February 4, 2000 | Backgrounder on Retirement Security
In a remarkably short period of time, there has been a significant shift in attitudes about Social Security. As recently as five years ago, very few policymakers voiced their support for fundamental reform of the old-age retirement program. Today, by contrast, there is a growing bipartisan consensus that transforming Social Security into a system based on personal retirement accounts is the best way to ensure retirement security for today's workers.
Dozens of countries around the world have successfully privatized their social security programs, helping to convince U.S. lawmakers that reform is politically practical, not just theoretically attractive.
Yet this change in attitude does not necessarily mean that America's Social Security system will be reformed. Defenders of the status quo are battling vigorously against reform. And the fact that public opinion is increasingly sympathetic to privatization is no guarantee that change will occur. America's political system is not conducive to momentous shifts in public policy. Indeed, by creating a system based on separated powers, the Founders sought deliberately to make it difficult to enact far-reaching changes.
Therefore, the opponents of change do not need to win the debate. They simply need to create enough uncertainty and erect enough hurdles so that advocates of reform are unable to muster the support they need to guide an important modification in how the system provides retirement income through the legislative process. Unfortunately, many of the arguments the critics use to hinder reform are irrelevant, misleading, and in some cases simply false.
Social Security will begin to run deficits in 2014, shortly after the baby-boom generation begins to retire. Even after adjusting for inflation, the projected deficits are staggering. Annual deficits (in 1999 dollars) will reach $100 billion in 2020, $200 billion in 2026, and $300 billion in 2037,1 making the total accumulated shortfall over the next 75 years a monumental $20 trillion.
If the numbers are not adjusted for inflation, the estimates are even more striking: $200 billion by 2021 and $1.5 trillion by 2048, with a cumulative shortfall between now and 2075 well above $100 trillion.
The Social Security trust fund is not a savings account. It consists solely of IOUs that, when redeemed as the baby-boom generation retires, will impose a significantly larger tax burden on tomorrow's workers.
The Social Security trust fund does not hold real assets. Surplus Social Security revenues are either spent on other government programs or used to pay down the national debt. All that the trust fund gets in return are IOUs from the U.S. Treasury. At best, these bonds simply give Social Security a claim on future income tax revenues. In the words of the Clinton Administration,
These balances are available to finance future benefit payments and other trust fund expenditures--but only in a bookkeeping sense....They do not consist of real economic assets that can be drawn down in the future to fund benefits. Instead, they are claims on the Treasury, that, when redeemed, will have to be financed by raising taxes, borrowing from the public, or reducing benefits or other expenditures.2
Proposals that supposedly dedicate the budget surplus to Social Security will have no impact on the program. The President's plan to pay down the debt and add IOUs to the Social Security trust fund will not change the program in any meaningful way. The Republican plan to create a "lock box" for the surplus would be equally ineffectual.
The only thing these proposals do is pay down the national debt. They do not reduce Social Security's unfunded liability, and they do not improve Social Security's low rate of return for workers. As explained above, the Social Security trust fund is not a pool of real assets. If adding more bonds to the trust fund could solve the problem, lawmakers could simply pass legislation adding three zeroes to every bond already in the trust fund. They will not do this, however, because they understand that an IOU written to oneself has no value, creates no wealth, and therefore has no meaning.
Elimination of Social Security's long-term deficit would require a 50 percent increase in taxes, a 33 percent reduction in promised benefits, or a combination of both. Changes such as these would impose immense hardship on workers and retirees.
Those who claim that Social Security can
be fixed by tinkering have been unable to put forward plans that
will fix the program. The reason, simply stated, is that the
long-term deficit is too big. By the time most baby boomers retire,
for instance, tax collections will cover only 70 percent of
promised benefits. To close that gap, according to Social Security
Administration (SSA) estimates, payroll tax rates would have to
rise from today's 12.4 percent to more than 17 percent by 2030 and
over 19 percent by 2075.3 Alternatively, SSA
figures show that promised benefit payments would have to be
reduced by one-third to erase the long-term
Most Americans presumably would not view as "modest" these types of tax increases and/or benefit cuts. The same can be said for other potential "fixes," such as boosting the retirement age and cutting the cost-of-living adjustment (COLA). All of these proposals, at least in theory, could bring Social Security into balance, but they would significantly alter the lives of current and future retirees.
