Revised 9/16/99 -- See fb-changes.html for what's new and a revision history.
Index Of All Web Pages (1)
Master Table of Contents (2)
Index page (fb-index)
<<###PREVIOUS(fb-76)
NEXT(fb-81)###>>
[A] The Social Security Trustees Are Using Pessimistic Demographic and Economic Assumptions? (234)
[A1] Introduction and Summary (234)[A2] Demographic and Economic Parameters, General (236)
[A6] Productivity and Real Wage Growth (241)
[A7] Workers Per Dependent (243)
[A8] The Rapidly Growing "Old Old" Population (251)
[A7.A] General Discussion (244)[A7.B] Total Number of Dependents, Including Children, Will Be No Worse Than In The 1960's (246)
[A7.B1] Population In Three Age Brackets. And Dependency Ratios Based On Age Brackets (247)[A7.B2] Aged Dependency Ratio To Increase 69%, From 0.210 Now to 0.354 In 2030 (248)
Table T234.1 Key Economic and Demographic Assumptions -- Social Security Trustees' Intermediate Forecast, March 1999 -- For Years 2000 - 2075 (235)
Table T239.1 Life Expectancy (240)
Table T241.1 Productivity and Real Wages (242)
Some people fault the Social Security (OASDI) Trustees' Intermediate projection as being very pessimistic with respect to real GDP growth -- about 1.3%/year after 2020 -- which is considerably less than the historic rate of real GDP growth. In the most recent 10 year period, 1989-1998, the real GDP grew at 2.4% per year. In the 1980s it grew at 2.7%/year. In the 1970s, 3.5%/year. In the 1960s, 4.4%/year. See section non35 for more on the real GDP growth rate.
However, the Trustees explain that the GDP growth is low because the growth in the labor force is expected to be very low -- 0.2% per year after 2020, and 0.1% per year after 2050 (See Table T234.1). Once again, it is demographics, demographics, and demographics. (A lot of people understand how there will be fewer workers per Social Security beneficiary in the future, and how that will put upward pressure on SS tax rates. But relatively few seem aware that there will be a slowdown in the growth rate of the labor force, and how that is going to slow down economic growth, barring a dramatic improvement in productivity). The twin demographic problems of low fertility and long life expectancy have many facets, and all impact Social Security and Medicare financing adversely.
The labor force growth rate projections are based on a 1.9 children per woman fertility rate (non237) -- a little lower than 1998's 2.04 but much higher than most European countries and Japan are experiencing. Also, between 1975 and 1987, the U.S. fertility rate was below 1.9. So a fertility rate of 1.9 is not an unrealistically low assumption.
Productivity (non241) has been declining steadily for the last several decades. For example, here are the productivity growth rates (in %/year) for the past four decades: 1958-67: 2.9%, 1968-1977: 2.0%, 1978-1987: 1.0%, and 1988-1997: 0.9%. The Trustees assume a future productivity growth rate of 1.3%/year. That is the average for about the last 30 years, but it is quite an improvement over the last 20 years. On the other hand, productivity during the last 3 years has been substantially better. We'll see after we get Y2K behind us if this rather short-term productivity improvement is sustained.
The growth rate in longevity (life expectancy) (non239) has been decelerating, though with a lot of ups and downs. See Section 3 of Table T239.1.
At first glance, it appears that the Trustees are greatly underestimating future life expectancy.
The Trustee's forecast of a 0.9 %/decade growth rate for life expectancy at birth is half the 1990-2000 growth rate of 1.8 %/decade.
The Trustee's forecast of a 2.1 %/decade growth rate for life expectancy at age 65 is considerably less than the 1990-2000 growth rate of 2.9 %/decade.
Given that there seems to have been a decelerating trend in longevity growth, it is hard to fault the Trustees' forecast. But one should read the Gray Dawn book material in the Life Expectancy section (non239) for some reasons why the author believes the Trustees are underestimating longevity growth.
The size of the "old old" population (non251) (those aged 85 and over) is expected to more than triple between now and 2040. This is one reason to expect Medicare costs per beneficiary to continue to grow at a much higher rate than general inflation.
Immigration (non238) is assumed to stay at the same level in the future as recently, 900,000 per year. This is probably too low an estimate. Higher immigration will be better for Social Security, as immigrants are mostly young working-age people and their children. However, note that immigration will have to at least double, and stay at that higher rate permanently, in order to have any noticable fiscal impact. And that even at today's rate, the population ethnic mix is changing rapidly. And don't expect immigrants to be thrilled to pay high taxes to support priveleged white native-born elders.
The number of covered workers per SS beneficiary is expected to drop from 3.4 today to 2.0 in 2040 and 1.8 in 2075. See section non243 for more on this.
See also a table of the ultimate (year 2075) demographic and economic assumptions for the SS Trustees' Low Cost, Intermediate, and High Cost Scenarios. And the actual 1989-1998 10-year average. The table is in section non168. This section also points out that even if the Trustees' Low Cost Scenario occurs, the general taxpayer will still have to provide funds to the SS program, beginning in 2019, in order for the SS program to continue to pay full benefits.
To sum this up, I don't see anything that is "pessimistic" about the Trustees' Intermediate forecast, at least when judging it against the historic data. If anything, some of their forecasted parameters seem a little optimistic, but this is certainly arguable. Also, I don't have any access to any deeper material or analysis that they have done. Just as one can overestimate real GDP growth by extrapolating historical real GDP data into the future (and thus not factoring in the expected decline to near zero of the labor force growth rate), I may be making similar errors by extrapolating life expectancy into the future from historic life expectancy data. And likewise for other parameters.
