The Salvation of Social Security

Social Security remains sacred in American politics while it threatens the entire economy. In the recent election campaign, practically all the candidates promised to “preserve Social Security,” to “resist any cuts in benefits,” and to “protect the elderly poor.” No one dared to say that without major reforms—including “cuts”—the Social Security system will run huge deficits, that these deficits will push our children into a situation of economic stagnation and social conflict and create a potentially disastrous situation for the elderly of the future.1

Politicians cannot hide a truth this important. When a recent national survey asked workers whether they felt they would end up receiving their Social Security benefits, 84 percent between the ages of eighteen and forty-four said no. Evidently many people sense that Social Security is in very serious trouble. Why do so many politicians pretend otherwise? They do so because they fear that the voters will punish messengers of bad news, and the future of Social Security is particularly bad news.

The elderly are bound to feel threatened, betrayed, and angry if politicians speak of taking away the one thing that has sustained many of them. At the same time, younger generations tend to see the elderly as far sicker and poorer, far more disabled and homeless than they really are. Younger people also believe the elderly are more dependent on their children than they are—or think they are. A recent survey, for example, found that people below the age of sixty-five estimated that about 30 percent of the aged receive income from their children. But in fact only 1 percent of old people—possibly fewer—receive their principal support from children or relatives; fewer than 5 percent receive any help that could be considered income; and the elderly are twice as likely to be providing financial help to children as to be receiving it.2

Social Security benefits are obviously of special importance to the elderly poor. Reforms must be carefully structured to avoid harsh consequences for those who depend on the system. But most retirement benefits do not go to those below the poverty level. Social Security is in many ways a middle- and upper-income program; it has become a political sacred cow not because of a humane concern for the poor but because of a political reluctance to impose fiscal restraint on middle- and upper-income voters. Still, if such reforms are carried out soon, they need only involve a gradual reduction in the rate of increase of Social Security benefits, giving the current working population time to adjust, and providing—through expanded private savings and private pensions—the savings people need for their retirement and the savings the economy needs for its long-term health.


The Myths of Social Security

If Social Security is to be saved, the obfuscation that has served to hide the true problems of the system must give way to informed debate. Understanding is now blocked by a set of myths—fantasies, really—about how Social Security works, whom it benefits, and how it will fare over the next decades. That the system is riddled with flaws—financial, ethical, and ultimately political—may come as a surprise to most citizens.

Myth One: The problems are not serious. Events are rapidly undermining the myth that Social Security’s financial troubles are minor and temporary. In November the retirement fund had to borrow from the disability fund—the first such borrowing in the history of Social Security. By late 1983, the disability fund will run out of money to lend and then will itself face certain bankruptcy. Like two drowning swimmers, both funds will have no choice but to reach out to grasp the third fund, the Hospital Insurance fund (HI, Medicare “Part A”), which is itself sinking fast. Unless the economy undergoes a truly miraculous recovery, providing a vast infusion of new payroll tax revenue, all three funds—the entire Social Security system—will run out of money sometime in 1984 or 1985.

Yet the myth that the Social Security system is fundamentally sound persists. According to the myth, Social Security will be saved by payroll tax increases between now and 1990, which have already been legislated, and by the fact that by the end of the 1980s the last of the postwar baby-boom generation will become taxpayers while the less numerous prewar generation will retire. Meanwhile, we are not to worry if all three funds happen to go broke before this. During the 1990s all the funds will be running in the black—so the myth goes—and can easily repay any interim borrowings from the Treasury. As for the long term, even the myth makers concede that gradually deepening deficits will begin in 2000 or soon thereafter, but we are admonished only to “monitor” the situation and to avoid “rash” solutions.

Nearly everything is wrong with this picture. As long as retired people continue to receive benefits wildly out of proportion (currently, three to five times as large) to their lifetime payroll tax contributions, plus interest, the system will remain fundamentally out of balance. In a chain-letter scheme in which everyone makes payments to someone else, not everyone can win the jackpot—unless there is an impossible acceleration in growth of the number or wealth of new players. It would take momentous economic good luck for the system to achieve a temporary balance in the late 1980s and 1990s. When the baby-boom generation retires early in the next century, the system will disintegrate.

Let us examine the future in detail. The Social Security trustees regularly publish projections of the future financial condition of the trust funds—the so-called “optimistic,” “intermediate,” and “pessimistic” projections. Our estimate of the imminence and severity of the collapse of Social Security as we know it depends on which of the projections we believe. As I argued in the first part of this article, the so-called “pessimistic” assumptions are not merely plausible, but, to me, are the most probable of the three possibilities. On the record of the last decade or so, the “pessimistic” assumptions look, indeed, optimistic.3 Table I, on the opposite page, shows what we have in store on the basis of these “pessimistic” assumptions.


