Social Security: Drifting Off Course (Fortune, 1967)November 20, 2011: 10:00 AM ET
Editor's note: Every week, Fortune.com publishes a favorite story from our magazine archives. This week, as the bi-partisan super committee faces a deadline to find ways to cut the deficit by $1.2 trillion, we take a look back at one of the programs that's on the table: Social security. As this 1967 Fortune story shows, the "swollen system" has been a political hot-button issue for decades.
The benefits are rising but so is the tax burden, and there's a question whether the swollen system is relevant to a modern America's needs.
By Dan Cordtz
FORTUNE -- Very few politicians have cared to challenge the assertion of Dr. Paul A. Samuelson, the M.LT. economist, that "social security is, by all odds, the most successful program of the modern welfare state." And small wonder. No other program has given so much to so many. Every month the Treasury mails out some 15,600,000 checks ranging from $16.50 to $368--a total of $20 billion a year in benefit payments. Nearly one American in eight receives some financial support from the system; it provides half the income received by those over the age of sixty-five. The system's political invulnerability is demonstrated repeatedly. Just recently, members of both the House of Representatives and the Senate have overwhelmingly supported legislation to liberalize benefits.
And yet in the last year or two a new current has emerged in public discussion of social security. The almost reverent tone once adopted ritualistically in such discussion is heard less often. One reason for the change seems to be demographic: recently, about 2,200,000 Americans have been entering the labor force in a typical year, while only 675,000 have been retiring under social security annually. As those millions of young workers reduce the average age of the labor force, they also reduce the numbers of Americans who view social security mainly in relation to its benefits. A young man in his twenties is more likely to think of the program as a tax than as a benefit. Meanwhile, the natural tendency of younger people to be more concerned about taxes than about benefits they may receive far in the future has been fortified by the series of steep increases in social security taxes--and the future steep increases now projected.
When collections began in 1937, no worker paid more than $30 a year into the Old-Age and Survivors Insurance trust fund. Now the maximum tax has reached $257 and Congress has been working on legislation to increase this figure to $540 a year. Since no exemptions are allowed, a middle-income family with several children and a substantial home mortgage might find its social security deductions as formidable as its federal income tax. Most legislators are confident that voters will accept a high level of social security taxes, because everyone can hope for a direct return on his money someday. But members of Congress admit that they are starting to hear grumbles.
Taking a closer look
Until recently, few Americans had really paid much attention to the social security tax. They have been largely unaware of the fact that, overall, 18 percent of all federal taxes are raised through the social security system. But as they begin to focus on it, they may well conclude that social security taxes--and the system itself—have become less equitable as they have grown over the years. There are, in fact, a number of reasons for concern:
--In a country that is firmly committed to the principle of progressive taxation, these taxes are highly regressive--i.e., the poor pay a higher proportion of their incomes than the well-to-do. And the taxes are likely to become still more regressive. Income levels probably will continue to rise faster than the ceiling on taxable earnings, so that relatively fewer Americans will be taxed on all their wages.
--The trust funds from which benefits are paid are expected to play a large fiscal role--a deflationary one--over the next decade. Under the present tax benefit schedule, receipts might exceed benefits by some $100 billion in this period, with the surplus to be squirreled away in special government bonds. Thus the program will create a "fiscal drag," as, indeed, it has at various times in the past. To be sure, Congress may elect to eliminate this drag by redistributing the surplus in the form of higher benefits. But that would exacerbate another problem associated with the system.
--The problem is that social security is a very inefficient and unfair way to redistribute income. Instead of redistributing income from the prosperous to the needy, the system now gives to the aged poor and the aged rich alike, principally at the expense of younger, middle-income families. And social security of course, does nothing at all for some of the poor--those who, because of one handicap or another, have not been employed most of their working years and cannot qualify for benefits even under the liberalized eligibility rules.
--Finally, some proposals for raising benefits pose a threat to the private pension schemes on which a growing number of Americans now depend for retirement at more than a subsistence level. The proposed heavy taxes, on employers as well as workers, could preempt the resources available to purchase more flexible and individually satisfactory pensions.
