieve actuarial balance (without even larger general revenue transfers than those plans already assume) if their traditional benefit reductions did not also apply to the disabled and children of deceased workers.

Appendix

In addition to the claims addressed in the main body of this note, we wish to set the record straight on misleading assertions made in the Executive Director’s memorandum about the substantive positions of each of the authors.  First, the Executive Director claims that Orszag’s support of the Clinton Administration’s proposed general transfers to Social Security is inconsistent with his opposition to large transfers (without other changes in the budget) now.  Second, the Executive Director claims that Diamond’s support of price indexing in the mid-1970’s is inconsistent with his criticism of Model 2 now.  Neither assertion is valid.  We address each of these claims in turn.

The Difference between the Clinton Proposals and the General Revenue Transfers

Under Models 2 and 3

The Executive Director argues that “one of the authors, in his service with the Clinton administration, participated in the advocacy of a Social Security framework consisting of nothing but general revenue transfers as a substitute for actual Social Security policy….For one of the authors of this paper to criticize the Commission on grounds of general revenue use is ironic, given that the Commission proposals reduce the system’s general revenue requirements and given that one of the authors has been associated with policies that are confined entirely to committing additional income taxes through general revenue transfers.”

This statement is misleading on two levels.  First, it is inaccurate factually.  Orszag left government service in May 1998, well before President Clinton’s 1999 State of the Union speech in which the President proposed specific levels of transfers from the rest of the budget to Social Security.  Second, and more importantly, the comparison is invalid because of the dramatic shift in the budget outlook that has occurred in the intervening period.  At the time President Clinton proposed to transfer some general revenues to Social Security, the budget was projected to be in surplus over the time period of the transfers.  Given the political debate at the time, the alternative to the dedication of such revenue to Social Security was likely to be large tax cuts.  Indeed, the whole logic of the transfers was to enhance national saving by preventing a large tax cut (as was subsequently passed in 2001).  Now, and in part due to the passage of that tax cut, the budget is projected to be in significant deficit.

At a time of surplus, it is plausible that previously legislated transfers to Social Security would occur.  At a time of deficits, there is much less likelihood that the transfers would indeed occur (unless they were specifically tied to a particular source of revenue).  In other words, dedicating projected surpluses to Social Security is an entirely different proposition (from a political economy perspective) from assuming that transfers would occur despite large projected deficits.  The failure to recognize this distinction underscores the Commission’s general neglect of how its transfers would be financed.

The Difference between the Price Indexing Proposal in the mid-1970s and the Price Indexing Proposal Included in Model 2

A panel on which Diamond served in 1974-5 (Panel on Social Security Financing consulting to U.S. Senate Finance Committee) recommended a different and less severe form of price indexing than the one included in Model 2.  At the time, furthermore, the deficits projected for Social Security were much more severe than today and the overall system for determining benefit levels was widely regarded as inappropriate because of the overindexing of benefits.  Thus the recommendation of that early panel differs from the recommendation of the Commission in two ways – it involves a less severe method of price indexing, and the recommendation was made under vastly different circumstances in the financial positions of Social Security.

The type of price indexing under Model 2 differs from the version proposed by the 1974-1975 panel:

Furthermore, it was widely recognized in the mid-1970s that the system then in place for setting benefit levels involved excessive indexation and needed to be changed.  The price indexing proposal of the panel in 1974-1975 was one way to address this problem.  Congress solved the problem in the 1977 Social Security legislation through an alternative approach.  A discussion of wage and price indexing today differs fundamentally from a discussion of wage and price indexing prior to enactment of the 1977 legislation, since today’s system does not have the fundamental flaw that the pre-1977 system did.

Another difference is also worth noting.  At the time of the earlier panel, the magnitude of the actuarial imbalance was so severe that it was difficult to see how Congress could legislate sufficient benefit reductions and/or revenue increases to restore solvency to the system while adopting wage indexation.  According to the 1975 Trustees’ Report, the actuarial deficit was 5.32 percent of payroll and the Trust Fund was projected to be exhausted in five years.  By 1976, the actuarial deficit had risen to 7.96 percent, with a slightly later projected exhaustion.  (The severity of the projected imbalance at that time was underscored by the fact that the subsequent 1977 legislation was followed very shortly after by the 1983 legislation.)  By contrast, the actuarial imbalance is now 1.87 percent of payroll and the projected date of Trust Fund exhaustion is nearly four decades away.  Since the financial situation now is far less severe, a dramatic alteration in the system is neither needed nor desired


End Notes:

[1] Peter A. Diamond is Institute Professor and Professor of Economics at the Massachusetts Institute of Technology.  Peter R. Orszag is Joseph A. Pechman Senior Fellow in Economic Studies at the Brookings Institution.  The views expressed here are the authors alone and do not necessarily represent those of the staff, trustees, or officers of the institutions with which the authors are affiliated.

