The elite press has urged Americans to be anxious about their Social Security pensions. Newsweek, the New Republic, and the Economist, among others, express a common fear that the system is going bankrupt and dragging down the economy. Polls show that the public supports the principles of Social Security but worries it will not be there when they retire. There is a building sense that the system has somehow not kept up with the changes in the nation's circumstances-in particular the age distribution of the population and the performance of the economy.
Peter G. Peterson, the reigning guru of gloom and the angel of the Concord Coalition, summed up these fearful attitudes in a recent feature article in the Atlantic Monthly. On the cover, above the obviously panic-stricken faces of Americans peering over the pension precipice, the headline reads "Social Insecurity: Unless We Act Now, The Aging of America Will Become An Economic Problem That Dwarfs All Other National Issues."
We too believe that Social Security pension reform belongs on the national agenda. But what is the core problem to be reformed? Is there a financial crisis that demands not only urgent action, but major revamping of the system? Must we "privatize" Social Security in order to save it? Or are circumstances that might warrant relatively minor adjustments in finance being used as a pretext for a more fundamental shift in national commitments?
We think Social Security finance requires prudent adjustment, but not major revision. Proposals to "privatize" substantial portions of the pension package have no relationship to solving financing problems. The real issues involve profoundly political choices, not technical ones: Do Americans want a society that insures all workers and their families against the dual risks of dying too young and outliving their private retirement savings? If so, we can "socialize" these risks only through collective social insurance. Or, do Americans want a society that merely mandates savings and investment for retirement, while leaving the ultimate security of workers and their families to be determined by their market success? If so, privatization is an appropriate response. But proponents of privatization are not practicing full disclosure. Privatization has much more to do with ideological preferences and economic interests than with the solvency of Social Security.
THE REAL CHOICE
According to Robert M. Ball, former commissioner of Social Security in both Democratic and Republican administrations. [See "A Secure System"], the system can be stabilized through moderate adjustments in pension formulas coupled with a new investment strategy. As Ball observes, the system is today accruing substantial surpluses, and total income will exceed outlays until about the year 2020. Thereafter, Social Security reserves must be retired to pay current benefits that exceed the level of current taxes. By 2070--that is, 75 years from now--benefits are projected to exceed taxes by about 5.5 percent of taxable wages. So unless some adjustments are made in benefit levels, taxation levels, or trust fund earnings, Social Security's retirement program would not be able--on current forecasts--to pay all its bills, as early as 35 years from now. From this perspective, Pete Peterson is technically correct: The current system is "unsustainable."
On the other hand, Peterson's assertion that something very like the current system cannot be financed is nonsense. Indeed, by gradually shifting to a partially funded system that invests in equity securities, there are ways to close this projected gap in future funding with no tax increases and extremely modest changes in current benefit levels.
For example, under the Social Security Advisory Council's option proposed by Ball and five other members, the total 75-year deficit can be eliminated, and then some, by six modest adjustments:
- extend Social Security coverage to currently excluded state and local employees;
- increase the length of the computation period for workers' average earnings from 35 to 38 years;
- tax Social Security benefits that exceed already-taxed contributions, as with private-defined benefit retirement plans;
- correct the current overstatement of the consumer price index used to calculate cost of living increases in accordance with the Bureau of Labor Statistics March 1996 proposal;
- credit income taxes on Social Security pensions to the pension rather than to the Medicare trust fund;
- shift approximately 40 percent of the Social Security trust funds, now invested in Treasury securities, into equity securities.
Not only can something very close to the present system be financed, but financing it is not very hard, so the proposals for more radical revisions reflect other goals.
For example, 5 other members of the 13-person Advisory Council favor substantial privatization. Every recipient would get a personal security account (PSA) financed by a diversion of nearly 40 percent of the payroll taxes that currently finance Social Security. Of the current FICA tax of 12.3 percent of taxable payroll, 5 percentage points would be diverted to PSAs.
