Social Security Then and Now
By all accounts, Social Security is the most successful domestic government program in American history. This year, more than $500 billion will be relatively costlessly taken from the pockets of American workers and transferred to those living in retirement. As a consequence of this program, the poverty rate among the elderly has fallen sharply over the last 65 years, and young people have largely been relieved of the burdensome responsibility of caring for their parents and grandparents in old age.
Yet, as almost everyone agrees, this model program is itself showing signs of old age. By 2018, revenues from Social Security taxes will no longer be sufficient to pay Social Security benefits. As President Bush pointed out in his State of the Union address, the number of workers supporting each retiree has fallen from 16 in 1950 to just 3.3 today, and those entering the labor force will get an infinitesimally smaller return on their contributions than did earlier generations.
The President has decided to make reform of this system a centerpiece of his second-term agenda. While there is as yet no detailed proposal for doing so, he has made it clear that personal accounts—along the lines of the popular 401(k) accounts commonly available to workers today—will be key to the plan. Basically, younger workers will be expected to give up some portion of their future Social Security benefits in exchange for such accounts, the return from which will compensate them for the lost benefits.
Understanding precisely how Social Security was developed, why certain features have been central to its operation, and how people may react to changes in the system will be critical to the success or failure of the Bush initiative. Following is a brief survey of some of the issues.
Since time immemorial, societies have had systems to care for the aged who were no longer able to work to support themselves. The foundation of these arrangements has been the family: until recently, the main method of providing for one's old age was to have as many children as possible. But even in ancient times, annuities were available to allow people to save specifically for retirement.
With an annuity, one pays a lump sum in return for specified payments for the balance of one's life. Records of annuities exist from ancient India, China, Babylon, Egypt, and Rome. In the Middle Ages, governments often sold annuities as a way of financing wars. Today, annuities are widely available through employers or can be purchased by individuals with funds that may have been saved through an Individual Retirement Account or a 401(k) plan.
Eventually, fraternal societies and guilds took on some of the responsibilities of caring for their aged members. According to the historian David Beito, on the eve of the Great Depression as many as 35 million Americans belonged to fraternal societies, many of which catered to immigrants and minorities and often provided full health and life insurance for their members. Some even owned hospitals.
Businesses also played an important role in providing for the elderly. For example, older workers were sometimes graduated into token jobs or given a stipend. As early as 1882, the Alfred Dolge company, a builder of musical instruments, had instituted a formal pension plan, withholding 1 percent of its workers' pay and adding 6-percent interest each year.
The Great Depression effectively ended the role of fraternal societies in the provision of social welfare. It also put an enormous strain on businesses, many of which terminated their pension plans. It was therefore inevitable that the federal government, which had already begun to provide pensions to veterans of wars and their families, would be forced to undertake this task. What was not at all inevitable was that the Social Security system would develop the way it did. Its precise form is mainly due to the vision of Franklin D. Roosevelt.
Roosevelt's chief political concern in his first term was the challenge from his Left. The Socialist candidate Norman Thomas had received almost a million votes in 1932, and populist demagogues like Huey Long and Francis Townsend were achieving wide popularity with their calls for massive income redistribution—far more than a capitalist economy could tolerate. With the rise of Communism in Russia (and National Socialism in Germany), Roosevelt was rightly concerned that failure to deal with social pressures here could lead to revolution.
Roosevelt's strategy was to institute modest liberal reforms accompanied by a few radical flourishes and rhetoric for effect. In this respect, he anticipated Bill Clinton's triangulation strategy, satisfying his liberal constituency mostly with words while the substance of his policies was much more conservative than is generally appreciated even today.
Concerning Social Security, Roosevelt faced strong pressure to institute a program that would be paid immediately out of general revenues. This would have been much more progressive and perhaps better for the economy at the time. Roosevelt, however, was adamant that he was not going to enact a welfare program but a funded system in which workers would have to earn their pension. Indeed, he chose just about the most conservative financing scheme possible, one that was strenuously opposed by the Left as being deeply regressive.
At the same time, it is clear that Roosevelt was at least as concerned with tying workers to the state as with helping them achieve a secure old age. This is proved by his strenuous efforts to kill a Senate plan that would have privatized the system right from the beginning. An amendment sponsored by Democratic Senator Bennett Champ Clark would have allowed workers to opt out of Social Security if their employers provided an annuity with benefits at least equal to those of the federal program. Although the amendment passed the Democratically-controlled Senate by a healthy margin, Roosevelt pressured the House to reject it in conference. In the end, the Clark amendment was dropped, killing the last serious effort to create a private alternative to the government-run system.