Some reasons for strain
These considerations are leading a growing number of economists and welfare experts, some of them inside the U.S. Government, to propose changes in the system. Late last month the Joint Economic Committee of Congress published the first three volumes of a compendium of more than sixty scholarly papers written in response to its 1966 staff study, "Old Age Income Assurance: An Outline of Issues and Alternatives." Although the proposals range across the spectrum, in one way or another most of them suggest that new methods be found to pay for some of social security's costs. The Brookings Institution next year will publish a book that also argues for changes in the way the system's benefits are financed.
From one point of view, there is no reason to be surprised that the system is showing signs of strain. It wasn't really very sophisticated in the first place--especially in comparison with some social-insurance schemes that had existed for years in European industrial countries. (Germany launched the first contributory, wage-related old-age pension system under Bismarck in 1889.) The U.S. program was an ad hoc response to the worst depression in the nation's history. The major concern of most of the planners that President Franklin D. Roosevelt brought to Washington was to find a way to provide an immediate source of income for those temporarily unemployed--not a permanent source of income for retirees.
Their efforts to aid the unemployed resulted initially in a nationwide, state-administered system of unemployment compensation. But the political threat of the Townsend Plan (which proclaimed a goal of $200 a month for every citizen over sixty) was hanging overhead, and so one small group of New Dealers was asked to come up with something for the old folks. They were given no detailed instructions, and carried out their work almost ignored by higher government officials, recalls Dr. J. Douglas Brown, now economics dean emeritus at Princeton and then one of the three members of the advisory committee that framed the legislation. The program they devised, later enacted almost without change by Congress--it became law in August, 1935--was a wage-related old-age pension for retired industrial and commercial workers to be supported by a specially earmarked payroll tax levied equally on workers and their employers.
From the start, Brown says, the scheme was jerrybuilt. He and his colleagues gauged what they could hope to get enacted and planned to improve and broaden it later. In fact, the politicians began tinkering with the benefit structure even before the first payments were made four years later. The original program, anything but a comprehensive and completely worked-out system of social insurance, was filled with compromises. For the architects were faced with a difficult problem of national attitude. Not even the trauma of the depression had overcome the Puritan strain in the American ethos, which held to the precept that hard work and thrift would always be rewarded and that turning to public charity--even in times of general economic disaster--was a disgrace. Even those who most needed help were unwilling to accept a permanent pension system that smacked of a government dole. And so the plan was packaged as one whose chief aim was, not relief for needy older Americans, but "insurance" for all. It was designed to provide only limited retirement protection and assumed that each beneficiary, by paying for his own pension over his working years, could accept his payments in old age with head held high.
How social is social insurance?
The New Deal social planners were shrewd judges of national psychology and political reality. But in emphasizing "insurance," they immediately encountered a large practical difficulty. Any rigid adherence to the usual rules of the insurance business--in which size of pensions is related strictly to size of contributions--would make it impossible to carry out the immediate social requirements, i.e., help for the elderly poor. And only if the date for beginning pensions was put off far into the future could early retirees possibly contribute enough money to justify even a minimal benefit level.
Thus the insurance concept was compromised to attain social objectives. A large share of the employers' contributions was assigned to subsidizing the near-term benefits. For the longer run, moreover, a minimum benefit with any economic meaning could never truly be "paid for" by the taxes on a worker with low wages. His contributions would be small, yet precisely because his wages were low his pension could not be much lower if he was to have enough to live on. Accordingly, the benefit scale was weighted heavily in favor of low-wage workers and, from the outset, they got back much more in relation to their tax contributions than did better-paid employees.
Such an arrangement was quite defensible, of course. Supporters of the system point out that, even in private pension plans, heavy company contributions are frequently made to finance benefits for workers who retire soon after the programs are established. They also argue that in social insurance the "social" aspect is just as important as the "insurance." But with every upward shift of the benefit level, the connection between contributions and benefits has become more tenuous; indeed, one can fairly describe the continued official emphasis on social security's wage-related character as little more than a politically useful fiction.
