Consulting Actuaries ¦ Specialists in Employee Benefit Plans
Re: The GAO's $0.5 million study fails to dispel disturbing doubts about Social Security's shaky actuarial projections

By David Langer, FCA, ASA, EA

The U.S. General Accounting Office (GAO) has now responded to Congressman Jerrold Nadler's January 1999 letter formally requesting the GAO to investigate my assertion that the Social Security actuarial projections violate the Actuarial Standards of Practice and, also, the conclusion in my Contingencies article, "Social Security Finances are in Fine Shape," that there would not be a financial deficit in the 75 year projections, if the Social Security actuaries had complied with the standards.

The GAO retained PricewaterhouseCoopers (PwC), at an estimated cost of over half a million dollars, to respond, and PwC submitted a 100 page report. The GAO then prepared a five page letter on January 14 to Rep. Nadler with its own analysis.

At the request of Rep. Nadler's office, I submitted to him the enclosed "Critique of the GAO/PwC documents." My critique notes that PwC and the GAO both echo the conclusions of the Social Security actuaries which, we now know, were surreptitiously dictated by the politically appointed Social Security trustees. My critique concludes that the PwC and GAO analyses fall short in significant ways; they do not therefore provide a useful rebuttal of my contentions. It has also evidently not occurred to the GAO and PwC that there is a manifest unsoundness in proposing major cutbacks to, and the privatization of, Social Security on the basis of highly dubious 75 year projections.

I remain confirmed in my belief that the Social Security trustees were politically motivated to produce an alarming financial picture of the Social Security program to worry the public into accepting the benefit cutbacks and privatization that have been promoted by those who would benefit from them.

A Christian Science Monitor reporter interviewed me on the above, and his April 10 article is enclosed.

David Langer

P.S. For a copy of the GAO's letter call or fax my office. For PwC's report, contact the GAO (publication GAO/AIMD-00-53R, Social Security Actuarial Projections)

May 3, 2000
Critique of the GAO/PwC documents

By David Langer, FCA, ASA, EA
Consulting Actuary

The General Accounting Office and PricewaterhouseCoopers (GAO/PwC) documents were produced in response to Congressman Jerrold Nadler's 1/27/99 letter to the GAO requesting a review of the validity of the actuarial assumptions underlying the cost projections prepared by Social Security's actuaries. The documents include PwC's 101 page study dated 10/29/99 and the GAO's two letters dated 1/14/2000, the second of which replies to Mr. Nadler's questions that arose from my talk before the 1/21/99 Congressional Conference on Social Security. The GAO accepts PwC's conclusions.

I found that the documents fall short of a thorough evaluation in significant ways and therefore do not constitute a useful rebuttal of my contentions, namely, (1) that the actuarial projections that appear in the annual reports of the Social Security trustees are in violation of the Actuarial Standards of Practice (ASOPs) of the American Academy of Actuaries, and (2) that a 75 year projection period is of little value in making decisions on whether to cut Social Security's benefits, raise its retirement age, or privatize it. Consider the following: ASOP violations

1. Failure to adequately recognize past experience.
The documents support the reliance by the trustees/actuaries on future economic and demographic trends in projecting GDP. I suggested a mix of 80% past experience and 20% for future trends. GAO/PwC, on the other hand, ignore past experience and stress demographic trends as one of the two key elements of GDP growth. This leads to a projected 1.5% average GDP (compared to the long-term 3.3%) using a 1.3% rate of productivity increase, and about a 0.2% rate for total hours worked based on future increases in the labor force (compared to 1.8% for the period 1990 to 1998).

The 0.2% rate also looks only at workers in the 18-64 age bracket. The GAO/PWC thus unrealistically ignore all of the increases in labor force that will arise because the projected worker shortage will result in workers over 65 continuing to work because of employer incentives, higher immigration, the growing wish to return to work out of financial need or boredom with retirement, and longer hours worked by the 18-64 group. There will undoubtedly also be increases in productivity necessitated by the diminished supply of workers. The GAO/PwC do not note that the number of agricultural workers required to feed 100 persons fell from 15 in 1910 all the way down to one in 1990, and a similar phenomenon has been occurring and will likely continue to occur in nonagricultural areas.

2. Failure of actuaries to reveal dominant role of trustees.
The GAO/PwC documents make no mention of the curious, if not improper, relationship that exists between the politically appointed trustees and Social Security's actuaries and how this affects the financial projections which fill the trustees' reports. The reports give the distinct impression that the actuaries make the key decisions on actuarial matters. The actuaries also prepare numerous studies on how the assumptions are derived and also discuss these at professional meetings. It is therefore not surprising that the many actuaries I have asked about this, who have not worked for Social Security, have come to believe, as I have, that the Social Security actuaries are in total control of actuarial matters. However, we now know that the trustees control the process.

ASOP No. 32, "Social Insurance," states in Sec. 4.1.8 that, "If any assumption was prescribed by someone other than the actuary, the actuary should disclose that fact and the source of the assumptions." This has not been done and is thus a violation of ASOP No. 32. The effect is to conceal the significant interactions between the trustees and the actuaries and the apparent manipulation of the projected deficit values by the trustees. This has been detailed in my January 4 Christian Science Monitor article.