Reform will save Americans trillions of dollars, since the transition cost of shifting to personal accounts would be far less than the $19.8 trillion transition cost of bailing out the current system.
Reforming Social Security would be like refinancing a mortgage; the up-front cost--paying promised benefits to current and future retirees--would be akin to paying points when taking out a new mortgage. That short-term cost is worth incurring, however, because Social Security's future unfunded liability will decrease (just as homeowners pay points to enjoy the benefit of smaller mortgage payments in the future). In other words, reform is a way to save money over time.
More specifically, a system of individual accounts would result in transition costs because workers would be shifting some portion of their payroll taxes out of Social Security and into their personal retirement accounts. And because some of that money is not being used to fund Social Security payments, the government will need to find some other source of revenue to finance payments to current and future beneficiaries.4 After a few decades, however, this transition cost diminishes. Indeed, the government actually begins to save money because future retirees will be able to self-finance the bulk of their retirement using the money they have saved in their personal retirement accounts.
The amount of money needed to finance this transition cost depends on the amount of payroll taxes that workers would be allowed to invest in their accounts and how quickly the new system would be implemented. The short-term transition cost can be minimized if personal accounts are phased in slowly. But a slow phase-in period also would mean that the huge unfunded liability of today's system would be reduced both more slowly and less completely. In other words, there is a trade-off: If workers are allowed to divert the bulk of their payroll taxes, the short-term cost will be higher, but the long-term savings will be larger.
A more sweeping reform program, similar to the reforms implemented in nations like Australia and Chile, would allow workers to put most of their payroll taxes into private accounts while guaranteeing all currently promised Social Security benefits. This would mean a large transition cost in the short run, but it would also reduce the unfunded liability more rapidly and allow workers to look forward to the prospect of much higher income in retirement.5
More modest reform plans would result in lower transition costs in the short run but fewer savings in the long run. Moreover, more modest reforms also mean that today's workers will enjoy less extra retirement income than they would enjoy with more complete reform. For example, legislation introduced in the House by Representatives Jim Kolbe (R-AZ) and Charles Stenholm (D-TX) would involve less transition financing in the early years--though the trade-off is that the costs would not fall by as much in future years, and workers would not receive as much retirement income in the future.6
Regardless of how quickly Social Security is reformed, the transition cost of reform can be spread over many years--even generations. Indeed, this may be the fairest way of dealing with the problem, since the benefits of reform will also be enjoyed over several generations.
Unless it was linked to other reforms--such as letting workers shift some of their payroll taxes to personal accounts--raising the retirement age to 70 or 75 would make people pay more for fewer benefits and would create particular hardships for those with physically demanding jobs.
Increasing the retirement age would reduce Social Security's deficit in two ways. First, because it would force people to spend more years in the workforce, more payroll taxes would be collected. Second, because these workers would have fewer years of retirement, the level of Social Security spending would be lower. In theory, the program's finances could be kept in perpetual balance by regular adjustments of the retirement age.
The strongest argument for adjusting the retirement age is that life expectancy has increased dramatically. A worker born in 1940, for instance, had a projected life span of 63.55 years. A worker born today, by contrast, is expected to live more than 76 years.
Unfortunately, although further changes in the retirement age (which already is slated to reach age 67 by 2027) would save a lot of money, such a policy would make Social Security an even worse deal for workers. The rate of return the average worker can expect from Social Security is already very low, so workers who are forced to spend more years in the workforce and fewer years in retirement would see their already dismal rate of return become negative.