Of all the hard facts about social security nowadays, probably the most difficult to justify is the fact that the taxes soak those of moderate means more heavily than they do the rich. A married man with four children earning $5,000 a year now pays a tax equivalent to 3.9 percent of his gross income. A bachelor earning $50,000, on the other hand, is charged only 0.5 percent of his income. When the social security tax was first collected thirty years ago, its regressive character was much less pronounced. Although the levy applied to only the first $3,000 of income, that figure took in the entire earnings of 97 percent of those covered. Today, with the taxable wage base set at $6,600, only 75 percent of American workers pay social security taxes on their entire income. Even under the Administration's recent proposal to raise the base to $10,800 by 1974, just 87 percent of all workers would be taxed on all their wages.
There is no longer much argument about the proposition that social security may have significant fiscal effects. These effects were, of course, quite unforeseen by the system's designers. But given the size of the system today, and given the public's manifest uneasiness when the trust funds pay out more than they take in, they tend to run surpluses arid so tend to restrain the economy in most years. This year, for example, the system will extract about $25.3 billion in taxes and pay out about $21.5 billion in benefits, which means that Americans will have had about $3.8 billion less to spend this year than they otherwise would have had. Since this has been an inflationary year, it might be argued that the restraint was desirable. In fact, during last month's Senate debate on social security, it was argued--by those wishing to raise benefits--that the government should continue this fiscal restraint early in 1968 by raising taxes even more sharply than it raised benefits; some Senators contended that the higher social security taxes could serve as a substitute for the proposed 10 percent income-tax surcharge.
Aside from the rather special current situation, those who support the system claim that in general it has a desirable countercyclical tendency, helping the economy when it is weak and dampening it when it seems too explosive. The argument is that in a weak economy taxes decline because payrolls fall, while benefits rise because marginal workers retire and begin to collect pensions; in a strong economy, both tendencies are reversed.
But some economists who have studied the system feel that it is too rigid and cumbersome to be used directly as an effective tool in fiscal policy. In a 1966 study, "Some Fiscal Implications of Expansion of the Social Security System," Nancy H. Teeters (then of the Federal Reserve Board staff) describes social security taxes as "the least responsive to changes in aggregate income" and warns that by increasing them "we will be limiting the built-in flexibility of the federal revenue system." From the fiscal point of view, she wrote, it would be desirable to make the tax "more responsive to changes in economic conditions." (Mrs. Teeters also wanted to make the tax less regressive.) While there is some countercyclical effect, it seems clear that on balance the tax has a strong deflationary tendency. According to current contribution-benefit schedules, the system is expected to remove $100 billion from the economy over the next decade; the sum would pile up in the O.A.S.I. trust fund. In its 1965 report, the government's Advisory Council on Social Security expressed concern over "the deflationary effect of the present contribution schedule in the years just ahead" and urged "a large reduction in the size of these accumulations." In short, those fiscal effects are not so desirable.
The poorest get a bargain
If social security is a cumbersome device for making fiscal policy, it is also impractical as a device for redistributing income. The redistribution arises out of a benefit structure heavily weighted on behalf of those at the lower rungs of the wage ladder. Dr. Elizabeth Deran, senior research analyst of the Tax Foundation, has analyzed the tax-benefit relationship in a study included in the Joint Economic Committee's compendium. Under the program's initial schedules, she notes, the maximum tax contribution entitled a qualified retired worker to a monthly pension equal to 24 percent of his taxed monthly wage. A retiree qualifying for minimum benefits, however, could receive 60 percent of his wage. The low-income worker, therefore, received some two and a half times as much for each dollar he paid into the system as the employee at the top of the scale. In 1950, when the first tax schedule changes went into effect, the multiple rose from 2.5 to 4.4. It has risen steadily until the low-wage worker today gets 8.5 times as much as the high-wage worker for each dollar of taxes paid. And under the latest set of Administration proposals, the multiple would go up to 13.5.
To the system's defenders, this rather drastic imbalance in benefit payments seems justified because it tends to offset the regressiveness of the tax. In other words, the tax overpayments made by the poor while they work are compensated for by the benefit overpayments they receive when they retire. But clearly, there are better ways to redistribute income. For one thing, there are sizable windfall payments made under social security to some who are not at all needy. Included among those who receive heavy benefit payments in relation to the contributions they have made are some who earn pensions as schoolteachers or government workers, but who also work just long enough in private jobs to qualify for social security payments as well. Miss Deran writes of "an astute lady who was anticipating retirement from administrative work in a public school system not at that time under social security. She persuaded her brother, who owned a large department store, to hire her to tie bows for gift wrapping, spending just enough time at the chore for the $50 quarterly earnings requisite for coverage. As she pointed out, she certainly didn't need the income from bow-tying, but it was silly to pass up the social security for which she could so easily qualify, and she accumulated quarters of credit just as assiduously as she accumulated growth stock."