The “pessimistic” assumptions themselves need to be examined. For the next two decades, demographics will hold very few surprises since almost every new worker by the year 2000 has already been born and can be counted.4 What counts is our assumptions regarding economic performance, and of these, two statistics have special significance: the future unemployment rate and the future rate of real-wage growth. The former tells us how many workers are paying taxes into the system; the latter tells us how quickly each worker’s tax payments are rising or falling relative to the Consumer Price Index (i.e., relative to each retired person’s indexed benefits). Table II, on the opposite page, shows both rates under the “pessimistic” projection.


The single most important variable for the near-term Social Security outlook is real-wage growth. Higher real wages immediately translate into higher total payroll tax collections. The striking point about Table II is that compared with the past five, ten, or even fifteen years, the “pessimistic” assumption about real-wage growth looks rosy. If real wages between now and the year 2000 were to perform as they did during the past fifteen years (that is, to decline by 0.2 percent per year), then the Social Security deficits would be truly appalling—more than double the already unacceptable “pessimistic” figures of Table I.

We may reasonably hope that new investment and the maturing of new entrants into the labor force will make it possible to improve on the labor productivity record of the past and thus allow us at least to match the pessimistic projection. But working against this will be the fact that the era of cheap energy and “costless” pollution is over.

The near-term outlook for Social Security is much less affected by changes in the unemployment rate than by changes in real-wage growth. Looking at the projected unemployment rates in Table II, most people will, I think, conclude that the estimates, although unpleasant, are highly plausible. In recent years, most economists have come to agree that the so-called “natural rate” of unemployment (that is, the unemployment rate consistent with a rate of inflation that is not accelerating) will be substantially higher in the foreseeable future than it was in the late 1950s and throughout the 1960s (when between 4.0 and 4.5 percent seemed about the norm). Young people today are taking much longer to decide on careers. Moreover, the number of families with several earners has been rising dramatically. Wives, husbands, or grown children tolerate being unemployed more frequently, and for longer periods, when their families enjoy other sources of steady income. The “pessimistic” projection of 7.25 percent unemployment for the period between 1982 and 2000 looks favorable when compared with the current rate of 10.4 percent and the projections of most economic forecasters that the rate is unlikely to fall below 9 percent until 1984 or 1985.

Any realistic—let alone prudent—assessment of our economy’s potential should take the “pessimistic” economic projection, or something close to it, as a guide for the next two decades. Further into the future the forces of demography impinge on any calculations. The central demographic question is really quite simple: as the twenty-first century wears on, how many retired (or disabled) beneficiaries will there be for every tax-paying worker? The answer to this question will depend almost entirely on the future fertility rate and on future life expectancy. More children, of course, means more tax-paying workers, which means more Social Security tax revenues for the same number of beneficiaries. Longer life spans, on the other hand, will tend to increase the number of retired people to whom benefits must be paid.

These future variables are inherently unknowable. They will be strongly influenced by the subtlest cultural trends (such as the social acceptability of having one as opposed to five children) and the most recondite medical discoveries (such as the production of cancer-fighting antibodies). Yet their influence on the future of Social Security is profound. By the year 2050, the projected “optimistic” and “pessimistic” deficits diverge sharply: an optimistic deficit of 0.7 percent of the GNP versus a pessimistic one of 10.9 percent of GNP. Over two-thirds of this difference is due to the cumulative effect of differing demographic assumptions.

According to the “pessimistic” projection, women in future decades will have only slightly fewer children than they are having now (about 1.70 versus 1.85 per lifetime). At least two current trends may actually reduce fertility rates below this “pessimistic” projection: the widespread availability of family planning and contraceptives, and the rapidly growing female participation in the labor force. Because fertility rates and economic growth have always been closely correlated, we may eventually fall into a curious sort of vicious cycle. Sustained low birthrates will put a greater burden on our system of retirement entitlements, which may in turn ensure that our economic growth remains very low.

The “pessimistic” projection also assumes that the life expectancies of both men and women will continue to improve at about the same rate as they have over the last fifteen years. Again, this is not an implausible assumption. Demographers and medical experts have still not identified a “biological limit” to human life spans. The “optimistic” assumption, by contrast, is gloomy indeed. It projects that from 1980 to the year 2060 the life expectancies of both men and women at age sixty-five will increase only by about two years.5

There is, of course, deep irony in the terms “optimism” and “pessimism” as we have been using them here. Many babies and short lives are not everyone’s idea of an optimistic prospect, even if they brighten the annual report of the Social Security trustees. I have brought this to the attention of some liberals who resist serious changes in the Social Security system. The only future that would give the present system even the slightest chance of surviving is a future they would find abominable—a future of fantastic, limitless economic expansion, buoyant population growth, and stagnant life expectancies. Why should it be considered “liberal” to bet on such a future? Since the future of Social Security and of the economy is at stake, why shouldn’t we bet on the more prudent assumptions of the “pessimistic” projection? And even if we do have a happier economic future ahead of us, we needn’t worry about reluctance to increase benefits once again.

