All Commentary
Monday, November 1, 1993

The Coming Financial Collapse of Social Security

Mr. Ferrara is a fellow at the Heritage Foundation. He is the author of Social Security: Prospects for Real Reform (Cato, 1985). This first appeared in the Freeman, November 1993. 

The first officially recognized financial collapse of Social Security occurred in 1977. Government projections at that time showed, and everyone agreed, that without major changes Social Security would be unable to pay all of its promised benefits within a couple of years, with a yawning, continually growing deficit after that time.

Consequently, by the end of the year, Congress and President Carter dramatically increased Social Security taxes and trimmed benefits. Payroll tax rates increased repeatedly through 1990, for a total increase of 30 percent. Moreover, the maximum income to which this tax rate applied was increased sharply from $16,500 at the time, and indexed to increase every year thereafter. Today this maximum taxable income is $57,600.

The American people were assured over and over by President Carter, the Social Security Administration, and the rest of the Washington political establishment that these changes guaranteed the financial soundness of Social Security “for the rest of this century and well into[ the next one.”[1] But by 1980 Social Security was already in deep financial trouble again. The government’s annual financial report for the program showed that without a change in the law, the program might not be able to pay its promised benefits as early as 1981.[2]

To address this second financial crisis, a bipartisan commission headed by Alan Greenspan developed a package of tax increases and benefit cuts enacted early in 1983. The truth is that if the economy had continued to perform as it had in the 1970s, with high inflation and periodic sharp recessions, the system would have collapsed again within four years. But price inflation was sharply reduced in the 1980s, and the economy continued to grow for about eight years after 1982 without a recession, and then slid into a relatively shallow slump. Consequently, another quick collapse was avoided.

But Social Security’s long-term financial problems are another question. A key strategy of the Greenspan Commission was to develop a large surplus in the Social Security trust funds from 1990 to about 2010, to be used to help finance the retirement of the huge baby boom generation starting after 2010. However, the latest government projections show the expected surplus shrinking into relative insignificance. Moreover, the so-called Social Security trust funds in any event are not a store of financial reserves that can assure the future ability of the program to pay its promised benefits.

Repeated financial crises are an inherent feature of the Social Security system. The Social Security trust funds are essentially a sham that cannot assure future financial security. Today’s young workers will never receive the benefits currently promised to them by the program.

An Inherent Problem

Most people seem to imagine that Social Security operates like a traditional fully funded retirement program. In such a system, the tax payments of current workers are saved and invested to finance their own future benefits. As a result, a huge financial reserve is built up sufficient to finance accrued benefits at any point in time. This reserve is used to finance benefits during retirement years, while current workers at that time will be building up their own reserves to finance their own future retirement.

Social Security, by contrast, fundamentally operates on a pay-as-you-go basis. The tax payments of current taxpayers are not saved and invested to finance their own future benefits. Rather, most current tax payments are immediately paid out to finance the benefits for current retirees. Future benefits for present taxpayers are to be paid out of the future tax payments of future workers when today’s taxpayers are in retirement. Consequently, large cash reserves to finance benefits are never developed in such a system.

Such a pay-as-you-go system is quite vulnerable to any adverse development that may upset the delicate balance between expected future taxes and expected benefits. If unemployment rises, revenues from the payroll tax will fall from expected levels. If price inflation accelerates, indexed benefit payments will increase faster than expected. If retirees live longer than expected, benefit expenditures will again grow faster than projected. If the birth rate drops, fewer workers will be available to pay promised benefits in the future that are already paid for and relied on by current workers. These and many other possible developments can quickly tip a pay-as-you-go system into financial crisis, leaving it without sufficient funds to pay promised benefits.

None of this is a concern in the first generation under a pay-as-you-go system. When such a system is begun, a full generation of taxpayers begins to pay taxes, but there are no beneficiaries entitled to benefits based on past tax payments. In a fully funded system, these initial tax payments would have to be saved and invested to finance the future benefits of current workers. But, of course, these initial tax payments are not saved and invested under a pay-as-you-go system. Consequently, in the start-up phase of such a system, there is no concern over bankruptcy, or the inability of the program to pay promised benefits. To the contrary, the system is awash in unclaimed funds, and the only issue is how much to pay out in virtually free windfall benefits to early retirees. Since the first retirees pay little or nothing for their benefits, it is easy to pay them only what can be comfortably paid out of the initial incoming revenues. The beneficiaries will be grateful for the windfall benefits they receive.