A poor poverty program
Despite that 8.5-to-1 ratio, furthermore, the minimum pension benefits are pathetically inadequate to anyone entirely dependent on them. They amount to only $44 a month for a single person or $66 for a couple. The Administration has proposed an increase to $70 and $105 respectively, and the House has authorized $50 and $75. In mid-November it was still unclear which levels would become law; but even the more generous levels would still leave in poverty recipients who have no other resources. Many of them would obviously still be obliged to ask for Old-Age Assistance, which normally requires a demonstration of need involving precisely the kind of humiliation that an "insurance"-oriented social security system was supposed to avoid.
Defenders of the social security system have several kinds of answers to such criticisms. Commissioner Robert M. Ball of the Social Security Administration says that it is unfair to characterize the system as an inadequate poverty program. Says Ball: "One objective of the program has always been to contribute to the prevention of poverty. There is not the slightest question that the social security program has been the biggest anti-poverty program in this country. But it is not solely a poverty program. It contributes to the economic security of anyone who works-provides him with a base upon which he can build." But Ball's boss, Under Secretary Wilbur J. Cohen of the Department of Health, Education and Welfare, acknowledges that there is now "moderate opposition" coming from younger workers who are not poor and who have doubts that the program is a good deal for them.
These doubts have to do with a subsidiary question about social security. The question is often raised by those who persist in thinking of the system as though it really were an insurance plan--and, accordingly, want to know whether, by private insurance standards, they're getting their money's worth. Dr. Colin Campbell of Dartmouth contends that, by any such standards, every worker now under forty is being short-changed, with the youngest and best-paid suffering most. To make his point, Campbell uses the example of a twenty-two-year-old who is entering the work force this year at a salary of $6,600. Assuming only presently scheduled increases in tax rates, the worker and his employer will accumulate a total of $68,000 during the forty-three years before he retires. (The figure includes $41,000 that the, worker could have earned in interest on the money if it had not been held by the government.) If 20 percent of these contributions are assumed to have paid for the disability and life insurance protection that goes with social security, then $54,000 was in effect used to purchase the worker's retirement pension. But if he and his wife live for the statistically average fourteen years past retirement age, only $32,000, with the interest it would earn, would be needed to pay benefits under current schedules. Thus, Campbell asserts, the average young worker at the upper end of the taxable income scale is being overcharged by $22,000 for the benefits he will receive.
Robert J. Myers, chief actuary of the Social Security Administration, contests Campbell's assertions. Myers argues that it is unfair to credit both the employer and employee tax to the employee's personal account and to assume that, if the employer's contribution were not taken by social security, it would then be available to the worker to put into a savings fund of his own. Most economists agree that, in the aggregate and in the long run, both the employee and the employer taxes fall on the wage earner. (This does not mean, of course, that every individual could expect to recapture his theoretical share of those employer-paid taxes if social security were eliminated.) And if the money were paid out in additional wages, it would then be subject to income taxes.
The adequacy of benefits
For purposes of his example, Campbell also assumed no changes over the years in benefits. Myers argues that this is wildly unrealistic. Because the working population is constantly expanding, and wage rates growing higher, he says, it would be possible to raise benefits in the future without raising either the tax rate or the taxable wage base on which it is applied. But it is also worth observing that in the past "fully covered" employees like our hypothetical twenty-two-year-old have not been getting significantly higher benefits except to the extent that they have had their own contributions increased.
It is clear, in any case, that criticisms can be directed at the adequacy of benefits paid both those who are fully covered and those who are covered only minimally. In both cases, the workers involved can legitimately feel that social security simply doesn't do enough for them. This raises a basic question about social security benefits: Should we try to expand the level of benefits dramatically, and make them sufficient to cover workers' total retirement needs-or should we operate on the traditional assumption that the system is only one source of retirement income, and that most Americans can expect to have other sources?