Myth Two: It is “my money.” Many retired couples I know have the understandable but nonetheless naïve view that they have a “contract” with the US government to get back “their” money from “their” account—by which they usually mean their lifetime tax contributions, without interest. They may hold this view because terms like “insurance” or “entitlement” or “pension” are used to describe Social Security. Whatever their technical nuances, such terms mean only one thing to most people: benefits belong to the beneficiary by right. Proposals to take any of them away are both unjust and immoral.

This view—probably the most damaging myth of all—bears no relation to reality. This year’s newly retired sixty-five-year-old who paid Social Security taxes during his entire working career and was an average wage earner contributed $7,209 in payroll taxes over the entire period. If we accept the estimates of longevity and inflation of the “pessimistic” projection (for the average wage earner with a nonworking spouse), he and/or his spouse will receive about $520,000, or about seventy-five times the amount that he contributed in dollars. In the first year of his retirement alone, this person received an initial benefit in January 1982 of $803 a month for an annual amount of $9,992.6 Put differently, in nine months the same person got back his lifetime tax contribution. This retired person or his spouse would continue to receive these payments, indexed upward annually to 100 percent of the increase in the Consumer Price Index, for a period of over twenty-five years—the average number of years that either the husband or wife would be alive to receive benefits, according to actuaries.

A more expansive definition of “my” money would include the employee’s contribution, plus interest. Then the payback period would be one year and ten months. An even more expansive definition would include both the employee’s and the employer’s contributions, again with interest. Here, our average wage earner could get back his money in only three years and seven months—still only a fraction of the expected twenty-five-year longevity of that worker and/or any nonworking spouse.

A far more refined “present value” analysis by James Capra, Peter Skaperdas, and Roger Kubarych has just been published by the Federal Reserve Bank of New York. The analysis elaborates what should be obvious but is too often a hidden point: a so-called self-financed program that returns more than is put into it (even with interest) is a logical contradiction—or, less politely, a Ponzi scheme. Clearly someone will have to pay up. Capra et al. conclude that the program as it exists today is “fundamentally flawed.” They quantify the extent to which Social Security is today a welfare program disguised as an insurance or retirement program. The person who retires and has earned average wages during his working life has a stunning “present-value ratio”—the ratio of the current cash value of all his expected benefits to his and his employer’s lifetime contributions with interest—of 5.0 if he has a wife who didn’t contribute to Social Security, and 2.7 if he is single.7

For a person receiving Social Security today this is, of course, a fabulous investment (which explains, in part, why the program is so popular). He could not have achieved a return remotely as good on any other type of “investment” made over the same period. The economy itself did not grow nearly fast enough for any “real investments” to show comparable results.

But what does this say about the argument that retired people have a moral claim on—an “entitlement” to—the level of benefits for which Congress has so generously provided? First, almost none of the person’s benefits are “his money” in any sense at all. Most of them are pure windfall benefits, unrelated to tax contributions made by him or his employer. Because people have come to expect this windfall, the government should be cautions about reducing benefits in absolute terms, at least. But that is because government should, wherever possible, avoid inflicting sharp changes on its citizens—a principle extending far beyond Social Security. However, the government does have every right to begin adjusting the annual growth of benefits immediately, in light of economic realities. The government has no “moral obligation” to provide windfalls forever.

Second, such vast windfalls could not continue forever, even if the government were of such a mind. Social Security proved a golden investment for today’s retired people. For tomorrow’s retired people it will almost certainly prove a bad investment, if not a monumental bust.

This is a matter of simple logic.

  1. For Social Security to produce such vast returns—windfalls, compared to “real” investments—benefits must grow faster than the general economy. This they have done. Since 1960 total benefits have grown 3.0 times faster than GNP.

  2. Because benefits are financed directly from the current tax on payrolls, and because payrolls cannot grow faster than the general economy for any appreciable period, benefits can grow ahead of the economy only if the effective rate of payroll taxes continually rises. This is what has been happening. Since 1960 payroll taxes have climbed from 2.3 percent of the GNP to 5.9 percent of the GNP.

  3. But, as payroll rates rise, those paying the taxes must—as a matter of arithmetic—suffer a progressive erosion in the ratio of their ultimate benefits to the taxes they will have paid into the system. That is, their Social Security “investment” must show an ever-decreasing return.