After the first generation under such a system, however, this situation completely reverses. The retiring generation will then have paid taxes for an entire lifetime and will have built up enormous benefit claims. At this point, there are no more unclaimed surpluses and no more free benefits to pass out. The issue instead becomes whether taxes from current workers will be sufficient to finance promised benefits. If not, then Congress must raise taxes or cut benefits, in stark contrast to the vote-buying spending sprees of the first generation.

During its first 40 years, Social Security was in its start-up phase, and short-term financial solvency was not an issue. Instead, free windfall benefits were paid out to retirees.

But by the mid-1970s, sufficient benefit obligations had accrued to make financing a problem. Adverse economic developments soon developed to tip the system into financial failure. Inflation soared in the 1970s, sharply increasing benefit payments indexed to inflation. At the same time, periodic sharp recessions caused unemployment to rise and wage growth to fall, sharply reducing expected revenues. The combination of these economic difficulties caused the first two financial crises of the system described above. The primary cause of the third wave of financial collapse of Social Security, however, will be demographic, as discussed further below.

The Trust Fund Fraud

Even if Social Security attempted to depart from the principle of pure pay-as-you-go financing and developed a substantial trust fund reserve, future benefits would not be any more assured because of the essentially fraudulent nature of the Social Security trust fund system. Any remaining Social Security revenues after benefits are paid are lent to the federal government in return for new, specially issued government bonds which are held by the Social Security trust funds. The federal government then spends the borrowed Social Security revenues on other programs. The Social Security trust funds hold no assets other than these government bonds.

When Social Security revenues are insufficient to finance current benefits, the government bonds held by the trust funds are to be turned into the federal government for the cash needed to finance the benefits. But the government holds no cash or other assets to back up the Social Security bonds. The trust fund assets are claims against the federal government, government IOUs which will have to be financed out of increased federal taxes or increased federal borrowing. In other words, the trust funds are part of the national debt which must be paid when Social Security needs the money.

As a practical matter, these Social Security trust funds are nothing more than a statement of the amount that Social Security is legally authorized to draw from general federal revenues in the future, in addition to payroll tax revenues. Therefore, if the Social Security trust funds hold $1 trillion at some point, that statement even if true, would not mean that the Social Security system is financially sound. Quite to the contrary, it would mean that Social Security would have an additional $1 trillion claim against the taxpayers, in addition to the claim against them for payroll taxes.

Because the Social Security trust funds do not hold any real assets, just a claim against future tax revenues, a growing trust fund by itself does not mean that paying for the retirement of future generations will be any easier economically. It just means that more of this burden will be met out of income taxes and federal borrowing rather than payroll taxes.

The inherent financial problems of Social Security could be successfully addressed if the system were changed so that it accumulates reserves in a fully funded system and those reserves are invested in productive assets in the private sector. But that would require the government to own so much of the private sector through the Social Security trust funds that it would fundamentally change our entire economic system in an unacceptable way. Consequently, financial problems of Social Security can be solved only by shifting to a private system of decentralized investment accounts controlled by workers individually or through voluntarily organized groups.

The Looming Retirement of the Baby Boom Generation

As indicated above, the primary cause of the next foreseeable financial crash of Social Security is a destabilizing demographic problem. The huge baby boom generation is now entering middle age. Around 2010, this huge generation will start to retire, causing Social Security benefit expenditures to rise. This generation will continue to have a major effect on Social Security spending for the following 40 years.

But something has happened to make matters worse. Starting in the early 1960s, after the development of the birth control pill, birth rates in the United States declined precipitously. With the legalization of abortion in the 1970s, the fertility rate, or lifetime births per woman, fell below 2.0 in the early 1970s. It continued to decline to a low of about 1.7 per woman, eventually stabilizing at these low levels until the end of the 1980s.

As a result, the baby boom was followed by a baby bust. This means that at the same time the huge baby boom generation will be retiring, causing benefit expenditures to soar, the generation of workers that is supposed to finance their retirement payments out of current taxes will be relatively small.

The devastating impact of this demographic double whammy on Social Security is shown by the Social Security Administration’s own long-range financial projections. We can examine these projections under the most widely cited intermediate set of assumptions. Table 1 (on the following page) shows the results under these projections if we combine all three trust funds financed by the payroll tax the Old-Age and Survivors Insurance trust fund (OASI), the Disability Insurance trust fund (DI), and the Hospital Insurance trust fund (HI). These three trust funds together are referred to as the OASDHI trust funds.