The original tax rate imposed by the Social Security legislation in 1935 was 1 percent of the first $3,000 of wages on both employer and employee, for a total of 2 percent. (Economists generally believe that in fact the employee pays the whole tax, in the form of lower wages.) Everyone, regardless of income, would pay at the same rate—in effect, a flat tax on a broad base with no exemptions whatsoever.
Taxes were first collected in 1937, but benefits would not be paid until 1940. Although he was criticized by Keynesian economists, who saw this initial build-up of Social Security surpluses as depressing consumer spending, Roosevelt again held firm: he saw the creation of a trust fund as essential to making Social Security look as much as possible like a private pension.
There was a political calculation here as well: tying contributions and benefits together would guarantee the long-term viability of the system and protect it forever against political attack. As he told a doubting adviser, “We put those payroll contributions there so as to give the contributors a legal, moral, and political right to collect their pensions and their unemployment benefits. With those taxes in there, no damn politician can ever scrap my Social Security program.”
Roosevelt was correct on the politics and better on the economics than many believed. The tight linkage between taxes paid and benefits received has ensured that workers tend to regard Social Security in the same light in which they regard private pensions. That is, they look upon payroll taxes as forced saving, akin to deductions for health insurance or 401(k) contributions, and not as a true tax for which nothing specific is received.
Critics still charge the Social Security tax with regressivity, but in doing so they overlook the nature of the benefit structure. True, low-income workers pay a greater percentage of their income in payroll taxes than do those with high incomes, but they also get much larger benefits relative to their contributions. On balance, the system—benefits and taxes taken together—is highly progressive.
This progressivity was especially pronounced in the early years of the program, when benefits were paid out to people who had paid virtually nothing in. The classic example is Ida May Fuller, the first person to draw retirement benefits. She paid a total of $24.75 into the system between 1937 and 1940, when she retired, drawing benefits from that point until her death in 1975 at the age of a hundred. Over the years, she received $22,888.92 from Social Security, an impressive return.
While few retirees did so well as Fuller, it is nevertheless the case that most have historically registered a very good return on their Social Security taxes. As recently as 1980, according to the Congressional Research Service (CRS), the average worker got back all of his and his employer's Social Security taxes in the form of benefits in just 2.8 years of retirement, and when he began to draw Social Security at age sixty-five, he received monthly benefits equal to almost half his pre-retirement monthly wage.
As the years have gone by, however, this rate of return has fallen sharply. The reasons are simple. More and more retirees have paid into the system during their entire working lives, and are thus not in a position to reap the benefits of Ida May Fuller and other early retirees. Also, both the tax rates and the wage base have risen steadily over the years. Today, the tax rate for Social Security's retirement program is 5.3 percent on workers and employers on all wages up to $90,000. Additional taxes for disability insurance and Medicare put the total payroll tax rate at 15.3 percent.
As a result, in 2000 it took a low-income worker 11.8 years to get back all of his and his employer's Social Security taxes in the form of benefits, while an average worker needed 17.5 years and a high-income worker 24.9 years. Absent any changes to the system, this return will continue to deteriorate. By 2030, the CRS estimates that it will take a low-income worker 14.1 years to recover his contributions, an average worker 25.6 years, and a high-income worker 55 years.
Meanwhile, Social Security's long-term finances are becoming increasingly precarious. This would be self-evident were it not for the existence of the trust fund, which, on paper, has sufficient assets to continue paying full benefits well after 2018, the point at which payroll taxes will be insufficient to pay current benefits.
The problem here is that the trust fund is and always has been invested only in U.S. Treasury securities. In effect, the government simply lends money to itself, using payroll taxes in excess of current benefits to fund general government programs. This does not amount to stealing, as is sometimes charged, but it does nothing to reduce the economic burden of Social Security.
The trust fund is best thought of as budget authority: legal permission to spend general revenues to pay Social Security benefits, which will become a growing necessity after 2018. If the Treasury simply created $10 trillion of new bonds out of thin air and deposited them in the trust fund, the effect would be only to extend the federal government's legal ability to use general revenues to pay benefits.