This latter point—the ongoing erosion in Social Security’s return to its participants—raises staggering questions of both solvency and fairness. People retiring today will receive their lifetime contribution (employer-employee taxes plus accrued interest) in about three years and seven months. Under current laws, today’s baby-boom twenty-three-year-olds who retire in about 2025 will receive their contribution in about nine years—a 2.5 to 1 ratio in favor of those who retire today. And if future taxes are raised in order to close the huge deficits shown on Table I, the ratio in favor of today’s retirees would be much higher.

What then does it mean to say that the patterns and trends of today’s benefits are sacred and untouchable? It means that people retiring today have some sort of moral right to reap windfalls from a tax-benefit mechanism that will inevitably shortchange their children and grandchildren. I see no ethical logic in that position.8

Myth Three: The elderly are, by definition, needy. A major obstacle to Social Security reform is the pervasive belief that old people are, virtually by definition, needy. Unwavering support for the Social Security status quo is often considered a test of one’s social compassion and sense of fairness. Polls invariably show that most of us think that most old people are very poor. For instance, a Harris survey last year found that 65 percent of those under sixty-five believe that “not having enough money to live on” is a very serious problem for most of the aged, and that 54 percent believe that the aged are worse off now than they were twenty years ago.

Oddly enough, the aged themselves see things quite differently. The same survey revealed that only 17 percent of the elderly regarded low income as a serious problem for them personally, and that 58 percent of the elderly thought it was hardly any problem at all. In mean per capita income—adjusted for family size and including all sources of income—old people are doing better than people under sixty-five. Furthermore, retired people typically have lower expenses than working people (e.g., two-thirds of the elderly own homes without mortgages and they spend less on education and transportation), except for health expenses, where the elderly’s disadvantage is powerfully offset by Medicare. Remember also that Social Security benefits are tax free.

The real average income of the elderly has been rising, not falling. During the 1970s, in fact, it rose more than that of the general population. The percentage of elderly people below the poverty level has fallen from 35.2 percent to 15.7 percent during the last twenty years. When benefits such as Medicare and housing subsidies are included in income, the elderly population now shows a lower poverty rate than the general population.

Some elderly people are indeed poor, very poor. Income averages can be deceptive. The elderly population has a somewhat higher percentage of relatively low-income people than does the general population. For the elderly with low incomes, Social Security is vital to maintaining a minimal standard of living.

This is particularly so for what Stephen Crystal in America’s Old Age Crisis calls the “multiple jeopardy” groups—those who are over seventy-five (a group expected to grow by over 50 percent by the year 2000), widowed, single, or divorced, in poor health, without private pensions, and nonwhite. For example, the mean income of the black elderly was only 54 percent of the mean income of the white elderly in 1980.

Thus, poverty among some of the aged is a serious problem. We have just seen that Social Security is, in large degree, a welfare program rather than an insurance program. In that case, how well directed is it to the problems of the elderly poor? According to the 1980 census, those below the poverty line—today about $10,000 for a family of four—received only 9 percent of total Social Security benefits. Roughly another 30 percent of benefits go to those whose earnings range from the poverty line to twice the poverty line. By contrast, the elderly with incomes in the top 20 percent (with family income over $30,000) received about 30 percent of total benefit payments. The total Social Security outlay for such relatively well-off families is more than the combined outlays of the major need-related welfare programs—food stamps, AFDC, Medicaid, and Supplementary Security Income. In other words, because of Social Security, welfare for the well-to-do now exceeds welfare received by the poor in these welfare programs—even before one takes into account tax breaks (e.g., the mortgage interest deduction) especially designed for middle- and upper-income groups. I do not see how the claims of social fairness are met by holding inviolate the benefits of the well-to-do.

Myth Four: The elderly are physically unable to work beyond age sixty-five. Any sensible reform package for Social Security involves raising the retirement age beyond sixty-five years (e.g., to sixty-eight or seventy years). To many, this seems heartless. In fact, polls show that a remarkable 79 percent of workers nearing sixty-five would like to continue working, if not in hard physical labor, at least part-time. But a far smaller percentage actually continues to work. A major reason is the disincentive to work built into the Social Security program: each dollar of earned income (wages) over $6,000 a year cuts Social Security benefits by 50 cents.9 That penalty, along with the federal income tax, puts the working beneficiary into a 70 percent tax bracket or worse once he has earned over $6,000. Clearly, any coherent program of entitlements reform for the aged must make it financially attractive for people over sixty-five to work.

Two generations ago, the remaining life of a sixty-five-year-old man was an average of twelve years, and these were usually spent in deteriorating health. Today sixty-five seems comparatively young. At sixty-five, one can expect another 16.6 years of life. Every medical study indicates that responsible work, at a reasonable pace, improves the physical and mental health of most of the elderly. In Japan, most men work beyond age sixty-five and the average Japanese life span is growing faster than ours. In America only about one man in five works after reaching sixty-five.

