February 2, 2000 | Center for Data Analysis Report on Social Security
The Social Security system, according to its own actuaries, faces a financial crisis of immense proportions. The Office of the Chief Actuary of the Social Security Administration (SSA) is projecting that within 15 years, the system will begin taking in less money than it needs to pay the benefits promised to participants. Moreover, within 30 years, it will have only enough money to pay less than 75 cents of every dollar of benefits it promises. By 2075, using its intermediate assumptions, Social Security will be running an annual deficit of $562 billion (in 1999 inflation-adjusted dollars).2
Despite these dire predictions, supporters of the current Social Security system blithely assert that the problem can be solved by faster economic growth. Without much evidence, they claim that current economic projections are entirely too pessimistic and that the financial shortfall will disappear if the numbers are made more optimistic.
Robert Reich, former Secretary of Labor in the Clinton Administration, for example, has called the SSA economic projections "wildly pessimistic."3 Economist James K. Galbraith claims that if higher growth rates were substituted for the SSA's projected rates, "future deficits disappear without any cuts in benefits or increases in taxes."4 These assertions were bolstered by the Department of Commerce's October 1999 adjustment in its economic growth figures, which shows that the U.S. economy grew at a faster rate from 1959 through 1998 than previously estimated.5
Regrettably, claims that Social Security can be saved by faster economic growth are wrong. If anything, the projections underlying the Social Security Administration's forecasts are likely to be overly optimistic.6 And even if the SSA massively underestimates the future rate of economic growth, higher growth will have little impact on the system's solvency. By some measures, faster growth could even add to Social Security's problems.
According to an analysis of Social Security's own projections, if the inflation-adjusted growth rate of average wages over the next 75 years increases over the current forecast by 56 percent (or from the SSA's intermediate or "best guess" forecast of 0.9 percent annually to its most optimistic forecast of 1.4 percent), then:
There would be only a modest decrease in Social Security's deficit as a share of taxable payroll.
Under the scenario of faster economic growth, by 2075 Social Security's annual shortfall would equal 5.12 percent of taxable wages, compared with 6.54 percent under the SSA's intermediate scenario.
Measured in terms of inflation-adjusted dollars, faster economic growth would cause an increase in Social Security's annual shortfall after 2055.
Although economic growth would increase revenues, it would cause an even larger increase in the system's benefit obligations over the long term. By 2075, the annual Social Security deficit under the scenario of rapid economic growth is $629.9 billion
(in 1999 inflation-adjusted dollars), $67.8 billion higher than the deficit under the SSA's intermediate projections.
In short, critics of reform are mistaken if they think faster-than-predicted economic growth will help the Social Security system avoid its financial crisis. Without fundamental reform that allows workers to invest their own Social Security taxes, deep benefit cuts or steep tax increases will be required, regardless of how rapidly wages grow.
However, long-term demographic trends are the driving force behind the financial crisis that now threatens the future of the Social Security system. Over the period 1950-1999, the life expectancy at birth of an average worker increased 8.2 years, from 68.35 to 76.55. During the same period, the birth rate dropped precipitously. In 1950, American women could expect to give birth to an average of three children; today, the average number of children per woman is just two. Social Security projects these demographic trends to continue. The SSA estimates that by 2075, life expectancy will have reached 81.85 years and U.S. women will give birth to an average of 1.9 children each over their lifetimes.7
The increase in longevity coupled with the declining birth rate is increasing the number of Social Security recipients relative to the number of younger workers whose payroll taxes support their benefits. In 1945, two persons were collecting benefits for every 100 workers paying Social Security tax. By 1999, this number had increased to 30 beneficiaries for every 100 workers; and the SSA projects that by 2075, there will be 54 persons receiving Social Security benefits for every 100 persons paying into the system--even though all of the baby boomer generation will have died.8
Put simply, demographic trends threaten the future solvency of the Social Security system by increasing the ratio of retirees who receive Social Security benefits to workers who pay Social Security taxes.
The major economic variable that affects the solvency of the Social Security system is growth.9 Moreover, because Social Security taxes are levied only on labor income, the key relevant measure of economic growth is the real growth rate of average wages and self-employment income.10
Consistent with the SSA's own approach, this analysis attempts to capture the impact of economic growth caused by many factors (such as new technology, improvements in skills and training, deregulation, and capital accumulation) to the extent that this economic growth is reflected in greater worker productivity and higher wages. An increase in other sources of income, such as dividends or rent, does not increase Social Security payroll tax revenues (or benefits).