This is not to say that the apocalypse will occur on the day the trust fund runs dry—a day now estimated to fall in the year 2042. Even if this were a looming threat, it would take Congress only a matter of hours to pass legislation providing for general-revenue financing of any shortfall in payroll taxes and trust-fund assets to cover promised benefits. The true guarantee of Social Security benefits is not the existence of a trust fund or words on a piece of paper but the overwhelming necessity of providing for the elderly—plus the continued ability of the government to take sufficient revenues out of the economy to finance those benefits.
The real question, therefore, is whether we can afford the benefits that have been promised. Under the Congressional Budget Office's long-term projections, Social Security spending will rise by 2 percent of the gross domestic product annually between 2010 and 2050. While large, this is not a tremendous economic burden to meet—roughly equivalent to increasing the payroll tax rate by 1 percent each on employer and employee, for a new combined rate of 17.3 percent.1
Unfortunately, though, the idea that a 2-percent tax increase is all we need to fix Social Security in perpetuity ignores deteriorating demographic trends that will make the burden much greater in the future than it is today. Because of declining birth rates and the aging of the giant baby-boom generation, the number of workers per retiree will continue to decline. One way to quantify this effect is by looking at those receiving Social Security as a percentage of those paying the taxes. This “dependency ratio” is projected to rise from 30 percent today to 54 percent in 2080.
Because of growing life expectancy, moreover, these estimates are likely to prove optimistic. In 1960, a sixty-five-year-old male could anticipate living another 13.2 years and a female 17.4 years. By 2003, a man of sixty-five could look forward to an additional 16.7 years and a woman to 19.5 years. Improving medical technology may sharply increase these figures even well beyond those now projected for 2080, when it is currently estimated that a sixty-five-year-old man will live another 21.1 years and a woman another 23.7 years.
Another problem: projections tend to ignore the fact that few workers today bother to wait until the normal retirement age before drawing their benefits. In 2002, 56.1 percent of those eligible began taking benefits at age sixty-two, and another 22.7 percent started before sixty-five. The critical issue here is not the outflow of payments, since those retiring early begin with an actuarially lower benefit—currently about 20 percent less than they would get at the normal retirement age. The issue, rather, is the economic impact of a regulation that severely limits how much wage income one may earn after retirement without losing Social Security benefits. Those below the normal retirement age lose $1 of benefits for every $2 earned above $12,000—the equivalent of a 50-percent marginal tax rate that pushes substantial numbers of otherwise productive people out of the labor force.
This brings us to the debate over personal accounts. One strong argument for such accounts is that they will give older workers an incentive to stay in the labor force rather than ceasing work at the earliest opportunity. Many now take early retirement because of a use-it-or-lose-it mentality and because benefits generally cannot be willed to heirs. With personal accounts, however, workers in effect capitalize some of their Social Security benefits and will have unlimited freedom to leave this asset to whomever they please.
Indeed, one of the principal, unheralded advantages of replacing some of Social Security with personal accounts will be to expand the supply of workers in their mid-sixties. At a time of an acute national shortage of experienced and well-educated workers, this will be a plus both for society and for many workers themselves. In the 1930's, when life spans were shorter and more people held physically demanding jobs, full cessation of work at age sixty-five was justifiable (and may have seemed a prudent way to help open up vacancies for younger workers). Today, with increasing life spans, improvements in nutrition and medicine, and more comfortable working conditions, the idea that people should totally cease work in their mid-sixties is increasingly out of date. People of this age today are likely to be in better physical shape than their parents were in their forties, and many have found that retirement at age sixty-five, with the accompanying loss of a sense of purpose in their lives, is both emotionally and physically harmful.
But are workers likely to welcome the inevitable risks that go with personal accounts? The critics are doubtful, but their doubts are belied by what we know about 401(k) accounts, which were first authorized in 1978. These accounts did not replace defined benefit plans, which like Social Security pay a fixed sum to retirees for life. But both workers and employers have taken with such alacrity to the 401(k) that it has gradually displaced traditional pension plans.
Our experience with 401(k) and other such defined-contribution plans suggests that workers enjoy knowing the capital value of their retirement savings and having the freedom to invest those funds as they see fit. They also like the flexibility to set both the date of their retirement and the form of their retirement income, rather than being tied to an arbitrary arrangement.