Therefore, changes in the retirement age should be considered only if they are accompanied by more fundamental reform of the program. Personal accounts, for instance, would help ensure that manual laborers, minorities, and others who have lower life expectancies would have the ability--depending on their level of private savings--to choose their own retirement age. This would relieve them of having to bear a disproportionate hardship if the eligibility age for government benefits was increased.
Supporters of the status quo sometimes argue that Social Security's problems would go away if the economy just grew a little bit faster. According to this argument, more jobs and higher incomes for more people would increase the amount of payroll taxes coming into the system. All of this is accurate, and it may even be true that the Social Security Administration's long-run growth estimates are too pessimistic.
However, even substantial increases in projected growth will have only a modest effect on Social Security's finances--in large part because higher wages entitle workers to larger retirement benefits. Consider what happens if inflation-adjusted wages grow 56 percent faster than currently forecast by the Social Security Administration (a 1.4 percent annual increase instead of a 0.9 percent annual increase):7
Initially, new immigrants would contribute to Social Security while imposing virtually no costs on the system. Eventually, however, this short-run infusion of revenues would be offset when these new workers retired and spending increased. In theory, policymakers could find a new and larger group of immigrants each year to offset the spending increases, but the house of cards would come tumbling down when there were not enough immigrants and it was time to pay benefits to all the new retirees.
Letting politicians invest Social Security funds is an open invitation to financial mismanagement and politically driven investment decisions.8
A handful of countries, including India, Kenya, Malaysia, and Singapore, have implemented government-controlled investment of retirement money. At best, workers in these nations get very low returns because the money is invested for political rather than economic reasons.9 In most cases, as shown in Chart 2, corruption and mismanagement result in negative returns; the workers would have been better off hiding the money under their mattresses. Private management of funds is much safer since there is both the competitive pressure to get a good return and the legal obligation to make investments in the best interest of the worker.
Another shortcoming of government-controlled investing is that it does nothing to improve retirees' income. Consider the plan introduced by President Clinton that would allow politicians to invest about one-fourth of the Social Security surplus. Even if it worked perfectly, workers would see no benefit. Retirement income would not even increase by one penny; all the returns would go to the government. To be sure, this added money could be used to stave off Social Security's bankruptcy, but the U.S. General Accounting Office estimates that Clinton's plan--if successful--would add only six years to the program's solvency.10
Legitimate changes in the CPI, based on scientifically sound updates of survey methodology, are reasonable. The CPI should be changed if it is too high (or low), but changing it artificially just to reduce Social Security's huge deficit would debase the integrity of government statistics and be a back-door way to force workers to pay more while reducing benefits to retirees.
The CPI, which is used as the basis for adjusting Social Security benefits to reflect changes in the cost of living, is generally thought to be too high because it overstates inflation, which leads to a steady rise in benefits above the intent of the program.11 Changing it to reflect reality is prudent and fair. Forcing the bureaucracy to tamper with CPI calculations so that politicians can claim credit for saving money, however, could open the door for politically motivated changes in other government statistics.
Moreover, an unjustified reduction in the CPI--beyond an appropriate change so that benefit increases accurately reflect the cost of living--will do three things, none of which would be good for people.
Means-testing the benefits not only would constitute selective punishment of those who saved and invested during their working years, but also would require penalizing seniors who have incomes as low as $40,000 in order to have any noticeable impact on Social Security's shaky finances.
Social Security is based on the notion that everyone pays into the program and everyone gets something out of the program. Means-testing violates this principle of equal treatment by telling some citizens that they must endure the costs but then receive none of the benefits. The benefits may be meager compared with what they could obtain by investing in personal retirement accounts, but that does not change the underlying principle of equal treatment.
Means-testing, which would mean the denial of some or all retirement benefits as income rises above a certain level, also would have adverse economic consequences. Taking away benefits as income rises creates a bigger gap between a senior's total income and his disposable income. This would reduce the incentive to engage in productive behavior. This is particularly pernicious because, for most seniors, income in retirement usually is the result of savings and investment that occurred during working years. In other words, means-testing really creates a significant disincentive to save and invest--and every economic theory, even Marxism, agrees that capital formation is the key to rising wages.