An increase in average wages has two effects on the balance between Social Security expenditures and revenues. First, an increase in the average earnings of workers and the self-employed increases the payroll tax base and the amount of payroll tax revenue that is available to pay benefits. This improves the solvency of the Social Security system.
However, there is also a second effect. According to current law, the Social Security benefit to which a retiree is entitled is calculated as a proportion of the value of the "Average Wage Index" on the year of retirement.11 Each year, the Average Wage Index is calculated by the Social Security Administration using the average value of wage, salary, and self-employment income for all workers in the economy. If, in a certain year, the Average Wage Index increases (due to the faster growth of wages in the real economy), the Social Security Administration must pay higher benefits to workers who are retiring in that year.
While this effect is initially small in the first year, the impact becomes greater over the long term as successive cohorts of workers retire, each of whom must be paid higher benefits as the average amount of wages rises because of this increase in the Average Wage Index.
In short, an increase in the average wage earned by workers not only increases the revenues that the Social Security Administration has available to pay benefits, but also increases the benefit obligations that the Social Security trust fund must meet. Over the long term, the positive revenue impact of wage growth comes close to being cancelled out by the increase in benefit obligations that it also generates.
The Projected Effect of Faster Economic
The Social Security Administration's "best guess" of future economic and demographic conditions is represented by the "Intermediate Assumptions" scenario that SSA publishes each year in the Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds. These projections include estimates of future mortality, fertility, inflation, economic growth, and interest rates that are used to forecast the expected revenues and expenditures of the Social Security system over the next 75 years.
The Social Security Administration also makes projections based on a pessimistic "High Cost Assumptions" scenario and an optimistic "Low Cost Assumptions" scenario. These two sets of forecasts represent, respectively, fiscal outcomes where all of the underlying demographic and economic conditions are much less favorable and a scenario in which conditions will be much more favorable than the Social Security Administration projects.12
Charts 2 through 6 show the future fiscal condition of the Social Security system under a scenario in which wages grow at the rate projected by the Social Security Administration in its "Low Cost Assumptions" (or extremely optimistic) scenario, but where all other economic and demographic variables (fertility, life expectancy, inflation, and interest rates) occur exactly as projected in its "Intermediate Cost" (or "best guess") scenario. In short, the projections hold constant all of the SSA's demographic and economic projections except for average wages, which the author of this analysis assumes will grow at the rate projected in the SSA's most optimistic scenario.13
The projections shown in Charts 2 through 6 assume that average wages grow at an annual rate of 1.4 percent after inflation over the next 75 years, rather than the 0.9 percent rate projected in the SSA's "Intermediate Cost" scenario. This 1.4 percent rate of wage growth was selected because it is the rate used by the Social Security Administration in its "Low Cost" or optimistic set of projections. In other words, the scenario reported in these charts assumes that over the next 75 years, wages grow at a rate that is 56 percent faster than the rate projected by the Social Security Administration in its "best guess" about future economic conditions.
To give an idea of the magnitude of this difference in growth rates, consider that if wages grow at the 1.4 percent rate projected in Social Security's most optimistic scenario, by 2075 the average wage per worker in the United States would be $83,374 (in terms of 1999 dollars), rather than the $57,438 predicted under the SSA's "Intermediate Cost" set of projections.
The degree to which the projection of 1.4 percent annual real growth in wages is extremely optimistic becomes clear when one considers the historical rates of growth of wages. According to the Social Security Administration's 1999 annual trustees' report, "the average annual rate of change in average real earnings for the total U.S. economy was an increase of 0.9 percent for the 40 years 1958-97."14 The report also states that the average annual real growth in wages during the decades 1968-1977, 1978-1987, and 1988-1997, respectively, was 0.4 percent, 0.0 percent, and 0.8 percent.
In other words, under the Social Security Administration's "best guess" scenario, wages will grow at least 12.5 percent faster than they actually have in any decade since the 1960s. The optimistic scenario used in this analysis assumes a 1.4 percent annual real growth rate in wages that lies more than 56 percent above the growth rate of U.S. wages in the entire period since the mid-1950s.15
As can be seen from both Chart 2 and Chart 3, the faster growth in wages pushes back the date of insolvency by a mere two years, from 2014 to 2016. By 2035, the annual deficit is equal to 3.85 percent of taxable payroll under the scenario of rapid economic growth, only 1.19 percentage points below the deficit of 5.04 percent of taxable payroll that occurs under the Social Security Administration's baseline assumptions. By 2075, Social Security's annual shortfall as a percent of payroll is equal to 5.12 percent under the assumption of faster wage growth, compared with 6.54 percent under the SSA's intermediate projections.
