Order Code 95-543 EPW Updated April 28, 2003 CRS Report for Congress Received through the CRS Web The Financial Outlook for Social Security and Medicare Geoffrey Kollmann and Dawn Nuschler Domestic Social Policy Division Summary The 2003 annual reports of the board of trustees of the Social Security and Medicare trust funds were released on March 17, 2003. The financial status of the Social Security program is little changed, but Medicare’s short-term financial difficulties are more severe and both programs continue to have projected long-range problems. Insolvency for the Disability Insurance (DI) part of Social Security is projected to occur in 2028, and for the retirement and survivors part, in 2044. On a combined basis, the two parts would become insolvent in 2042, 1 year later than projected last year and 13 years later than projected in 1997. Insolvency of the Hospital Insurance (HI) part of Medicare is projected to occur in 2026, 4 years earlier than projected last year. In the long run, the financing gap for both programs is larger than in last year’s forecast. Over the next 75 years, the Social Security program’s average deficit is 1.92% of taxable payroll, equal to about 14% of the program’s projected income. The average deficit of the HI program is 2.4% of payroll, equal to about 71% of the program’s projected income. By the end of the projection period, Social Security and HI combined expenditures are projected to be nearly double their combined income. Currently, taxes collected for Social Security and HI are higher than their expenditures. However, the situation will reverse in 2013 for HI and 2018 for Social Security, and expenditures are expected to exceed tax revenues thereafter. Overview of the Outlook for Both Programs Social Security’s financial condition is assessed annually by its 6-member board of trustees, consisting of the Commissioner of Social Security, three members of the President’s Cabinet, and two representatives of the public. The board’s reports have projected long-range financing deficits for the system since 1984. Although the trustees’ 2003 report continues to show a near-term buildup of Projected Points of Insolvency Social Security— Disability . . . . . . . . . . . . . . . . . . . . . Retirement & survivors . . . . . . . . . . Disability & retirement combined . . Medicare— Hospital insurance . . . . . . . . . . . . . . Congressional Research Service ˜ The Library of Congress 2028 2044 2042 2026 CRS-2 trust fund reserves, their “best estimate” for the next 75 years shows that on average Social Security’s expenditures will be 14% more than its income. By 2080 the income shortfall would be 50%. The near-term buildup of reserves would peak at $7.5 trillion in 2027, and then be drawn down as the post-World War II baby boomers retire. The trustees estimate that the disability fund would be exhausted in 2028 and the retirement fund in 2044. On a combined basis, the two funds would be exhausted in 2042. At that point, revenues would be sufficient to pay only 73% of scheduled benefits; by 2080 only 67% would be payable. Although the estimates imply that Social Security can be kept solvent for 39 years, they also show that the program’s taxes would begin lagging expenditures in 2018. At that point, the program would begin relying in part on general revenues in the form of interest payments to the trust funds. By 2028, interest payments and tax revenues would no longer be sufficient to cover the program’s expenditures, and the balances of the trust funds would begin to be drawn down. These reserves consist exclusively of Treasury bonds. By 2028, $1 out of every $5 of the program’s outgo would be dependent upon these claims against the general fund. The government has never defaulted on the bonds it records to its trust funds, but the magnitude of future claims has prompted many observers to ask where the government will get the money to cover them. The picture is more troublesome for Medicare. Its projected expenditures are rising at a faster rate than are Social Security’s. The HI part of the program is projected to become insolvent in 2026. On average over the next 75 years, its costs would be 71% higher than its income. By 2075, they would be 226% higher. While Supplementary Medical Insurance (SMI), the part of Medicare that pays for physician care, does not have the same type of financing problem as HI (SMI relies heavily on annual general revenue payments, not a fixed tax rate), its costs also are rising rapidly. As a share of GDP, SMI’s costs are projected to rise from 1.12% today to 4.14% by 2075. By 2055, the combined costs of HI and SMI are projected to exceed that of Social Security. Background Social Security is the Nation’s largest retirement and disability program providing cash benefits to 46 million retired and disabled workers and to their dependents and survivors. Medicare provides 40 million of them with health