With the huge baby boom generation entering its peak-earning middle-age years, and paying Social Security taxes on its earnings, the program should be doing quite well financially right now. Indeed, as indicated previously, the government’s plan is for Social Security to depart somewhat from its usual pay-as-you-go policy during this period and accumulate some substantial trust fund “reserves” to help finance the retirement of the boomer generation.

But Table 1 shows that under the “intermediate” assumptions, tax revenues for all three trust funds combined start to fall short of benefit promises in 2005, only twelve years from now. The federal government must cover these deficits by raising taxes, cutting other spending, or increasing the total federal deficit and federal borrowing. Besides tax revenues, the Social Security trust funds receive imputed interest income on their trust fund bonds. But since the federal government must pay the interest on the bonds, which it does by issuing additional bonds to Social Security in the amount of such interest, that interest does not help the government pay its promised Social Security benefits. To finance these benefits, the federal government must come up with the full amount of cash needed to close the deficit between Social Security taxes and Social Security expenditures. Effectively, the Social Security trust funds must begin redeeming some of their bonds for cash to cover these deficits, though counting the additional bonds received for interest each year the total trust fund assets may continue growing for a few more years.

As shown in Table 1, this annual Social Security deficit grows to almost $40 billion per year in constant 1993 dollars by 2010. By 2015, this annual deficit grows to $120 billion in 1993 dollars. By 2020, the annual deficit is an incredible $226.5 billion in 1993 dollars. The federal government again must either raise taxes, cut other spending, or increase the total federal deficit and federal borrowing by these amounts in order to pay all promised Social Security benefits, even before the Social Security trust funds are exhausted. The financial impact of the long-term Social Security financing crisis will start to hit less than a dozen years from now.

But that is not all. The federal government finances about 75 percent of Medicare Part B, also called Supplementary Medical Insurance (SMI), out of general revenues rather than payroll taxes. SMI pays doctors’ bills and other health expenses, while Medicare Part A, or Hospital Insurance (HI), which is financed entirely by payroll taxes, provides coverage for hospitalization. Table 1 also shows the projected amount of this general revenue contribution for SMI each year. The federal government must come up with these funds each year as well through either federal taxes, reductions in other spending, or increased government borrowing.

The general revenue contribution this fiscal year for SMI is about $48 billion. But Table 1 shows that by 2005 this will almost double to about $91 billion in constant 1993 dollars. By 2010 the general revenue contribution for SMI will grow to about $112 billion. Counting the $39 billion deficit in Social Security, this adds up to a total general revenue burden on the Federal government that year to finance all promised Social Security and Medicare benefits of about $150 billion in 1993 dollars. This is again before the Social Security trust funds are even exhausted.

By 2015, this general revenue requirement to pay promised benefits grows to $263.9 billion in constant 1993 dollars. Paying all promised benefits in that year for Social Security and Medicare alone would consequently create a total federal deficit almost as large as today’s federal deficit, unless taxes are raised or other spending cut. By 2018, the general revenue drain to pay all promised benefits would grow to $343.1 billion in 1993 dollars.

Counting the imputed interest on the government bonds held by the Social Security trust funds, all three trust funds combined actually hit their peak in nominal dollars in2011, as shown in Table 2. In constant 1993 dollars, the combined trust fund assets in 2011 would total about $915 billion. Yet, this is only 75 percent more than current Social Security trust fund assets, which will total about $525 billion for all four trust funds combined this year. Moreover, as also shown in Table 2, the projected trust fund assets in 2011 would only be sufficient by themselves to cover about one year and four months of projected Social Security expenditures. Yet, the current Social Security trust assets of $525 billion are sufficient to cover about one year and four months of projected benefit expenditures as well. Therefore, relative to the size of Social Security and the general economy, the total trust fund assets at their nominal peak in 2011 will not be significantly larger than today. Indeed, the largest the trust funds ever grow relative to expenditures under these projections is only one year and nine months worth of expenditures in 2005.

Consequently, the government seems to be failing to accumulate substantially larger Social Security trust funds. In fact, every year the projected growth in Social Security trust fund accumulations is getting smaller and smaller. What was cited as the incredible projected Social Security trust fund surplus a few years ago is now an incredible shrinking Social Security trust fund surplus. Within a few more years, we can expect the projected Social Security trust funds to deteriorate further, making the specter of the financial collapse of Social Security even more immediate.