Table T241.1 below has the March 1999 Trustees' Intermediate Forecast projections for some key demographic and economic parameters.
===================== (235) ==========================
===================================================Table T234.1 Key Economic and Demographic Assumptions -- Social Security Trustees' Intermediate Forecast, March 1999 For Years 2000 - 2075 Average Annual Percentage Change Life ---------------------------- Fert- Expect- Labor Product- Real Real ility ancy at Force ivity Wages GDP Rate Birth 1960s 1.7 2.2 4.4 1970s 2.4 0.5 3.5 1980s 1.7 0.5 2.7 1990s 1.1 1.2 2.5 2000 1.0 0.9 2.0 2.03 76.7 2010 0.6 1.0 1.8 1.97 77.8 2020 0.2 0.9 1.4 1.92 78.5 2030 0.2 0.9 1.4 1.90 79.2 2040 0.2 0.9 1.4 1.90 79.8 2050 0.1 0.9 1.3 1.90 80.4 2060 0.1 0.9 1.3 1.90 81.0 2070 0.1 0.9 1.3 1.90 81.6 2075 0.1 1.3 0.9 1.2 1.90 81.9Note: See Table T241.1 below for historical productivity and real wage growth.
TR99 is the March 1999 Social Security (OASDI) Trustee's Report
The Average Annual Percentage Changes in Real GDP, Real Wages (real wage differential), and Labor Force come from TR99's Table II.D1, p. 58 {74}. Table II.D1 also contains historical data back to 1960.
The ultimate rate of productivity growth comes from the TR99 text P. 54 {70} and 148 {164}. There aren't any projections for earlier years given, but I believe 1.3%, or something quite close to it, is used throughout the 2000 - 2075 period.
The Fertility Rate and Life Expectancy at birth come from Table II.D2 p. 62 {78}. The Life Expectancy at birth is given separately for males and females. I calculated a simple average life expectancy as ( male life expectancy + female life expectancy ) / 2.
Real GDP growth exceeds Real Wage growth because of increasing population size. Real GDP per capita growth would be more comparable to Real Wage growth.
Here is more discussion from TR99 p. 54 {70} on the economic and demographic assumptions:
[Begin TR99 p. 54 {70} quotes:]
For alternative II [the Intemediate scenario], the annual growth in real GDP is assumed to average 2.0 percent over the short-range projection period (1999-2008), a slower rate than the 2.4 percent average observed over the most recent historical ten-year period (1989-98). This 0.4 percent slowdown is mostly due to slower projected growth in labor force and employment.
After 2008, the projected rates of growth in real GDP, for all three alternatives, are determined by the assumed rates of growth in employment, average hours worked, and labor productivity. The Trustees assume an intermediate trend growth rate of labor productivity of 1.3 percent per year, roughly in line with the average rate of growth of productivity over the last 30 years.
The Trustees project much slower growth in the working age population in the future based on the natural consequence of the "baby-boom" generation approaching retirement and the succeeding lower-birth- rate cohorts reaching working age. The projected slowdown in labor force growth also reflects the cessation of relatively rapid growth in labor force participation rates, particularly by women, by about 2010. The annual rate of growth in total labor force decreased from an average of about 2 percent per year during the 1970s and 1980s to about 1.1 percent from 1990 to 1998. After 1998 the labor force is projected to increase at about 0.9 percent per year, on average, through 2008, and to increase much more slowly after that, ultimately reaching 0.1 percent toward the end of the 75-year projection period. Thus, the projected growth of real GDP falls toward trend productivity growth because of the Trustees projected decline in labor force growth over the long term. With productivity growth at 1.3 percent and labor force growth slowing to essentially zero, ultimately trend GDP growth slows to about 1.3 percent as well.
[End TR99 p. 54 {70} quotes]
The Trustees' Intermediate Forecast projects a 1.9 children per woman fertility rate -- a little lower than 1998's 2.04 but much higher than most European countries and Japan are experiencing. Also, between 1975 and 1987, the U.S. fertility rate was below 1.9 (See TR99 Table II.D2). So a fertility rate of 1.9 is not an unrealistically low assumption.
TR99 is the March 1999 Social Security (OASDI) Trustee's Report
Here are some perspectives of Gray Dawn about fertility rates:
[From Gray Dawn, Excerpts on Fertility:]
[p. 49:] The total fertility rates of the G7 countries in the 1995-2000 period, per a U.N. 1998 projection are: U.S.: 2.0; UK: 1.7; France: 1.6; Canada: 1.6; Japan: 1.4; Germany: 1.3; Italy: 1.2.
[p. 60:] Likewise, on fertility, most government and global agencies assume an eventual return to the replacement rate of 2.1 births per woman, no matter how far below that rate the developed countries have already drifted. In the United States, the Social Security Administration assumes that the fertility rate will continue to hover indefinitely at 1.9, just below replacement. Yet this too can be regarded as optimistic, since 1.9 is higher than the current fertility rate in 19 of the other 22 developed countries (excepting only New Zealand, Ireland, and Iceland). [The 23 developed countries are the OECD member countries during 1973 - 1993 and are listed on p. 244 of Gray Dawn]. Even the SSA's "pessimistic" or "high-cost" scenario assumes an ultimate future rate of 1.6 -- which is greater than the average rate in Europe today. Demographers now know of nothing that is likely to reverse a long-term fertility decline which -- with a few episodic exceptions such as the postwar U.S. baby boom -- has been underway for over a century.
[End from Gray Dawn, Excerpts on Fertility]
(hits since June 22, 1999)
Get your own Free Home Page