insurance. In 2000, Social Security accounted for an estimated 38% of the income of the elderly (people age 65 and older). Medicare provided more than 95% of them with basic health coverage. Today, one out of six Americans receives Social Security; one out of seven receives Medicare. In 2002, an estimated 156 million workers paid taxes to support the two programs. Workers gain eligibility for Social Security and HI by working in jobs where Social Security and HI taxes are levied. They pay a flat-rate tax of 7.65% on their earnings (6.2% for Social Security and 1.45% for HI), which is matched by their employers. The self-employed pay a tax of 15.3% (with adjustments that effectively reduce the rate). The Social Security portion is levied on earnings up to $87,000 in 2003; the HI portion is levied on all earnings. In 2003, payroll taxes comprise 85% of Social Security’s estimated income and 85% of HI’s. The rest comes mostly from government credits, the largest of which is for interest on federal securities held by their trust funds. There is no SMI tax; 76% of its estimated 2003 income comes from general revenues of the government and 24% from premiums paid by enrollees ($58.70 per month in 2003). CRS-3 The taxes and premiums people pay flow into the Treasury, with each program’s share credited to separate trust funds (one for retirement and survivors benefits, another for disability, and two others for Medicare). The government credits the money to the Social Security and Medicare trust funds by recording new interest-bearing federal securities to the appropriate fund (these securities earn interest at the average rate prevailing on outstanding federal bonds with a maturity of four years or longer). When the government makes payments, it writes some off. While these securities are represented as assets for the trust funds, they also represent liabilities for the government. Their primary role is to be reserve “spending authority.” What this means is that as long as a trust fund has a positive balance, the Treasury Department is authorized to make payments for it from the Treasury; however, the funds themselves do not contain the resources to do so. The Social Security Picture For more than three decades after Social Security taxes were first levied in 1937, the system’s income routinely exceeded its outgo, and its trust funds grew. However, the situation changed in the early 1970s. Enactment of major benefit increases in the 1968 to 1972 period was followed by higher inflation and leaner economic growth than had been expected. Prices rose faster than wages, the post-World War II baby boom ended (leading to a large cut in projected birth rates), and Congress adopted faulty benefit rules in 1972 that overcompensated new Social Security retirees for inflation. These factors combined to sour the outlook for Social Security and it remained poor through the mid1980s. Before 1971, the balances of the trust funds had never fallen below one year’s worth of outgo. Beginning in 1973, the program’s income lagged its outgo and its trust funds declined rapidly. Congress had to step in five times to keep them from being exhausted. Although major changes enacted in 1977 greatly reduced the program’s longrun deficit, they did not eliminate it, and the short-run changes made by the legislation were not large enough to enable the program to withstand back-to-back recessions in 1980 and 1982. A disability bill in 1980 and temporary fixes in 1980 and 1981 were followed by another major reform package in 1983. These 1983 changes, along with better economic conditions, helped to alter the picture. Income began to exceed outgo in 1983 and the trust funds grew substantially. Cumulatively, the changes were projected to yield $96 billion in surplus income by 1990, and to raise the trust funds’ balances to $123 billion. The funds actually were credited with $200 billion in surplus income by 1990, and their balances reached $225 billion by the end of that year. By the end of 2002, the trust funds’ balances had risen to more than $1.4 trillion. This is equivalent to 288% of estimated expenditures in 2003 (or more than 2½ years’ worth). The long-range picture for Social Security has worsened considerably since 1983. By raising Social Security’s age for full benefits from 65 to 67, subjecting benefits to income taxes, and making federal and nonprofit workers join the system, Congress had attempted in 1983 to eliminate the long-run problem. In fact, projections made then showed that it had, at least on average, for the following 75 years. However, the average condition of the two trust funds did not represent their condition over the entire period. The funds were not shown to be insolvent at any point, but their expenditures were CRS-4 expected to exceed their income in 2025 and to remain higher thereafter. Simply stated, 40 years of surpluses were to be followed by an indefinite period of deficits. With