Chart 2 shows the annual expenditures and revenues of the Social Security system as a percent of taxable payroll under two sets of assumptions: the Social Security Administration's "Intermediate Cost" (or "best guess") assumptions and a scenario in which wages grow at an annual rate that is 56 percent faster than under the SSA's "Intermediate Cost" assumptions. Chart 3 reports Social Security's annual deficit or surplus as a percent of taxable payroll in each of these two scenarios. Expressing Social Security's revenues and expenditures as a percent of taxable payroll enables us to examine the surplus or deficit relative to the tax base that is potentially available to fund the system's needs.
The implications of the projections shown in Chart 4 can be expressed in terms of the increase in payroll taxes needed to balance the Social Security system under each set of assumptions.By 2040, even under the scenario in which wages grow at a rate that is 56 percent faster than is currently projected, the Social Security OASDI tax rate would have to be raised by 30 percent (or 3.76 percentage points) over the current rate of 12.4 percent to pay promised benefits. By 2075, OASDI payroll tax rates would need to be increased by 5.12 percentage points above the current level of 12.4 percent to 17.52 percent in order to generate sufficient revenues to pay the benefits promised under current law.
Social Security's financial position also can be measured in terms of inflation-adjusted dollars. While the system's deficit or surplus as a percent of taxable payroll is a good measure of the balance of the system relative to the potential tax base, expressing the balance in terms of inflation-adjusted dollars gives a very good sense of the absolute financial burden of any imbalance.
Chart 4 shows the effect on Social Security revenues and expenditures (in terms of 1999 inflation-adjusted dollars) of a 56 percent increase in the growth rate in real wages above the rate projected in the SSA's "Intermediate Assumptions" scenario. As Chart 4 shows, the increase in economic growth causes an increase in both Social Security benefits and taxes over the next 75 years.
As can be seen in Chart 5, an increase in the rate of wage growth has little impact on Social Security's financial health, measured in 1999 inflation-adjusted dollars. If the economic growth rate is raised by 56 percent, the Social Security system's insolvency date--when revenues can no longer cover benefits--is pushed back by a mere two years (from 2014 under the "Intermediate Assumptions" scenario to 2016).
In fact, increases in economic growth lead to increases in Social Security deficits during the years after 2055. Under the scenario of a 56 percent jump in the growth rate of wages, by 2075 Social Security's annual deficit is $629.9 billion (in 1999 inflation-adjusted dollars), an amount that is $67.8 billion higher than the $562.1 billion deficit existing under the SSA's "best guess" assumptions. This hike in the deficit occurs because faster economic growth eventually results in larger dollar increases in benefits than in revenues.
According to projections made by the Social Security Administration's own Office of the Chief Actuary, even a massive 56 percent increase in the rate of economic growth above projected levels will do little to solve Social Security's financial crisis. If the average annual growth rate of wages over the next 75 years is increased from 0.9 percent to 1.4 percent, by 2035 the Social Security system will still be able to pay out only 77 cents for every dollar of promised benefits. In that year, benefits must be cut by 23 percent, or payroll taxes increased by 30 percent from the levels promised in current law, if the system is to remain in balance.
In fact, by some measures, the projections from the Office of the Chief Actuary suggest that faster economic growth actually may hurt rather than help the system's long-term financial health. According to the Social Security Administration's projections, if wages grow at an annual real rate of 1.4 percent, by 2075 the OASDI program will be running annual deficits of about $630 billion (in inflation-adjusted 1999 dollars). By comparison, the program's annual operating deficit will be "only" $562.1 billion by 2075 if wages grow at only 0.9 percent per annum.
The long-term financial crisis facing Social Security is very real and cannot simply be wished away by assuming future rates of economic growth that are higher than any figure that is justified by the evidence. Even if wages grow for the next 75 years at a rate 56 percent above the actual growth rate achieved in the period since World War II, Social Security will continue to face dire financial problems. Under the current pay-as-you-go structure, these problems can be solved only by drastic cuts in benefits or by massive hikes in payroll taxes, which would hurt current and future generations and drive down their rates of return from the program.