Finally, means-testing is unlikely to solve Social Security's financial problems because, simply stated, there are not enough rich seniors. The only way to make a significant dent in the program's long-run deficit is to impose means-testing on the middle class. Yet even if means-testing was forced on seniors making as little as $40,000 annually, it would reduce the long-run deficit by less than 50 percent.12
The returns available from private investments are dramatically larger than the returns available from Social Security--even after subtracting the tiny fraction of account balances that would be used to pay administrative costs.
Personal retirement accounts would be subject to fees for funds management and information processing, but such charges in a well-designed system would be less than one-half of 1 percent (0.5 percent) of assets annually. Workers with small accounts would be likely to pay less than $10 per year. And since private investments produce a much larger return than Social Security, the net effect is that workers would have significantly more income when they retire.13
It also is worth noting that estimates of Social Security's administrative costs (supposedly less than 1 percent of annual benefits) are misleading, largely because they do not include the compliance costs of payroll tax collection that are imposed on workers and businesses. To be fair, the payroll tax, which is basically a flat tax with no deductions, is not nearly as onerous as the income tax. Nevertheless, the compliance costs are substantial when compared with the budget of the Social Security Administration, making the real administrative costs of Social Security several times larger than the official number.
Chart 3 shows how much retirement income a worker will have if allowed to place 3 percent of income in a private account, as well as the impact of administrative costs using two different assumptions. In either case, the worker will wind up with about twice as much money as he would receive if the same amount of money was paid to Social Security.
"Since Social Security redistributes money from some types of families to others, privatization would mean less retirement income for certain groups, such as low-income, single-earner couples."
All demographic groups would enjoy more retirement income if they were allowed to have personal accounts. Moreover, some groups that are particularly disadvantaged by the current system, such as black Americans, would reap large benefits.
Because Social Security's benefit formula is tilted against higher-income workers, it is sometimes thought that the program is a good deal for the poor. Since there is a link between income and life expectancy, however, the poor do not get a measurably better rate of return from Social Security. Simply stated, their average life spans after reaching retirement age are too short. For some groups with particularly low life spans, such as African-Americans, Social Security is a terrible deal.
Social Security also is designed to redistribute income to single-earner couples at the expense of single workers and dual-earner couples. It actually does achieve this result, but this does not mean that single-earner couples would not benefit from personal accounts. As the Chart 4 illustrates, all demographic groups would enjoy more retirement income if given the opportunity to steer a portion of their payroll taxes to private accounts.14
The disability program (which provides payments to workers who become disabled) and the survivors program (which provides payments to children of workers who die) are separate parts of the Social Security program. Allowing workers to shift some of their payroll taxes to a personal retirement account would have no impact on these other programs.
Social Security reform sweeps across the globe, nations at all
stages of development are shifting to systems based on personal
retirement accounts. Among the nations that are similar to the
United States, Australia has implemented a fully privatized system
that was enacted by a Labor
government, and Britain has taken a partially privatized approach. In Western Europe, Denmark, Sweden, and Switzerland have moved, to varying degrees, to compulsory retirement savings.15
Chile set up a very successful system nearly 20 years ago, and seven other countries in Latin America--Argentina, Bolivia, Colombia, El Salvador, Mexico, Peru, and Uruguay--have adopted similar systems of mandatory retirement savings.16 Singapore has a private system (although government-controlled investment has resulted in paltry returns), and Hong Kong is implementing one.17 In the former Soviet empire, Croatia, Estonia, Hungary, Kazakhstan, Latvia, and Poland either have privatized or are privatizing their pension systems.18 In all of these cases, policymakers realized that reform was a good deal for workers, taxpayers, and retirees.