During this period of high unemployment, it may seem economically perverse to expand the labor force by encouraging later retirement. But the economy will soon run the other way. The latter part of this century will be a period in which the growth of the work-force will slow down dramatically. Then we will especially need the skills of our most experienced citizens.


Principles of Reform

The myths being circulated to block reform all grossly underestimate the moral capacities of old people. Some assume, for instance, that today’s elderly would demand abnormally high benefits even if that meant destroying the prospects for economic security of future generations. Politicians are therefore reluctant to antagonize what they take to be a vehement interest group. Most retired people I have talked to simply do not feel that way. In my experience, many old people have an unusual interest in and sense of responsibility toward later generations. I know of very few old people who, when informed of the facts, would hesitate to accept reasonable adjustments in their Social Security benefits.

Similarly the myth that the old see no distinction between the moral claims on benefits of the poor and of the well-to-do has, so far as I’m aware, little basis in fact. Indeed, I would guess that most rich people would be uneasy about their indexed, tax-free Social Security benefits if they knew how much their checks exceed their original contributions.

Those resisting Social Security reform are protecting not the elderly but only a fanciful vision of the elderly. The “gray power” lobbies, which make a profession of blocking reform, have a vested interest in promoting these myths. But their constituents, I believe, would far more readily face facts and support constructive efforts aimed at improving the financial soundness and general fairness of the Social Security system.

Three principles should, in my view, underlie such proposals. First, we must recognize that raising payroll taxes once again is not a solution. That approach to saving Social Security will inevitably fail. To close the deficits in the Social Security system under the “pessimistic” projection would take a payroll-tax rate of about 44 percent in 2035. This simply is not going to happen. Working people would not accept it. Even if they did, the dead hand of such taxation would drive millions of unskilled workers into permanent unemployment. We worry now that high income-tax rates may be eroding incentives to work, save, and invest, and there has been much talk of a “flat-rate” income tax below 20 percent. Yet we are moving inexorably toward a system where payroll taxes for Social Security alone will scoop away more—much more—than 20 percent of wages and salaries.

The approach of raising taxes has already proved its ineffectiveness. Since 1949, average wages are up 470 percent, average income taxes 570 percent, average Social Security taxes 3,960 percent, and maximum Social Security taxes 6,480 percent. The employers’ share of payroll taxes will be equivalent to about 50 percent of 1982 corporate profits—against 7 percent in 1955. Today, fully one-quarter of our workers pay more in Social Security taxes than they do in federal income taxes.

And what is the result of all this tax raising? The result, as we have seen, has been to worsen the inequity between generations and to put off confronting the coming collapse of the system. Every new tax increase on Social Security has been followed by pressure to raise payroll-tax rates yet again. The latest proposal is to “accelerate” the Social Security payroll-tax-rate increases scheduled through 1990 but to offer an “offsetting” income-tax credit at the same time. Alas, this will simply widen the already grotesque deficit in the non-Social Security part of the budget. Ultimately this proposal is one more scheme for a self-deceiving and self-defeating tax increase.

The second principle is that we must avoid deficits in the long term. We are hearing much about “temporary” transfers from general revenues to “solve” the Social Security retirement and disability fund deficits. To that, I have two questions. First, what general revenues? As one senator put it: “There is no such thing as general revenues. Only general deficits.” Second, if we exhaust that nonexistent treasure now, what will we use to finance the enormous Medicare deficits facing us?

Finally, any reform proposal should be equitable. Under the current benefit and tax structure, the present generation of retired people will receive a much greater return on contributions made than will those in the future. At a minimum, Social Security reform should reduce this imbalance. Equity between generations therefore requires that restraint on benefits not be postponed. Reform should also seek to protect the neediest—those who are very old, have low incomes, and are in poor health.


A Program of Reform

How can we build a program that both assures financial solvency, without additional taxes, and satisfies the requirements of equity? Here is one proposal—offered in the hope that others will try to improve it.

One: Freeze cost-of-living adjustments (COLAs) for at least one year. This would make up for at least some of the excesses that occurred between 1978 and 1981 because benefits were tied to the Consumer Price Index, which has an upward bias. During that period, retired people received cost-of-living adjustments that were too high by 10.8 percent, compared with those measured by a more accurate index of inflation, the personal consumption deflator. A freeze on the July 1983 or July 1984 COLA could within a year save as much as $18 billion in the cost of Social Security and other federal pension plans.

