•Press Release Economic Growth Health and Social Programs
December 6, 1999
Despite Robust Growth of Recent Years, Projections are Lowered
Forecast Sees 21st Century America as Poorest of Industrial Nations
For Immediate Release: December 6, 1999
The newly issued report of the 1999 Technical Panel on Assumptions and Methods is considerably more pessimistic on prospective economic and productivity growth than the report issued by the previous panel in 1995. This increased pessimism is striking, since there has been exceptionally strong economic growth in the four years that have elapsed between the two panels.
Specifically, the new panel:
In addition, the panel:
The key downward revision in the panel’s report is in the projections for productivity and real wage growth. This issue is complicated somewhat by changes in measurement by the Bureau of Labor Statistics (BLS) and Commerce Department over the last five years. These changes in measurement must be taken into account when comparing the projections from the two panels.
The 1994-1995 panel split on its recommendation for productivity growth, with the average annual rate being 1.35 percent. Changes in the procedures for measuring productivity since the 1995 report have the effect of raising the measured rate of productivity growth by approximately 0.3 percentage points annually, so that the previous panel’s projection for productivity growth would be approximately 1.65 percent annually, using the current methodology.1 The 1999 panel recommended projecting a 1.35 percent annual rate of productivity growth, approximately 0.3 percentage points less than the previous projection, when adjusted for the changes in methodology. According to the most recent data from BLS, productivity growth averaged 2.1 percent annually between 1994 and 1998.
The 1994-5 panel also split on the projected rate of wage growth, with the average being a projection of 0.9 percent annually. The changes to the consumer price index (CPI) since the last report have the effect of raising the measured rate of real wage growth by 0.5 percentage points annually, as noted in this report (p 53). This means that if the projected rate of real wage growth from the 1995 report was adjusted for the changes in the CPI, it would be raised to 1.4 percent annually. The current report instead projects a 1.1 percent rate of real wage growth. The average rate of real wage growth over the last four years, using the Social Security Administration’s data, has been 2.6 percent annually.
It is worth noting that the latest report assumes slower productivity growth than the United States has ever experienced over any prolonged period. According to the most recent data from BLS, productivity growth has averaged 2.0 percent annually since 1959. In the quarter century since the productivity slowdown began in 1973 it has averaged 1.5 percent annually.
The panel’s projection also assumes that productivity growth in the United States will be substantially lower than that experienced in most European nations in recent decades. For example, according to the BLS, annual productivity growth since 1980 in France, Sweden, and Germany has averaged 1.6, 1.7, and 1.8 percent respectively.2 If these nations sustain their recent pace of productivity growth, the panel’s projections imply that their citizens will enjoy substantially higher living standards in the next century than workers in the United States.
Other noteworthy recommendations in the panel’s report also have the effect of worsening Social Security’s financial prospects. The panel recommended lowering the projected real interest rate on the government bonds held by Social Security to 2.7 percent. According to the panel, this debt has paid an average real interest rate of 3.98 percent over the last decade.
The panel also recommended a more optimistic assumption about the rate of increase in life expectancies. This piece of good news has the effect of worsening the program’s finances since it means that future generations will be collecting benefits for a longer period of time than workers do at present.
The net effect of these recommended changes is to raise the projected shortfall by 0.53 percentage points. However, there is likely to be relatively little effect on the projected date of the trust fund’s depletion (currently 2034), since most of the impact is due to longer projected life expectancies later in the century.
The increase in projections for life expectancies will help to focus attention on the real nature of the problem facing the Social Security system. In public debate, it has wrongly been portrayed as a generational issue, with the large baby boomer cohort bankrupting the system. In fact, the large size of the baby boom cohort has relatively little net effect on the program’s finances (they get more in benefits after they retire, but they pay more in taxes during their working years).
The real problem is simply that it will cost more money to support the longer retirements that future generations are projected to enjoy, because of progress in medical technology and the general affluence generated by the economy. As a result of the projected increase in life expectancy, the annual deficit in the year 2075 (when all the baby boomers are long gone), is projected to be 7.7 percent of payroll. Future generations will have to decide for themselves how best to divide longer life spans between work and retirement. A perverse implication of the seventy-five year planning period for Social Security is that these decisions should be made by the current generation of politicians.
One of the most striking changes in this panel’s recommendations, compared with those of previous panels, is that it assumes that there will be a continued redistribution of income from wages to profits. In the past, following standard economic theory on this topic, the projections had always assumed that the shares of income going to capital and labor remained constant over the long-term. As a result of the large shift in income from wages to profits in the last decade, this panel recommends that the trustees project a continued shift at the rate of 0.06 percent annually. This projected shift in income can in effect be seen as a "capital tax" from the standpoint of workers. While the size of the shift is not that large, an increase in the Social Security tax of the same magnitude (which would have the exact same effect on workers’ living standards) would be sufficient to balance the program’s finances indefinitely.
The panel recommended a significant reduction in the projected rates of return assumed from stock, from a 7.0 percent real annual rate to a 5.7 percent rate. It also recommended an accounting treatment that recognizes the greater risk associated with stock returns. Unfortunately, the panel did not attempt to construct a set of projections for stock returns, derived from projections of its components — dividends and capital gains — in the same rigorous manner that it derived its projections of wage growth from productivity growth. There is no explanation for the panel’s failure to use a systematic approach to projecting stock returns, particularly since other economists, such as M.I.T. Professor Peter Diamond and Dean Baker, have shown exactly how such projections can be made.3 Had the panel used such a systematic approach, it would have projected an even lower return for equities of approximately 4.0 percent, or alternatively, a decline of 50 percent in the stock market in the near future.
See "What Stock Market Returns to Expect for the Future," by Peter Diamond, Center for Retirement Research at Boston College, 1999; "Saving Social Security With Stocks: The Promises Don’t Add Up," by Dean Baker, Century Foundation (formerly the Twentieth Century Fund), 1997; or Letter to Martin Feldstein, May 15, 1999.