Social Security faces an enormous future deficit: Between today and 2075, the inflation-adjusted shortfall is projected to reach a staggering $20 trillion. Although the problem with the current system is due in part to changes in demographics, the root of the problem lies in the fact that the Social Security system itself is poorly designed. Workers, particularly those under age 50, are slated to receive very low benefits in return for a record amount of payroll taxes they send to the federal government.1 These workers could enjoy substantially greater levels of retirement income if they were allowed to place the bulk of their payroll taxes in professionally managed individual retirement accounts,2 which historically have had significantly higher rates of return.
Defenders of the current system generally admit that personal accounts would make workers better off, but they also argue that the "transition cost" of privatizing would be significant. More specifically, because a major share of the payroll taxes now used to pay benefits would be invested instead in private accounts, policymakers would need to find several trillion dollars to finance benefits for current retirees and those nearing retirement (and, therefore, too old to take advantage of private accounts).
These critics are only correct in a technical sense because they omit the other side of the story. Yes, privatization entails a sizable transition cost, but keeping the current system in place and putting it on sound footing would involve a large transition cost as well. The important question to ask is whether the price tag for moving to a private system is smaller or larger than the amount of money lawmakers would have to find to fulfill the promises of the current system. As it turns out, privatization is less expensive. The level of savings, needless to say, would depend on the particular plan that lawmakers ultimately adopt.
HOW TO MEASURE SOCIAL SECURITY'S LONG-TERM DEFICIT
The Social Security Administration produces estimates of annual income and expenses for every year through 2075.3 These data are included in the appendices to this paper: Appendix A shows the annual data in nominal (non-inflation adjusted) dollars, while Appendix B expresses the same information using inflation-adjusted dollars and present-value calculations. In order to calculate Social Security's annual funding gap, the appendices combine the projected cash payments into the system (payroll taxes and the income taxes that the elderly pay on their benefits) and then subtract estimated benefit payments.
As seen in the accompanying charts, the nominal deficit is immense. Beginning in 2010, with a deficit of under $1 billion, the system's funding gap grows rapidly, reaching nearly $500 billion in 2025, $1 trillion in 2035, and more than $7.5 trillion in 2075. Between 1998 and 2075, the cumulative shortfall in nominal dollars would reach $143 trillion.
These figures give a false impression, however, because they do not account for inflation. The Social Security Administration estimates that long-term annual inflation will be about 3.5 percent. Appendix B adjusts the annual figures with the inflation estimates used by the Social Security Administration. The long-term deficits fall dramatically when expressed in today's dollars, but the gap is still huge. The inflation-adjusted deficit reaches $200 billion in 2025, $300 billion in 2035, $400 billion in 2056, and $500 billion in 2068. The total gap between now and 2075 is more than $20 trillion in 1998 inflation-adjusted dollars.
Present value is another way to calculate the long-term debt. In addition to considering the effects of inflation, present value calculations recognize that a dollar today is worth more than the same dollar--even after adjusting for inflation--in the future. In other words, because money today can be invested to earn a return, the unfunded liabilities of the Social Security system could be offset completely if lawmakers came up with a big enough pile of cash to invest today. The "good" news is that the present value of Social Security's unfunded promises is "only" $5.2 trillion. The bad news is that collecting that much money today would require imposing tax rates of more than 100 percent on everyone in the country. Moreover, the viability of such an approach rests on politicians' prudently investing the money and using all the funds--interest and principal--to do nothing except pay for promised benefits. Chart 3 shows the annual present value deficit using a real interest rate of 2.8 percent (the rate used by the Social Security Administration).
Q & A ABOUT SOCIAL SECURITY'S DEBT
Q. When calculating Social Security's total debt, should short-term surpluses be used to offset a portion of the future deficits?
A. If the government saved excess payroll tax revenues, then the answer would be yes. Surpluses are spent on other programs, however, and the Social Security Trust Fund receives IOUs from the Treasury in exchange. These IOUs represent claims on future taxpayers, not a store of wealth. The law could be changed so that the surpluses were invested in private, income-producing assets (much as occurs with the pension systems of state government employees). Under this approach, the Trust Fund would hold real assets that properly could be used to offset long-term debt. Because of widespread concerns that politicians would use such a fund to finance pet projects and engage in misguided industrial policy, however, this generally is not seen as a desirable option.
Q. If the Social Security Administration provided annual spending and revenue estimates beyond 2075, wouldn't the system's total shortfall be higher than $20 trillion?
A. Yes. Annual inflation-adjusted deficits are more than $500 billion, and continue to rise in each of the years leading up to 2075. There is no way to tell how long this trend will continue, but the cumulative shortfall certainly is far greater than $20 trillion; it may be unlimited. Likewise, the present value debt also is higher than the $5.2 trillion described above. The only "good" news is that present value debt peaks at about $110 billion and then begin a gradual decline, falling to less than $70 billion by 2075. Assuming this trend continued, the total present value debt could be less than $8 trillion.