The real solution to both the solvency crisis and the rate of return crisis confronting Social Security is to create a system that allows workers to fund their retirement by investing their own payroll taxes rather than relying on the payroll taxes paid by younger generations.
Gareth G. Davis is a former Policy Analyst in the Center for Data Analysis at The Heritage Foundation.
1. The author wishes to acknowledge the contribution of William Butterfield, who worked as an intern in the Center for Data Analysis during the summer of 1999 and co-authored several sections of this paper.
3. Quoted in James K. Galbraith, "I Don't Want to Talk About It," published on the Electronic Policy Network (EPN) Web site at http://www.epn.org/galb/jg980410.html.
6. Indeed, in November 1999 the technical panel advising the SSA's Advisory Board recommended that the Social Security Administration's Office of the Chief Actuary adjust the assumptions that it uses to forecast the program's future financial status. The aggregate impact of these recommended changes would be to increase the program's 75-year deficit by 26 percent. See Social Security Administration, The 1999 Technical Panel of Assumptions and Methods, Report to the Social Security Advisory Board, November 1999.
9. The Social Security Administration reports that changes in the rate of inflation can have a minor short-term impact on the solvency of the system. The impact of inflation is minor and transitory because, under current law, all benefits and all tax variables are adjusted annually for inflation. For reasons of brevity, the role of inflation is not analyzed in this study. For a discussion of the impact of inflation on SSA finances, see 1999 Trustees' Report, Section II.G.
10. The term "economic growth" as used in this paper refers to growth in per capita income. This statistic measures the change in average income available per person. The total size of a country's national income can be measured by multiplying per capita income by population. Among the ways in which the size of a country's economy can grow are an increase in the size of the country's population (due to demographic changes) and an increase in income per person. The purpose of this paper is to isolate the impact of faster economic growth as measured by changes in per capita incomes, rather than the impact of demographic variables (such as birth rates or life expectancy). The impact of demographic factors, which may affect the size of the U.S. economy by changing the population, will be examined in a subsequent Center for Data Analysis paper. The preliminary analysis to be published in this paper suggests that the SSA's demographic assumptions may be overly optimistic.
11. The glossary of the 1999 Trustees' Report defines the Average Wage Index as the "average amount of total wages for each year after 1950, including wages in noncovered employment and wages in covered employment in excess of the OASDI contribution and benefit base."
13. The Social Security Administration uses a static model to estimate the future revenues and expenditures of the system. This means that the SSA model does not generate "dynamic" economic feedback between macroeconomic variables. For example, according to the SSA's model, an increase in interest rates has no impact on economic growth or employment. Because of this static construction, it is possible for the Social Security Administration to examine the fiscal impact of a faster post-inflation growth rate in wages in isolation, while assuming that all other economic variables remain unchanged. If the SSA used a dynamic model that took account of the dynamic interaction between economic variables, it is likely that the problems faced by the system would actually be worse than portrayed under the current static projections. For example, the huge financial deficits forecast under current law would likely drive employment below baseline levels, further worsening the system's financial status.
15. In October 1999, the U.S. Commerce Department reported that adjustments to its previously reported data revealed that real gross domestic product grew at an annual rate of 3.4 percent over 1959-1998 rather than at the previously estimated rate of 3.2 percent, for a difference of 0.2 percent. This adjustment was not reflected in the 1999 Trustees' Report, which was published in March 1999. It is not yet apparent that these new data will support the view that the Social Security Administration has underestimated the growth rate of average real earnings over the period 1959-1997. However, if re-examination of available data suggests that average wages grew at a rate that is 0.2 percentage points faster than previously thought, and if the Social Security Administration uses these data to boost projections of future wage growth by 0.2 percentage points, then the "new" projected growth rate of wages would be 1.1 percent. This 1.1 percent growth rate still lies well below the 1.4 percent growth rate used in the "optimistic" scenario reported in this analysis. Given the minimal impact of an increase in the growth rate to 1.4 percent, it is highly likely that adjusting projections of future growth rates to conform with newly available data for 1959-1998 will have a minuscule impact on the program's projected solvency. The SSA's 1999 Technical Panel of Assumptions and Methods drew on the available evidence and recommended that the projected growth rate in average wages be increased to 1.1 percent; however, it also recommended a variety of other changes in the economic and demographic assumptions used by the SSA in making their annual projections. The net impact of implementing all of the Technical Panel's suggestions would be to increase the program's long-term projected deficit by 26 percent.