Two: Limit future COLAs after the freeze is over. We are told that it is politically impossible to reduce Social Security COLAs. I am not so sure. Surveys of public opinion on Social Security show a great deal of confusion about these adjustments. For example, a recent survey conducted by the Gallup organization suggests that the public does not approve of automatic COLAs. Most respondents believed that adjustments ought to be annual decisions based on current circumstances.10 Also, among those who favored automatic COLAs, a majority favored adjustments equal to the automatic adjustments in labor contracts—which, historically, have equalled about 60 percent of the increase in the CPI. Holding COLAs to 60 percent of the increase in the CPI would go a long way toward ensuring the future solvency of the system and would affect current and future beneficiaries about equally, according to the analysis of the New York Federal Reserve. Savings could reach $21 billion by 1986.

An alternative plan would be to index the increase in average wages minus 1.5 percent. An advantage of this proposal is that it would reduce the vulnerability of the system to unforeseen economic developments. Wage growth is the primary determinant of how fast the Social Security system’s income from payroll tax revenues will increase, while COLAs are an important factor in how fast benefit payments will increase. If COLAs were limited to the increase in wages minus 1.5 percent, it would prevent a repetition of the experience of recent years in which rapid growth of the Consumer Price Index compared to wages depleted the Social Security system’s cash reserves even more quickly than would have been the case otherwise. In fact, this proposal for wage indexing could save $18.7 billion by 1986, almost as much as indexing to 60 percent of the increase in the CPI. However, in the long run the savings are probably less than half of those that would result from indexing to 60 percent of the Consumer Price Index. Indexing to wages minus 1.5 percent may be a little like buying weapons to fight the previous war. On the other hand, it is far preferable to the current method of full CPI indexing.

A freeze on cost-of-living adjustments and lower future COLAs need not hurt the poor. For those who receive benefits under Supplemental Security Income (SSI), the welfare program for the aged and disabled, the loss in income would be made up dollar for dollar with higher SSI benefits. Also, the restraints on COLAs could be applied selectively—they could exclude, for example, the Social Security beneficiaries who receive the smaller monthly benefit checks. By whatever means, we must continue to provide adequate benefits to the poorest, oldest, and sickest.

Three: Tax those benefits that are received in excess of contributions.11 For every other retirement program, whether private or public, any benefits that are received in excess of employee contributions (whether they are the result of interest on those contributions or represent the employer’s contributions or both) are subject to federal income tax. It’s unclear why this principle has never been applied to Social Security.12 Taxing benefits in this way would have a number of advantages. In the long run it has the largest effect of any proposed change except for the indexing reforms. It also reduces the imbalance in the returns on contributions between those who are now retired and those who will retire in the future. Finally, the proposal has no perceptible effect on the aged who are in serious need; their incomes are not large enough to be subject to income taxes, especially in view of the extra tax exemption available to the elderly.

Four: Selectively lower initial benefits for retired persons. When someone reaches retirement age, a calculation is made in which the lifetime monthly earnings on which he paid Social Security taxes are put into current dollars, or indexed, and then averaged. If he is single, he will receive, in 1982, 90 percent of his first $180 in average monthly earnings, 32 percent of all average monthly earnings between $180 and $1,085, and 15 percent of average monthly earnings between $1,058 and $1,258 (the latter figure represents the maximum possible average monthly earnings on which someone retiring in a 1982 could have paid Social Security taxes). We could cut future benefits, while still protecting the poor, by lowering the 15 percent figure or both the 15 percent and the 32 percent figures, which apply to the upper-income categories. This would reduce the replacement rate (the percentage of pre-retirement wages that are “replaced” by Social Security benefits) for those retired people who had a record of high earnings.13

Five: Raise the retirement age by at least three years, to age sixty-eight, on a gradual basis. Adding three months per year starting in 1990 would significantly improve the long-term financial status of the system.

Even with eligibility shifted to age sixty-eight, the average retired person would enjoy benefits for a longer period than did his parents or grandparents. Life expectancy has increased at a rate of about one year per decade. Many of the poorest elderly—those subject to the multiple miseries of low income, no assets, and poor health—are often unable to find work as they approach their sixties. They are often unskilled, and employers are unwilling to train them. I suggest not raising, and perhaps even lowering, the eligibility age for SSI as well as increasing SSI benefit levels in order to protect this group.

Six: Once these reforms are made, bring civil service workers into Social Security. Right now the retirement system for federal employees is 2.5 to 3 times more generous than are pensions for private workers. The entire federal pension structure should be rebuilt from the ground up, providing a pension plan no more nor less generous than private plans. While cashing in on unusually generous early retirement benefits, over 70 percent of government pensioners now launch second careers that make them eligible to tap into Social Security. At the same time excessive Social Security benefits are paid to retired federal employees who have worked only the minimum time to enable them to qualify for Social Security. This is not only unfair, it’s a waste of money.