Some types of investments are volatile in the short term. The record stock market decline for one day is 20 percent, for instance, and the record drop in one month is 30 percent.19 Over time, however, boom markets offset these occasional downturns, and the longer-term performance has been very positive. Indeed, over the past 70 years--a period that includes both the Great Depression of the 1930s and a one-day decline of 20 percent in 1987--annual returns in the stock market have averaged more than 10 percent (more than 7 percent after adjusting for inflation). This is far higher than the 1.3 percent average rate of return that a two-earner married couple with two children can expect from Social Security.20
Personal retirement accounts are long-term investments. As such, they allow workers to ignore periodic fluctuations and reap the benefits of compound interest over long periods of time. This does not mean, incidentally, that some workers will not achieve better returns than others from private investments. Looking at the best and worst 46-year periods (the time an average person will spend working) in market history, a worker who retired in 1987 would have enjoyed a return of nearly 13 percent, while a worker who retired in 1974 would have realized a return of 7.32 percent.21 In both cases, though the worker would have had much more retirement income than Social Security provided.
For workers who are extremely risk-averse, there are investment options that have virtually no risk but still outperform Social Security. Series I bonds issued by the U.S. Treasury, for instance, currently pay a guaranteed inflation-adjusted 30-year return of 3.4 percent.22 This is lower than the returns that will be available from stocks and corporate bonds but significantly better than the returns from Social Security.
Personal retirement accounts certainly would be good news for those with higher incomes, but it is not as if they will suffer if the program stays the way it is now. As Chart 5 shows, Social Security provides only a fraction of their total old-age income. Workers with more modest incomes, by contrast, have the most to gain. Not only would they be likely to enjoy the largest percentage increase in retirement income, but the extra $500 to $1,000 per month they could expect from personal accounts would make a big difference in their quality of life.
This demeaning claim is irrelevant, since personal retirement accounts presumably would be professionally managed. Nonetheless, if people can be trusted to choose their careers, vote for a President, buy homes, and raise families, they certainly can be trusted to make basic choices about investments.
The argument that workers are incapable of participating in a private Social Security system is morally, analytically, and empirically flawed. First, people make very important choices every day of their lives. They get married, change jobs, and select insurance policies. Many of these choices are just as important as--if not more important than--picking a pension fund. Critics respond by saying that many poor people are financially illiterate, but this claim assumes that they are not capable of learning once there is a reason to do so.
In any event, this assertion is a red herring. Almost all of the proposals to create personal retirement accounts require professional management of the funds. At most, workers could choose from a list of approved pension fund providers. Indeed, it is far more likely that the government will over-regulate the new system than that workers will be thrown into a system they cannot comprehend.
Finally, one need only look at the experience of other countries to see that people are perfectly capable of making choices and planning for their retirement. Americans may not be financial experts, but they are presumably as knowledgeable about finances as are the British, Chileans, Mexicans, and Hungarians. In all of these countries, as well as about two dozen others, workers are responsible for choosing an appropriate private pension fund. And while it would be an exaggeration to claim that any of the systems set up in other countries is perfect, the shift to personal accounts has proved successful, and there is no campaign to reverse those reforms.
Defenders of the status quo are trying to have it both ways. They condemn privatization with demagogic charges while avoiding any discussion of what they would propose in lieu of reform. The goal of this strategy is clear: If the choice is between reform and doing nothing, many voters might feel more comfortable with the current system, which at least has the advantage of being a known commodity.
Yet, as the following points illustrate, the choice for policymakers is not between reform and current law, but rather between reform and draconian changes. Even though the current Social Security surplus means that these harsh changes might not be necessary for another decade, Social Security's gigantic deficit means that changes will happen. The only question is whether the changes will give Americans a better way of saving for retirement or come from the following menu:
As bleak as this picture is, it will only get worse if Social Security is not reformed. If promised benefits are to be paid, the payroll tax rate will need to rise above 19 percent--an increase in the burden of about 50 percent. The amount of income subject to the tax, which already is rising rapidly under current law, also could be increased. Such a step, however, would destroy Social Security as an insurance program.