Even among those who take the latter view of social security, there has never been full agreement on just how much income it should provide. But it's doubtful that most Americans would consciously choose to tum social security into a system that tries to furnish the entire retirement income of most elderly people. Large numbers of workers now feel secure enough about their futures to prefer spending to more saving. In any case, it seems unfair to many of them to compel workers to save more than they want to in order to guarantee that they will not become public charges in their old age. And those who prefer to save may believe that social security isn't the best way to do it: they might also invest in securities or purchase real estate, for example, preferring to control the money themselves rather than see it used to increase their stake in social security.
Finally, of those who want pensions, an increasing number must be concerned about private group pension plans. It is true that such plans are still far smaller, in the aggregate, than social security. As of 1962, the date of the most recent detailed study of sources of pension payments, private group pensions furnished only 6 percent of the income of married couples over sixty-five. (Interest, dividends, and rents together accounted for three times as much and earnings for more than four times as much.) And although private pensions are growing steadily, it appears that even by 1980 fewer than 30 percent of retired workers will receive payments from them. Most plans involve lengthy waiting periods before pension rights are vested, and offer little protection to the worker who changes jobs several times. For those who do receive them, however, the benefits will often be larger than those received through social security. Many private plans can pay more than social security for two reasons: because contributions have been related to all the employee's earnings, not just to a "base" figure; and because the funds have been invested, on the whole successfully, in stocks as well as bonds. An average Ford production worker with thirty years of service now stands to get $172.50 a month from the non-contributory pension plan included under the company's contract with the United Auto Workers. A Ford executive who averaged $25,000 a year for thirty years might draw $1,070 a month from the separate contributory pension plan that covers salaried personnel. Any proposal to raise social security taxes drastically would make it harder to secure and expand such private plans.
Paying the bill
How might all these large difficulties of the present social security system be dealt with? Of all the proposals now in the wind, the most promising seem to be those that would reshape the present tax and benefit arrangements so as to deemphasize the role of the trust funds. Most such proposals envisage using the general revenues of the U.S. Treasury to support or supplement the trust funds.
So far, most discussion of this idea has taken place among academicians and welfare specialists. But it also has some powerful political champions. Senator Robert F. Kennedy has introduced a bill, endorsed by ten of his colleagues, that would have the Treasury gradually assume 35 percent of the costs of social security benefits. In effect, it would change the present two-part contribution system (taxes on employer and employee) into a three-way system, with employers, workers, and the government sharing the load more or less equally. His bill would also increase all benefits by an average of 50 percent, so that even with the general revenue contribution trust-fund taxes would rise sharply. By 1977, in fact, a man earning $15,000 a year would pay $750 in social security taxes. His employer would pay another $750 and the Treasury would contribute $900 out of general revenues to finance the benefit structure.
Kennedy wants to turn to Treasury support, primarily because he believes it is the only practical way to finance the much higher benefits he would like to see enacted. But there are a number of other arguments to be made for the idea of using general revenues; and the idea need not be coupled to a program of sharply expanded benefits. Perhaps the best reasons for moving in the direction of general revenues are that taxing could be on a progressive rather than regressive basis; that the fiscal effects of the program could be managed with much greater flexibility than at present (taxes or benefits could be raised or lowered independently, whereas now they are directly tied together); that any redistribution of income could take place as part of a declared and debated national policy, and not as a "side effect" of a pension program; and that there would, or at least might, be less financial pressure on private pension plans.
Support of social insurance from general revenues is a common practice in other Western industrialized countries. And it has been suggested by two special Advisory Councils on Social Security appointed in the past thirty years. But the idea is anathema to many influential U.S. politicians, including Chairman Wilbur D. Mills of the House Ways and Means Committee, which has original jurisdiction over all social security and tax legislation. "I'm unalterably opposed," Mills declares, and he is supported by Representative John W. Byrnes, ranking Republican on the committee. They fear that payments and taxes would go through the roof once the presumed "discipline" of self-finance was removed. They also believe that even if we started out using general revenues only for a portion of the costs, there would be heavy pressure to enlarge that portion and there would be no logical place to stop-short of billing the Treasury for everything.