each passing year since 1983, the trustees’ 75-year averaging period has picked up one deficit year at the back end and dropped a surplus year from the front end. This, by itself, would cause the average condition to worsen. In subsequent reports, however, birth rate and wage growth assumptions were lowered, and actuarial methods were revised, causing further deterioration in the outlook. A small long-range deficit appeared in the 1984 report and the gap grew larger (with the point of insolvency generally coming closer) in subsequent reports. Projections reported over the last six years, however, have shown small improvements in part due to favorable near-term economic conditions. The 2003 report shows an average 75-year deficit equal to 14% of the program’s income, and projects that the trust funds would become insolvent in 2042 (one year later than projected last year). As a percent of the Nation’s payrolls, their income would average 13.78%, their outgo, 15.7%, and the deficit would be 1.92% (compared to 1.87% reported last year). This average deficit is just a little less than the deficit tackled by Congress in 1983. However, an examination of the back end of the projection period shows chronic revenue shortfalls. By 2080, the yearly deficit would be equal to 50% of the system’s projected income. These long-range projections assume that GDP (adjusted for inflation) will rise annually at rates ranging from 2.9% in 2003 to 1.8% in 2080, wages would rise at an ultimate rate of 4.1% per year, the cost of living would go up at a rate of 3.0%, unemployment would average 5.5%, and that Social Security retirement benefits would fall in relative terms as the age at which full benefits are payable rises from 65 to 67 over the 2000 to 2022 period. These assumptions by themselves would seem to bode well for the system; however, looming demographic shifts are projected to overwhelm them. During the next two decades, the baby boomers will be in their prime productive years, and the baby-trough generation of the 1930s will be in retirement. Together these factors will lead to a stable ratio of workers to recipients. However, as the baby boomers begin retiring around 2010, this ratio will erode quickly. By 2025, most of the surviving baby boomers will be 65 and older. By then, the number of people 65 and older is projected to grow from 35.9 million in 2000 to 63 million, an increase of 76%. The number of workers is projected to grow from 153.5 million to 178.4 million, or by only 16%. Consequently, the number of covered workers per beneficiary will decline from 3.3 today to 2.2 by 2030. By 2075, there will be 1.8 workers per beneficiary. These declining trends are reflected in a new portrayal in the most recent Trustees report that shows that, if projections are made beyond the 75-year window, the status of the program is even more dire (e.g., instead of 1.92% of taxable payroll over the next 75 years, the long-range deficit looking indefinitely into the future would be 3.77% of taxable payroll). Under this forecast, the trust funds (on a combined basis) would be credited with surplus income through 2027 bringing their balances to a level of $7.5 trillion. They would decline in 2028 and thereafter, and would be depleted by 2042. However, tax receipts begin lagging outgo much sooner, in 2018. At that point, the program would have to rely on the interest credited to its trust funds for part of its income, which would have to be drawn from general revenues. In 2028, the reserve balance of the trust funds would begin to be drawn down, and $1 out of every $5 of the program’s outgo would be dependent upon general revenues. The government has never defaulted on the securities CRS-5 it posts to its trust funds, but the magnitude of these potential claims has prompted many observers to ask where the government will find the money to cover them. In the absence of surpluses for the rest of the government’s operations, policymakers would have three options: raise other taxes, curtail other spending, or borrow money from the financial markets. Economists argue that if the surplus taxes projected for the next 15 years were to cause the government to reduce the federal debt held by the public, more money would be available in the financial markets for investment, which could lead to greater economic growth. If this occurred, extracting resources from the economy in the future to honor Social Security claims would not necessarily be so burdensome. Said another way, if one accepts the premise that reductions in the federal debt held by the public today will increase the resources available for investment, then surplus Social Security taxes today could help build a higher economic base from which to draw the needed resources. However, rolling surplus Social Security taxes into Treasury bonds will not by itself reduce government borrowing from the markets. Reductions in the debt occur when