After 2011, the projected Social Security trust funds decline, as the deficit between taxes and expenditures begins to exceed the annual interest on trust fund bonds paid by the issuance of new bonds. The federal government must continue to raise funds to cover the entire deficit between taxes and expenditures during this period, in the amounts shown in Table 1, effectively re deeming trust fund bonds equal to the entire deficit amount each year.

The total combined trust funds would be exhausted under these projections by 2019. Paying all promised Social Security benefits after that time would require huge payroll tax increases sufficient to close the deficit between taxes and expenditures in the system each year. The necessary tax increases are shown in Table 3. Paying all benefits promised to young workers entering the work force today would require a total Social Security payroll tax rate of about 27 percent, compared to 15.3 percent today. In other words, projected revenues, even under the intermediate assumptions, would be sufficient to cover only about half of promised benefits.

So far we have only discussed the “intermediate” projections. We must examine as well the projections under the so-called “pessimistic” assumptions. These assumptions are actually quite plausible. The “intermediate” assumptions assume regular price inflation of 4 percent per year, after a period of ups and downs, while the “pessimistic” assumptions assume regular price inflation of 5 percent per year. The intermediate assumptions similarly assume regular unemployment of 6 percent per year, while the pessimistic assumptions assume unemployment of 7 percent per year. A critical assumption for projected payroll tax revenues is the rate of growth of real wages. The intermediate assumptions assume real wage growth of 1.1 percent per year, while the pessimistic assumptions assume real wage growth of 0.6 percent per year. Actual experience in recent decades has been roughly halfway between these two assumptions.

Another critical assumption for future revenue is the fertility rate or rate of lifetime births per woman. The intermediate assumptions assume an ultimate regular rate of 1.9 while the pessimistic assumptions assume a regular rate of 1.6. Actual experience over the last 20 years has again been generally between these two rates, with experience in most other Western industri alized countries even lower.

Still another major assumption is life expectancy in retirement. The intermediate assumptions assume life expectancy at age 65 grows about 20 percent over the next 75 years. The pessimistic assumptions assume that such life expectancy grows 40 percent for males and 32 percent for females over this period. No one can know for sure, but given the potential developments in high technology medical care and other advances over the next 75 years, the pessimistic assumptions certainly seem quite plausible and may even underestimate the real possibilities. Indeed, from 1940 to 1990, life expectancy at age 65 grew about the same rate for males as assumed in the pessimistic assumptions and at an even faster rate for females.

Under these quite plausible “pessimistic” assumptions, tax revenues for all these trust funds combined start to fall short of benefits in 1996, only three years from now, as shown in Table 4. In that year, the federal government would have to come up with an additional $16.1 billion in 1993 dollars to pay promised benefits. The short-fall grows to $45.9 billion in 2000, and $100 billion in 2006, again in 1993 dollars.

Table 4 also shows that the annual general revenue contribution for SMI would grow to $77.6 billion in 1993 dollars by 2000. Counting the $45.9 billion Social Security deficit in that year, the total general revenue burden on the federal government to finance all promised Social Security benefits is $123.5 billion in today’s dollars. By 2006, just over a decade from now, this total general revenue requirement grows to $205.3 billion, again even before the trust funds are exhausted.

All three Social Security trust funds combined actually hit their peak in nominal dollars under these projections in 1999, as shown in Table 5. In constant 1993 dollars, the combined trust funds would total $527.6 billion in 1999, about the same as today. Indeed, the projected total 1999 trust fund assets would be sufficient to cover just over one year of benefit expenditures by themselves, compared to one year and four months for the current trust funds. Consequently, under these projections, the expected Social Security trust fund buildup to help fund the retirement of the baby boom generation never occurs.

The combined trust funds under these projections would be exhausted in 2007 just 14 years from now. Paying all Social Security benefits in 2010 would require an increase in the total Social Security payroll tax rate of about one-third, to about 20 percent from the present 15.3 percent, as shown in Table 6. By 2020, the total Social Security payroll tax rate would have to almost double to about 27 percent. In other words, total Social Security revenues by that date just 17 years from now would only be sufficient to pay about half of all Social Security benefits. In later years, paying all benefits promised to young workers entering the work force today would require a total Social Security payroll tax rate of over 40 percent, an increase of almost three times the current rate of 15.3 percent. In other words, projected Social Security revenues under these assumptions would be sufficient to cover only about one third of promised benefits.