There is a way, however, in which general fund revenues could actually be used to strengthen rather than weaken the self-disciplined, contributory nature of the present trust-fund system. The Treasury obligation, furthermore, need not contribute toward a social security system subject to continuous pressure for enlargement. One proposal, for example, would split off the disguised welfare or antipoverty tasks now assigned to the program and meet their costs out of general revenues. This would permit exclusive use of old-age trust funds to provide adequate, reasonably equitable retirement pensions on a truly wage-related contributory basis--and probably without further increases in the tax rate or unreasonable extensions of the wage base.
Minimum pensions for all
The outline of some such "dual-maintenance" approach, submitted as a possible alternative for study by responsible officials, is included in one of the papers in the Joint Economic Committee's compendium. Its author is Margaret S. Gordon of the University of California's Institute of Industrial Relations. Much simplified, the plan she describes would inaugurate a new minimum basic old-age pension to be paid as a matter of right to all persons sixty-five years or older. The recipients would not be required to prove they had no other financial resources; such a requirement could create a disincentive to save and prompt many elderly people simply to assign their assets to children or other relatives. In recognition of the popular resistance that probably still remains to such outright grants and as a matter of simple economic justice, however, an income test might be applied to limit the payment of pensions to those who really need them and to keep costs from rising as high as they would go if there were no restrictions. An income test would not impose conditions more objectionable than those already in effect under the social security law. (Benefits now are reduced if a recipient earns more than $125 in any one month.) Regular social security benefits could then be added to the universal minimum-and, of course, scaled down to reflect them.
Payment of the minimums from the general revenues would be expensive. Mrs. Gordon estimates the bill might run as high as $9.9 billion a year for $100 monthly pensions ($150 for a couple). But there would be important offsets. For one thing, the federal-state Old-Age Assistance program could be reduced substantially-perhaps eventually eliminated. There would probably be reduced outlays from the trust funds, and the reduction would permit a corresponding reduction in present social security tax rates. Obviously, assuming minimum benefits were increased substantially from their present levels, the total of payments under both parts of the dual system would be greater than today's aggregate outlays for retirement. But since payments seem headed higher in any event, welfare costs would be identified for what they really are and be shared in a more equitable manner by all taxpayers under a dual-maintenance plan. Some such plan would obviously minimize--it would not eliminate entirely--the problems associated with tax regression and fiscal policy in social security.
Dual-maintenance plans are not new; they are already in effect in several European countries including Sweden, Norway, France, the United Kingdom, Austria, and Italy. Canada established such a system in 1965. The concept is familiar to most economists concerned with social insurance. Not all favor it and few think it can soon be made into law. Professor Brown of Princeton says that anyone who believes the American people will accept such a universal pension system does not reckon with the continuing public distaste for "handouts." Even Mrs. Gordon concedes that it would be easier to put the minimums into effect only for persons over seventy, since few of them have any earned income and a high percentage are already drawing Old-Age Assistance benefits. Canada's plan initially covered only those seventy or older, although the eligible age is already being lowered.
Professor Samuelson, on the other hand, wants to go much further than Mrs. Gordon. He would prefer to see general revenues used to support the present benefit structure without restrictions on their use; he sees the dual-maintenance approach useful as an "opening wedge" for getting to that goal. "The first time we spend out of general revenues," he says candidly, "we are probably going to need some mumbo jumbo to do it."
New needs of a changed society
When the U.S. social-insurance mechanism was put together, it was not especially difficult to reconcile an antipoverty program with a system of appropriate old-age pensions for the overwhelming majority of American workers. At the time, most workers were poor and only a very visionary planner could have anticipated that in three decades they would attain their present level of affluence. But they have attained it; and, while poverty is still a large and nagging problem, it is plainly of an entirely different order--and requires different kinds of solutions--from the problem of getting retirement income for most workers.
As a national ailment, poverty should be the responsibility of all society. The burden of financing old-age pensions for the poor, who cannot earn them through their own contributions, should be shared by all taxpayers, not just by wage earners. But the heaviest load should certainly not be borne--as it is under the social security trust-fund tax system—by middle-income families.