The violation is a sensitive point for the SSA actuaries. This became evident when I raised it with the deputy chief actuary who called me in response to a letter to the Commissioner of Social Security by Rep. E. Clay Shaw, Chairman of the Social Security Subcommittee of the House Ways & Means Committee. The actuary initially denied the violation by stating that the SSA actuaries have, in fact, expressed disagreement with the trustees' assumption choices in prior annual reports. When I asked him to specify, he said I would find this in the trustees' annual reports on Medicare's Hospital Insurance trust fund. I then asked, but what about the annual reports on Social Security? He hesitated, but then acknowledged the SSA actuaries had never objected to any trustees' choice of assumptions in his tenure at Social Security since the 1970's.

The deputy chief actuary also confirmed to me that the preliminary reports the SSA actuaries prepare for the trustees and the records of their discussions are classified at a top level of secrecy and are thus not available to the public. This is unfortunate from their point of view, since it suggests they have something to hide, especially in light of the fact that the actuaries release volumes of data in all other actuarial areas. 75 year projections lack validity

If the trustees/actuaries are going to base their projections on a 75 year period relying on a crystal ball, even one peered into by actuaries, economists, and demographers, then two unfortunate facts need to be recognized by GAO/PwC, but aren't: (1) it is well established it is foolish to make financial projections over extended periods; those that make them invariably wind up with egg on their faces, and (2) the assumption selection process for a 75 year period is highly susceptible to political manipulation, and the evidence indicates this happened regularly over the years to prejudice the projections into showing deficits of over 2%.

The GAO/PwC documents demonstrate a lack of appreciation of these two negatives. For example, the second GAO letter states on page 2 that my study "overwhelmingly relied on historical rates of GDP growth and did not include readily available information on future demographic trends." Therefore, the GDP growth I projected "over the next 75 years does not seem reasonable."

The GAO and PwC thus give their stamp of approval to the trustees' use of assumptions derived from projections of demographic and economic factors for 75 years. But, in addition to the two objections to 75 year projections cited above, ASOP No. 32, Sec. 3.1 also states that, "Because no one knows what the future holds with respect to economic and other contingencies, the least an actuary can do is use professional judgment to estimate possible future economic outcomes based on past experience and future expectations." And, it continues in Sec. 3.3: "... the actuary should consider recent economic data. However, the actuary should not give undue weight to recent experience." While the statements refer here to economic data, they are no less true for the demographic assumptions.

See the last page of my Contingencies article, "Social Security Finances Are in Fine Shape," for the opposing conclusions about the future drawn from the work of the eminent demographic economist, Richard A. Easterlin, which indicated in the early 1990s that we are now on the verge of an improving economy. On this basis, his colleague, Diane Macunovich, concluded that the Social Security actuaries, in arriving at an over 2% deficit, were basing assumptions on the lowest point of the demographic-economic cycle, and were therefore unduly pessimistic. Her observation has, in fact, been born out for many years now, the GDP substantially exceeding that assumed for the actuarial projections.

Shaky assurance offered by the use of three projection levels

The use of the range provided by three levels of projections (Optimistic, Intermediate, and Pessimistic) by the trustees/actuaries does not right the pitfalls of the 75 year projections. Their range departs so radically from the extensive past experience for the GDP as to appear exaggerated. Consider that they arrived at a 75 year average GDP of 1.5% for the Intermediate level and then added and subtracted about 0.7% to obtain the Optimistic and Pessimistic average values of 2.2% and 0.8%. Now compare this to the range derived from the actual average GDP of 3.3% plus and minus the same 0.7%: 4.0% to 2.6%. The latter 2.6% Pessimistic value is higher than the trustees/actuaries 2.2% Optimistic value! This can be clearly seen on the enclosed graph, "GDP averages: prior years and future years (Optimistic, Intermediate, and Pessimistic)."

There is thus a surprising lack of proportion in the judgement of the trustees/actuaries: they are willing to accept a shaky basis for their projected deficits as being sufficiently reliable for making substantive reductions to Social Security that would affect over 200 million persons. Conclusions

The GAO/PwC documents largely accept the financial projections of the trustees/actuaries. In doing so, they have not fully considered, or even considered at all, the violations of the American Academy of Actuaries' Actuarial Standard of Practice. Without these violations, there would be no long-term deficit. The GAO/PWC also give no thought to the notion of the fundamental unsoundness of making major cutbacks to Social Security using 75 year projections as a basis.

In my opinion, the documents do not constitute a rebuttal of my contentions. I remain confirmed in my belief that the Social Security trustees were politically motivated to produce an alarming financial picture of the Social Security program to worry the public into accepting the benefit cutbacks and privatization that have been promoted by those who would benefit from them.

David Langer, FCA, ASA, EA
Consulting Actuary

References prepared by David Langer
"Social Security Finances Are in Fine Shape" (Contingencies, May-June 1999) "Cooking Social Security's deficit" (Christian Science Monitor, Jan. 4, 2000) "MEMO RE: Social Security Actuaries's Changes in Methods and Assumptions,1979-1998." Prepared for Congressional Conference on Social Security, Jan. 21, 1999. Graph (enclosed): "GDP averages: prior years and future years (Optimistic, Intermediate, and Pessimistic)"