If we bring civil servants under Social Security, the effect on Social Security solvency during the next few years would be substantial, since the system would receive an infusion of tax revenues from the new participants but would not have to pay additional benefits for a while. In the long run, this change by itself would simply make Social Security an even larger welfare program with more beneficiaries. Bringing federal employees into Social Security is an inevitable part of a much-needed reform of our federal pension system. However, because the proposal would have salutary effects in the near term, I fear that it may be used as a quick remedy instead of a long-term cure. Consequently, a program for reform of Social Security should include federal employees only if it is part of a comprehensive revision of both Social Security and the federal employee retirement system.

Table III, below, shows the effects of these various reforms in achieving the goals of solvency and equity for the pension system.


One result is that the system would become solvent, i.e., would be able to meet all its obligations, with no future increases in the tax rate. In other words, the total savings from these reforms about equal the currently projected deficits through the year 2000.

Seven: Reform Medicare and reduce medical cost inflation. Even if all the foregoing reforms were carried out, as we can see from Table III, there would still be a deficit in 2025, largely from Hospital Insurance, equivalent to 8.2 percent of taxable payroll. This requires that we go further, to reform Medicare. We have seen how swiftly rising Medicare deficits are a driving, even dominant force in prospective Social Security deficits. The Medicare program is the fastest-growing part of the Social Security system. Between 1972 and 1982, outlays for the Hospital Insurance trust fund increased more than five-fold, from $6.5 billion to $35.7 billion, 2.8 times as fast as GNP; disbursements per beneficiary increased 300 percent since 1972. Under the projections in the most recent report of the Social Security trustees, expenditures in the future are expected to increase between about 3 and 4 times as fast as GNP. Prospective costs are immense, especially in view of the huge increases in the number of people over seventy-five and the emergence of new life-saving technologies. The latter problem is already illustrated by the estimate, according to experts, that between one third and one half of Medicare payments have gone to keep elderly people alive in their last year of life.

Ultimately, the runaway expansion of Medicare can be slowed only by reducing the rampant inflation in hospital and other medical costs. This inflation is partly fueled by perverse incentives and archaic insurance arrangements that shelter the medical sector from normal competitive forces.

While reform of the Medicare program is beyond the scope of this article, the entire structure needs comprehensive reorganization with new incentives to reduce costs. For example, from the standpoint of the providers of health services, we have a system in which costs are determined retrospectively (that is, it reflects the sentiment, “You don’t get a dollar unless you spend a dollar”; “The more you spend, the more you get”). I would strongly support more experiments with prospective price systems, based either on unit prices for specific forms of care or on inclusive budgets for a package of services, provided through a hospital or local community, that would be negotiated in advance. From the standpoint of the patient (and his employer), there are now few incentives to keep costs down. Thus such notions as increased deductibles and co-payments, voucher systems, and treating some part of employer contributions for medical benefits as ordinary income for tax purposes must receive serious consideration. 14


By conventional political standards this program of reform may seem impossible. But without such a program the Social Security system, and economic growth itself, will become impossible. That will tend to alter political conventions. The real question, therefore, is whether we possess the wit and determination to change our own politics or whether we must wait until chaotic change is forced upon us, including a full-fledged rebellion by young workers, crushed by taxes, against the entire notion of supporting the elderly at a decent level of income.

Republicans and Democrats have both shared amply in the decisions which turned a sound, though controversial, program into a vast scheme of welfare for well-off citizens. The radical distortions in Social Security have had strong bipartisan support. The huge benefit increases of the early 1970s were supported jointly by Wilbur Mills and Richard Nixon. And both parties have failed to inform the public about the system’s problems—and have done much to give false reassurance about its future.

There is no partisan method to save the system. The rescue of Social Security is not a liberal or a conservative cause; it will depend on combining a realistic understanding of fiscal deficits with a humane sense of social fairness. Politicians are thoroughly convinced that any party or faction proposing serious reform will face certain annihilation at the polls. Inaction may remain smart politics for a few more years, but I am not so sure. When the system comes crashing down, smart politics will be small consolation. Thus Social Security poses one of the deepest challenges to democratic politics in our history, a challenge above all to the middle- and upper-income citizens of this country who largely determine the course of public policy. They are the ones willing to stake the future of Social Security on absurdly optimistic assumptions, and the ones to whom the system dispenses a big share of its welfare. The excessive Social Security and public pension benefits that flow today to the relatively well-off are, literally, stealing capital from tomorrow’s citizens and making cuts in government programs to the poor irresistible.