Can workers expect a high enough return on personal accounts to compensate fully for the fixed benefits they will be ceding? The Social Security Reform Commission concluded that a real (inflation-adjusted) return of 2 percent per year would be sufficient to break even. Historically, a broad portfolio of common stocks has returned (according to the economist Jeremy Siegel) 7 percent per year on average. While one can always find some isolated period in which this was not the case, most economists believe that the risks essentially disappear over a reasonable span of time—say, 20 years.
A more important question about personal accounts has to do with the macroeconomic impact of the loss of a steady source of payroll-tax revenue to the government. If the Treasury ends up covering that loss by borrowing, then the government's negative saving will exactly offset the new private saving, leaving nothing for additional investment. Therefore, it is essential that personal accounts not be financed solely through higher budget deficits.
Cutting government spending could accomplish this end, but with the budget already running a large deficit, it will be very hard to find another $1 to $2 trillion in cuts over the next couple of decades to cover the loss to the Social Security trust fund. Conceivably, taxes could be raised, but one of the main reasons for reforming Social Security in the first place is to avoid a large tax increase, which would anyway tend to depress private saving and investment and thus reduce economic growth still further.
Consequently, it may be necessary to consider other reforms. In 1983, a Democratic Congress initiated a rise in the normal retirement age from sixty-five to sixty-seven, even though neither the Greenspan commission nor the Reagan administration had recommended it. One suggestion now would be to phase in a further increase beyond sixty-seven to keep pace with rising life expectancy. For a number of reasons, this may be the easiest long-term solution that could be enacted: raising the retirement age clearly would have no effect on current retirees or on those nearing retirement, who would be exempted from any such provision, and so could avoid arousing the most vocal group of potential opponents.
Another proposal would be to change the Social Security payment formula so that it no longer raises real benefits annually. This happens because benefits are indexed to wages, which normally rise faster than inflation; indexing them to prices instead would almost entirely finance the transition to personal accounts, while still giving future retirees the same benefits that today's retirees get in real terms. They just won't get more.2
Finally, there is the partisan political dimension of the issue. Some Republicans in Congress are deeply fearful of tinkering with Social Security lest highly organized elderly voters penalize them at the polls. That is unlikely to happen, however. As we saw when the retirement age was raised, the elderly are mainly concerned with their benefits today, and historically have not strongly resisted cuts in benefits that only affect retirees in the distant future.
Although AARP, the principal lobbying group for the elderly, has declared war on any effort to replace any part of Social Security with personal accounts, the organization may be more of a paper tiger than it used to be. Its year-in, year-out attacks on Republicans for trying to destroy Social Security have taken on a boy-who-cried-wolf quality that many of its own members no longer respond to. Moreover, with many AARP members living comfortably off 401(k) plans, Individual Retirement Accounts, and Keogh plans, the prospect of adding personal accounts to Social Security is not likely to frighten people the way it may have done 25 years ago. Then, too, the recent spate of reports about corporate pension plans going belly-up has had the effect of lessening the sense of riskiness attending other proposed schemes, especially ones that promise greater personal control over one's own money.
As for Democrats and liberals, some of them have never really opposed privatizing some of Social Security, at least in principle—and at least when they themselves have been in office. Bill Clinton, for one, conceded the desirability of privatization, although he differed with Republicans on how to go about it. A compromise plan could easily have been financed with the budget surpluses that emerged in 1998, and if the Monica Lewinsky scandal had not intervened, Clinton might well have ended up with the historical legacy he so badly craved.
In any event, the strenuous opposition now being mounted by Democrats has more to do with tactics than with conviction. The Democrats intend to bargain from the strongest position possible, which at the moment translates into straightforward resistance and obstructionism. There will be plenty of time later on to establish, if necessary, a floor for negotiations.
Where conviction enters the calculus is over the question of the proper role of government in people's lives. Republicans want to wean people away from the state; turning workers into investors, to the greatest extent possible, is one way of accomplishing that goal. Democrats have a different governing philosophy, and a deep political worry. The more that American voters become sympathetic to the Republican message of personal responsibility and limited government, the greater will be the peril to the long-term prospects of the Democratic party in its current incarnation.
With the political stakes so high, it is little wonder that Bush's opponents have declined to focus on the simple question of whether personal accounts are an appropriate addition to our current system—especially because they are.
1 This is still well below the rate in many other countries. Italy, for example, has a rate just for retirement and disability programs of 32.7 percent, and a total payroll tax rate of 41.1 percent.
2 This would not affect cost-of-living increases in post-retirement benefits.