Less Retirement Income. Policymakers could address the crisis by cutting benefits. The options for doing this include reductions in benefits, changes in cost-of-living adjustments, and increases in the retirement age. Regardless of the method, however, the impact would be dramatic. Benefits would have to be slashed by about one-third in order to balance the system 30 years from now.
Budget Deficit. Social Security's long-run deficit will reach about 2 percent of the nation's gross domestic product (nearly $200 billion today). Closing a gap this large, whether by tax hikes, changes in spending, or a combination of these two, will require changes of a magnitude not seen since World War II. Needless to say, Americans were willing to make sacrifices in order to defeat Nazi Germany and Imperial Japan. It is not very likely, however, that voters will accept large tax increases or benefit reductions just to prop up a program that most would opt out of if given a choice. This would mean a return to large budget deficits.
Economic Stagnation. Previous Social Security crises have been addressed largely by means of higher taxes. Yet the payroll tax is a levy on jobs, and as the European experience has demonstrated, high payroll taxes contribute to unemployment. Many other policy mistakes are contributing to Europe's decline, it is true, but it would be unwise for the United States to mimic even one policy that causes joblessness. Another proposal--to turn Social Security into an income redistribution plan by applying the payroll tax to all wage and salary income--would involve the largest tax increase in U.S. history and push marginal tax rates to levels not seen since the 1970s.23
Intergenerational Conflict. Younger workers already are dissatisfied with Social Security and have extremely low expectations of receiving anything from the program. Just imagine, however, what would happen if they faced a 50 percent increase in the payroll tax. Or if they saw that their promised benefits were to be cut by one-third.
Social Security reform will not be a free lunch. It will involve decisions about how much workers should save, whether the accounts should be employment-based, how the accounts should be taxed, the level of regulation, the structure of the safety net, and how to make the transition from the current system.
As many other countries have demonstrated, it is possible to make these decisions and create a system that will be good for workers, taxpayers, and retirees. For these decisions to be made, however, the debate should be conducted in a rational and honest manner, with full understanding of the consequences of both action and inaction.
It is regrettable that many opponents of reform, fearing that such a debate will lead to a system of personal accounts, have decided that "victory" requires the demonization of privatization and adoption of a head-in-the-sand approach to serious consideration of alternative ways to bail out the current system.
Daniel J. Mitchell, Ph.D., is McKenna Senior Fellow in Political Economy at The Heritage Foundation.
1. Social Security Administration, 1999 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds, at http://www.ssa.gov/OACT/TR/TR99/index.html .
4. Generally, it is assumed that the transition cost will be financed by using non-Social Security revenues. The transition cost can also be addressed, however, by making changes in Social Security revenues and benefits.
9. World Bank, Averting the Old-Age Crisis: Policies to Protect the Old and Promote Growth (New York: Oxford University Press, 1994).
11. Final Report of the Advisory Commission to Study the Consumer Price Index, Toward a More Accurate Measure of the Cost of Living, Committee on Finance, U.S. Senate, 104th Cong., 2nd Sess., December 5, 1996. The commission is also known as the Boskin Commission after its chairman, Michael J. Boskin.
12. Report of the 1994-1996 Advisory Council on Social Security, Washington, D.C., 1997, at http://www.ssa.gov/policy/adcouncil/toc.htm .
13. David C. John and Gareth G. Davis, "Keeping Administrative Costs Low," in David C. John, ed., Improving Retirement Security: A Handbook for Reformers (Washington, D.C.: The Heritage Foundation, 2000).
15. Estelle James, "Social Security Reform in Other Nations," Heritage Lecture No. 618, June 4, 1998, p. 4.
17. World Bank, "Current
News and Events," at
18. Cato Institute, "Estonia and Latvia May Be Next to Privatize Social Security," August 6, 1999, at http://www.socialsecurity.org/dailys/08-06-99.html .
22. U.S. Department of the Treasury, Bureau of the Public Debt, "US Savings Bond Home Page," at http://www.publicdebt.treas.gov/sav/sav.html .