[2] Our papers are: Peter A. Diamond and Peter R. Orszag, “Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security,” The Century Foundation and the Center on Budget and Policy Priorities, June 18, 2002, available at http://www.cbpp.org and Peter A. Diamond and Peter R. Orszag, “An Assessment of the Proposals of the President’s Commission to Strengthen Social Security,” The Brookings Institution, June 2002, available at http://www.brookings.edu/es The response from the Executive Director of the Commission (Chuck Blahous, “Problems with the Diamond/Orszag Paper on the Proposals of the President’s Commission to Strengthen Social Security,” June 18, 2002) is available upon request to the authors.

 

[3] Annuitized income is income that is provided each month (or year) that the retiree (or the retiree’s spouse) is alive.

 

[4] If multiple baselines are used to evaluate a reform plan, it is crucially important that each baseline be used to evaluate all aspects of the reform (rather than selectively using one baseline for evaluating one aspect of the reform and another baseline for a different aspect of the reform).  One possibility is to use a single baseline to evaluate all aspects of a reform plan in the main presentation of an analysis and then to present alternative baselines in a box or appendix.  That is the approach we adopted in our paper.

[5] The linkage occurs with respect to the contributions made to the accounts, not the assets in the accounts.  As our paper explains, the resultant traditional benefit reductions under Models 2 and 3 are insufficient to compensate the Trust Fund for the diversion of revenue into individual accounts.  The accounts therefore impose a permanent cost on the Trust Fund.

[7] Peter A. Diamond and Peter R. Orszag, “Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security,” The Century Foundation and the Center on Budget and Policy Priorities, June 18, 2002, page 35.

[8] The Executive Director’s figures are based on “gross cash flow” between the Social Security system and the rest of the budget.  Our paper, on the other hand, measured general revenue transfers from the rest of the budget to Social Security.  The primary difference between these two approaches involves their treatment of the existing Social Security Trust Fund and future cash-flow surpluses in Social Security.  Under the Executive Director’s gross cash flow approach, only future cash-flow deficits in Social Security are included.  Under the general revenue transfer approach we adopt, by contrast, both cash-flow surpluses and cash-flow deficits in Social Security are included, as is the value of the existing Trust Fund (reflecting previous surpluses).  For example, assume that Social Security is expected to run a $100 cash-flow surplus some year within the next 75 years.  Under our approach, that surplus would be subtracted from the value of the deficits in other years to reach a summary 75-year figure.  Under the gross cash flow approach, however, the projected surpluses in Social Security would be entirely ignored.  We prefer the general revenue transfer measure, which is consistent with how the actuaries evaluate 75-year solvency.

[9] Memorandum from the Office of the Chief Actuary, pages 27-28.

[10] President’s Commission to Strengthen Social Security, Strengthening Social Security and Creating Personal Wealth for All Americans, page 101.

[11] Our constant-dollar figures for the amount of general revenue transfers required under Model 2 and required to finance scheduled benefits with no other changes over the next 75 years differ very slightly from those presented by the Executive Director.  His figures, which are consistent with those shown in the Commission report, show $6.9 trillion for two-thirds participation under Model 2 and $21.7 trillion for scheduled benefits.  Our figures were computed directly from the underlying cash flows presented in the actuaries’ analysis of Model 2, and the underlying data for our calculations are available upon request to the authors.  The differences are relatively minor and do not affect the conclusions.

[12] As noted in a footnote above, “gross cash flow” excludes the value of the Social Security Trust Fund and any cash-flow surpluses in Social Security.  It only counts cash-flow deficits in Social Security.

[13] If a worker dies before retirement and is not married, the individual account balance would be transferred to the worker’s heirs.

[14] Alternatively, even if the annuitant chose a variable annuity, the annuity payment would be dependent on the risky stock market.  Our use of an inflation-adjusted annuity could then also be seen as adjusting the variable annuity payments for the riskiness of stocks relatively to bonds (under the assumption that the higher return to stocks merely reflects their riskiness relative to bonds).

[15] Memorandum from the Office of the Chief Actuary, page 20.

[16] The Commission Report similarly uses confusing terminology to refer to its multiple baselines.  In the Report itself, the term “current law” is used to refer to payable benefits in the table comparing benefits on page 122 and then is used to refer to scheduled benefits in the figures on financing requirements on pages 125 and 126.

[17] If multiple baselines are used to evaluate a reform plan, it is crucially important that each baseline be used to evaluate all aspects of the reform (rather than selectively using one baseline for evaluating one aspect of the reform and another baseline for a different aspect of the reform).  One possibility is to use a single baseline to evaluate all aspects of a reform plan in the main presentation of an analysis and then to present alternative baselines in a box or appendix.  That is the approach we adopted in our paper.

[18] Peter A. Diamond and Peter R. Orszag, “Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security,” The Century Foundation and the Center on Budget and Policy Priorities, June 18, 2002, page 5.

[19] Peter A. Diamond and Peter R. Orszag, “Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security,” The Century Foundation and the Center on Budget and Policy Priorities, June 18, 2002, pages 32-33.

[20] For further detail on the alternative approaches, see Stephen C. Goss, “Long-Range OASDI Financial Effects of a Proposal to CPI-Index Benefits Across Generations,” Office of the Chief Actuary, Social Security Administration, May 3, 1999.