the government runs an overall or unified budget surplus, not when one of its programs generates surplus taxes. Also, if economic growth were enhanced in the coming decades by reductions in government debt, Social Security’s problems would not necessarily be resolved. Its costs would grow as the economy grows (since economic growth would likely result in higher wages, which in turn would lead to larger benefit claims). Further, as their numbers swell, the baby boomers and subsequent retirees will raise financial demands on all retirement systems, not only Social Security. The goods and services to be consumed by society cannot be stockpiled in advance, and the economy will have to adjust. The question is whether the adjustment would be mild or severe. The Medicare Picture The trustees present a worsened near-term picture for Medicare and an even more troublesome long-range one. Although major constraints in Medicare payment rates were enacted as part of the Balanced Budget Act of 1997 (P.L. 105-33), HI’s rapid growth is projected to continue indefinitely. Those changes and an improved economic outlook extended the HI trust funds’ projected insolvency point by 25 years (from 2001 to 2026), and cut the average 75-year deficit by nearly half (from 4.32% of taxable payroll to 2.4%). However, the remaining deficit is large. On average, HI’s projected costs would be about 71% higher than its income. By 2075, its costs would be more than three times larger than its income. This pessimistic outlook reflects not only the persistent high rate of inflation in the health sector of the economy and growth in the quantity of services provided, but also the general graying of the population. Whereas there are about four workers per HI beneficiary now, there will be an estimated 2.4 workers per beneficiary in 2030. Shown as a percent of the Nation’s payrolls, HI’s costs would rise from 2.92% in 2003 to 5.1% in 2030 and 11.19% in 2075. On average for the 75-year period, HI income would be 3.37% of payroll, HI outgo would be 5.77%, and the deficit would be 2.4%. Although this average deficit is only a little larger than that for Social Security, it is much CRS-6 larger given the size of each program. The average gap between HI’s income and outgo equals 71% of the program’s income in contrast to a gap of 14% for Social Security. Because SMI is financed with general revenues and premiums that are reset annually, it does not have the same type of financing problem as HI. However, its expenditures are expected to rise even faster than HI’s. Projections show that SMI’s expenditures as a share of GDP would double by 2030 (rising from 1.12% today to 2.32% in 2030). From 2003 to 2075, the combined costs of HI and SMI are projected to rise from 2.6% to 9.1% of GDP. The Combined Scenario Although the trustees’ 2003 projections show that Social Security overall will generate sufficient taxes to cover its commitments for the next 15 years, and that the trust funds will have a positive balance until 2042, the long-range outlook is not sanguine. HI’s problems are more imminent, as insolvency is projected for 2026. Resources could be reallocated to HI from Social Security; however, this would only move Social Security’s problems closer. If Social Security and HI are considered together, their outgo as a percent of the Nation’s payrolls would rise from 13.81% today to 21.98% in 2030 and 30.98% in 2075, levels that contrast sharply with a combined tax rate that is set now in the law at 15.3%. As a percent of GDP, outgo for Social Security and HI combined would rise from 5.82% today to 8.86% in 2030 and 11.85% in 2075. Including SMI would raise it from 6.94% today to 11.18% in 2030 and 15.99% in 2075. However, the taxes and premiums dedicated to support the programs (i.e., payroll taxes, proceeds from the taxation of benefits, and SMI premiums) are projected to hover only in the 7% of GDP range throughout the period. Although these projections are based in part on economic assumptions, they are driven by demographic factors — the post-World War II baby boom, the subsequent birth dearth, and the general aging of society. They imply that to restore long-run solvency, increases in income, cuts in expenditures, or some combination thereof, need to be made. Beyond possible changes to the programs themselves, important unknowns that can alter the outlook include whether an effective means can be found to rein in the spiraling cost of medical care generally and whether future technological advances will propel productivity. Also unknown and little understood is the effect of potential shifts in society’s wants and needs: from raising families, buying houses, and educating children to meeting the health and service demands of an older population. Will the higher future costs of Social Security and Medicare place large strains on the economy or merely reflect a shift of the Nation’s consumption priorities?