But even this is not the worst plausible scenario. Economic performance in the 1990s could be like the 1970s, with unemployment rising and real wages falling behind rapidly rising inflation. At the same time, high-tech medical breakthroughs could rapidly advance old-age life expectancy beyond even the “pessimistic” assumptions, while fertility rates could fall to Western European levels at or below the pessimistic assumptions. Some or all of these quite possible developments would create even more gaping deficits, and require even more draconian tax increases to pay promised benefits.

But tax increases approaching two to three times current levels could never be adopted. Indeed, the current payroll tax is already far too high, seriously hampering economic growth and limiting job opportunities for today’ s workers. The payroll tax is basically a tax on employment. To the extent it is borne by employers, it discourages them from hiring. To the extent it is borne by workers, it discourages them on the margin from working as much as otherwise. The overall result is fewer jobs, less work, and slower economic growth. Here, as elsewhere, the result of taxing something, in this case employment, is that there is less of it.

Indeed, one study estimated that just the payroll tax rate increases that went into effect in 1988 and 1990, raising the total payroll tax rate from 14.3 percent to 15.3 percent, ultimately eliminated one million jobs and reduced GNP by $25 billion per year.[3] In a society supposedly deeply concerned about employment opportunities, the tax burden the government places on employment is absurd. The debate should be over payroll tax cuts, not increases. In any event, increases of the magnitude necessary to pay promised Social Security benefits in the future are clearly economically and politically infeasible.

As a result, today’s workers will never receive the benefits currently promised to them by Social Security, even though they are paying thousands of dollars each year for such promised benefits, and will do so for their entire careers.


The long term financial crisis of Social Security is not the only problem justifying the abolition of the system. The program’s payroll taxes are now so high that even if all the promised benefits are somehow paid, these benefits would still represent low, below-market returns, on the thousands of dollars today’s young workers must pay into the system each year for their entire careers. For most young workers the benefits would represent a real rate of return of around 1 percent or less, and to many the return would be close to zero, or even below zero. These workers could now receive much higher returns and benefits investing through the private sector. Average-income workers could accumulate over half a million dollars in today’s terms by retirement, and more than a million for two-earner-average-income families, for the same sums now paid into Social Security. This makes the inevitable inability of Social Security to pay even the currently promised inadequate benefits all the more troubling.

The Social Security benefit structure is also rife with inequities, paying some workers much less in returns on their tax dollars than others. Indeed, many workers are forced to pay for benefits under Social Security that they can never even qualify for or receive. Most fundamentally, Social Security deprives workers of the freedom to control the large sums they are now paying into the system each year. The long-term financial problems of Social Security should not be allowed to obscure these many other critical problems.

Social Security’s financial problems, as well as the other problems discussed above, can ultimately be solved only by shifting to a fully funded system of private savings and investment. Such a system would avoid the inherent vulnerability of pay-as-you-go financing and accumulate a vast reserve of economically productive private sector assets to back up benefits. Through such a system, young workers could also obtain the much higher returns and benefits now avail-able to them through the private market, in the process accumulating large family nest eggs in their retirement accounts. Workers would also have control and freedom of choice over the large sums they would pay into and accumulate in such a system. The same market returns would be available to everyone, and workers could tailor their benefit packages to suit their personal needs and preferences. They would never have to pay for benefits they did not need or could not even qualify for.

Such private systems have been adopted in recent years in Chile and, in part, in Great Britain, and have been broadly popular in both countries. There is no reason why such a system could not be adopted in the United States as well.

1.   This statement was quoted over and over again from the 1978 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds (Washington, D.C.: U.S. Government Printing Office, May 15, 1977), p. 3.

2.   1980 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Fund (Washington, D.C.: U.S. Government Printing Office, June 17, 1980).

3.   Aldona and Gary Robbins, Effects of the 1988 and 1990 Social Security Tax Increases, Institute for Research on the Economics of Taxation, Economic Report #39, February 3, 1988.

  • Peter Joseph Ferrara is an American lawyer, policy analyst, and columnist who is an analyst for The Heartland Institute. He is former general counsel for the American Civil Rights Union. A libertarian scholar, he is known for supporting privatization of the Social Security program.