I can imagine few ethical goals more important than fairness to our children and fairness to the poor. If the children and the poor deserve better treatment, the sacrifices must come from those who are left—the adults who are relatively well off. I suspect that many readers of this article fit that category. It is from us that moral leadership is required. Someone must step forward and say, “Cut my welfare. I am willing to have my ox gored.”

There is, I think, some cause for hope. With the election behind us, Social Security now has a chance to become again a “problem” rather than an issue to be exploited. For example, a bipartisan group of several hundred people—former cabinet officers, business leaders, and academics—is about to make a nationwide appeal for balanced fiscal reform to the Congress and to the president. But this is only a start. Americans who are better off must now lead by example—by making sacrifices for the common interest. Otherwise, although America will of course survive, it will be a poorer country.

(This is the second of two articles on Social Security.)

  1. 1

    These economic consequences were discussed in the first part of this article in the New York Review, December 2, 1982.

  2. 2

    See Stephen Crystal’s America’s Old Age Crisis: Public Policy and the Two Worlds of Aging (Basic Books, 1982), which contains a great deal of useful information on the elderly and on attitudes toward them. I have also found most informative the statements and speeches of Robert Beck, chairman of the Prudential Insurance Company of America.

  3. 3

    Throughout the 1970s almost every “intermediate” projection for Social Security has, in the light of actual subsequent experience, turned out to be overly optimistic. I find today’s widely used “intermediate” assumptions optimistic and the “optimistic” assumptions sheer fantasies.

  4. 4

    One exception to the rule of “no surprises” may be immigration. Projections of the work force (and, therefore, aggregate earnings and the receipts of the Social Security system) are based on the premise that current immigration laws can and will be enforced. Either a flood of illegal immigrants or a substantial loosening of immigration restrictions could therefore significantly increase Social Security receipts by offsetting low birthrates (while avoiding the costs of child-rearing). Yet such an expansion of the work force would intensify the problem of insufficient capital formation—on which I have earlier laid considerable stress.

  5. 5

    The “pessimistic” demographic assumptions are that the fertility rate will become stable at 1.7 after 2005, that male life expectancy at sixty-five will reach 20.8 years in 2050, and that female life expectancy at age sixty-five will reach 29.3 years in 2050. For the “low-intermediate” (II B) assumptions, the figures are 2.1, 17.3, and 23.2, respectively.

  6. 6

    In July 1982 retired people received a 7.4 percent cost-of-living adjustment.

  7. 7

    These ratios include only retirement cash benefits. Inclusion of Disability and Hospital Insurance would make the ratios even larger.

  8. 8

    Other issues of equity lie beyond the scope of this article. One relates to sex and changing patterns of life and work. For example, retirement benefit levels are increased by 50 percent when a retired person has a nonworking spouse over sixty-five (generally a woman). Furthermore, working women get a larger return on their tax investment because they live longer. Homemakers, on the other hand, do not get any benefits for work not rewarded in cash (though divorced women are treated more favorably). For a lucid analysis of these tangled issues see The Coming Revolution in Social Security by A. Haeworth Robertson (Security Press, 1981).

  9. 9

    This is a particularly regressive provision when you consider that unearned income (e.g., interest income) has no effect on Social Security benefits. In other words, a retired person with little income would be penalized for earning wages, while a retired person who is receiving $100,000 per year in interest income would not be penalized for such income.

  10. 10

    I have been unable to find a single major corporation that provides automatic, 100 percent indexing for its pensions. Most provide no indexing at all, except on an ad hoc basis.

  11. 11

    In the past, some have suggested taxing 50 percent of benefits, under the assumption that the employee contribution was taken from income before taxes and that the employer contribution was tax-deductible. Such a proposal only perpetuates the myth that lifetime employee-employer taxes are equal to benefits.

  12. 12

    Curiously, on private pension plans, the retiree pays taxes not on contributions coming in, but on everything going out. On Social Security, it is the reverse.

  13. 13

    According to the Social Security Administration, net replacement rates (after taking into account taxes and work expenses) for maximum wage earners are much higher today (42 percent for a retired single person and 57 percent if he or she has a nonworking spouse) than they were under the original 1935 Social Security Act, where the net replacement rate was 35 percent for both single and married retired people. Since these are usually the people with the highest post-retirement incomes, reducing replacement rates to maximum wage earners would protect the needy. Another way to do the same thing would be to freeze the so-called bend points—$180 and $1,085—for several years, since under the current law these are raised every year on the basis of growth in an index of average wages.

  14. 14

    Unlike direct changes in the benefit levels of cash retirement programs, for which effects on outlays can be estimated with some precision, proposed changes in Medicare will have largely indirect effects on benefit levels (e.g., from the incentives provided by higher deductible amounts). The resulting effects on total outlays cannot be precisely predicted.