WHY PHANTOM FUNDS DON'T COUNT
Some defenders of the current system assert that Social Security's finances are stronger than these figures indicate. Instead of running a deficit in 2010, they argue that the system will enjoy a surplus until 2021. Moreover, they claim that the Trust Fund has enough assets to pay full benefits until 2032.
These assertions, however, are made possible only by counting bookkeeping entries as real income. The Social Security system currently is collecting more money than is needed to pay benefits. These surplus revenues are spent on other government programs, and the Social Security Trust Fund is given an IOU from the Department of the Treasury. Specifically, the Trust Fund receives U.S. government bonds.
IMAGINE RUNNING A HOUSEHOLD USING TRUST FUND FINANCES...
To understand the reasons that the IOUs in the Social Security Trust Fund are meaningless, consider what would happen to a household that operated its finances in the same way.
Imagine that a husband and wife decided they needed to set aside $1,000 annually so that their newborn would be able to attend college. Instead of investing the money in real assets, however, the parents followed the government's example: The family spent the money and issued itself an IOU--exactly as the federal government does when it spends the Social Security surplus--that it proceeded to place in a safe deposit box. Moreover, like the government, the family kept a ledger that showed the IOU growing each year because of interest. By the time the child turned 18, the family would have IOUs in the safe deposit box totaling about $34,000 (assuming they promised to "pay" themselves 7 percent interest).
Now imagine that this family took the child to the college tuition office and attempted to pay with these IOUs. Needless to say, college officials would point out that the IOUs were meaningless because the $34,000 "asset" was offset exactly by the family's $34,000 "liability." With no actual money to pay the tuition fees, the college would refuse to register the child.
Every year, these bonds supposedly "earn" interest. In 1998, for example, the Trust Fund claims that it will receive more than $49 billion in interest from its $650 billion collection of bonds. The only problem is that none of this is real money. All that happens is that the amount of IOUs in the Trust Fund will increase by that amount (plus new IOUs issued as the annual cash surplus of Social Security is spent on other government programs). The interest payments simply represent one part of the government's pretending to make a payment to another part of the government.
Opponents of reforming Social Security dispute this analysis, arguing that the bonds in the Trust Fund are backed by the "full faith and credit" of the U.S. government. All this means, however, is that the bonds are a claim on future taxpayers. In short, all future Social Security benefit payments will be financed by revenues collected that year. The bulk of those revenues will continue to be raised through the payroll tax. Some of those benefits may be paid for from income tax revenues (in which case, the government will undertake the meaningless exercise of retiring IOUs held by the Trust Fund). And some of the benefits may be financed by government borrowing (in which case the IOUs in the Trust Fund will be replaced by IOUs held by the public).
Legislators could enact a law that doubled the size of the Trust Fund. They even could pass legislation arbitrarily that made the Trust Fund ten times larger than it is today. Nothing they could do, however, would change the fact that the Trust Fund is nothing but a pile of IOUs, and that the interest paid to these IOUs is meaningless.
In addition to the phantom interest payments, another source of make-believe revenue for the Social Security system are payroll taxes supposedly paid by employees of the federal government. Like other workers, federal employees are covered by Social Security. And like other workers, their employer is responsible for withholding and paying those payroll taxes. In the federal government's case, however, each agency of the government pretends to make a payment on behalf of its workers and the Social Security system pretends that it has received the money.
Chart 4 shows the amount of "income" the Social Security Trust Fund pretends it will receive from interest payments and the payroll taxes of federal government workers. Neither of these amounts, however, represents a real transfer of resources.
The Social Security system is hovering on the brink of a financial abyss. Bringing the system into balance would require imposing a 54-percent increase in payroll taxes, reducing benefits by 33 percent, or using a combination of both approaches. These drastic measures are the transition cost of maintaining the current system and paying out promised benefits. These policies would exacerbate the Social Security crisis by making the system an even worse deal for workers. Younger workers already face real rates of return that are barely above zero; in some cases they face negative returns. Forcing them to pay more and to get less hardly represents good public policy.
Privatization, on the other hand, would mean that taxpayers would realize transition benefits because the additional costs needed to finance the shift to a private system would be so much less than the additional costs needed to preserve the status quo.
Daniel J. Mitchell is McKenna Senior Fellow in Political Economy for The Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation.
3. The OASDI Trustees' report on the Trust Fund includes annual estimates for the next ten years and for every fifth year thereafter. Annual numbers through 2075, however, are obtainable through the Social Security Administration's Office of the Actuary.