This system would provide an extremely modest basic pension of $410 a month, which is considerably lower than the average benefit under the current supplemental security income program (SSI), the means-tested portion of Social Security. The money diverted to personal security accounts could be invested in financial vehicles of a worker's own choosing. Upon retirement, workers would receive the base pension, and whatever had accumulated in their PSA would be available for lump-sum payment, investment, or annuitization.
Because the PSA proposal diverts 5 percentage points of FICA taxes from the trust fund, additional revenues and benefit reductions are required to maintain pension benefits for current and near-term retirees. The PSA plan would raise the retirement age to 67 by 2011, increase payroll taxes by 1.52 percent, and increase public borrowing by an additional $1.2 trillion over the next 35 years. [For further details, see Joseph F. Quinn and Olivia S. Mitchell, "Social Security on the Table," TAP, May-June 1996.]
According to the Office of the Actuary at the Social Security Administration, each of the Advisory Council proposals satisfies the statutory requirement that the scheme be in long-term (75-year) actuarial balance. But, if that is true, why go beyond the Ball solution? The real debate is between those who view social insurance as worth saving and those who want to reduce government's role in retirement security in favor of market outcomes. To simplify matters, we will illustrate the true stakes in this debate by contrasting the Ball plan with the PSA proposal.
RISK AND REWARD
Both plans tie Social Security's fortunes more closely to the performance of private capital markets. The Ball plan would retain a collective system, while the PSA proposal would create individualized, funded accounts. Over the last 60 years (the life of the Social Security pension system) capital markets have out-performed real wage growth by several percentage points. In principle, this change will improve the intergenerational equity of the system by allowing future retirees (who will collect less of a windfall than their parents) higher returns.
Nevertheless, there are risks. Tying Social Security retirement pensions to the performance of the capital markets would have looked quite bizarre in 1935, when the Social Security system was constructed. And even if we do not experience another Clutch Plague, the events of October 1987 should remind us that the stock market can make quite precipitous "corrections." Although on average, over long periods of time, one may do better by investing in the market, some individuals and some cohorts of individuals will do substantially worse.
This concern about smoothing out the vagaries of market returns underscores the first striking difference between the Ball plan and the PSA model. In effect, the Ball approach puts the market risk on the government--on all of us collectively. The PSA model puts that risk on individuals. The choice between these proposals is thus similar to the choice between a "defined benefit" and a "defined contribution" retirement program. Private savings, whether in IRAs or otherwise, are the equivalent of a defined contribution pension plan. Workers save a certain "individually defined" amount and have assets at retirement equal to the performance of their portfolio.
Social Security, by contrast, has always been similar to a defined benefit scheme. The federal government makes long-term promises to pay certain pension benefits and bears the risks that the performance of the economy will make those promises either harder or easier to fulfill. Historically, many private firms have also provided defined benefit retirement plans. But only about 30 percent of Americans now retire with a company pension, and companies are rapidly shifting from defined benefit to defined contribution pension arrangements. In short, Social Security is the only vehicle available to most Americans in which they do not bear market risks to their basic retirement income.
The Ball proposal allows pensioners to capture the higher returns of financial markets, while keeping the risk collectively shared rather than individually borne. The projections for the PSA plan show it, on average, outperforming the Ball proposal. But this has nothing to do with "privatizing" the investment choices and everything to do with adding 1.52 percent to the tax rate, and hence the total amount invested. If we added the same tax hike to the Ball plan, the returns would be even greater than the average PSA because the administrative costs are lower.
ECONOMICS OR IDEOLOGY?
The real issue here is not economics but political ideology. Most privatizers want to privatize because they do not trust the government, or because they believe that the American people do not trust the government, or because of concerns that Social Security depresses savings rates.
The lack of trust argument takes two forms: In one incarnation it asserts that Americans prefer market risk to political risk. Hence to preserve the political acceptability of a mandatory retirement program, Americans will demand that they, rather than a government agency, should have control over investment decisions. The second form of the argument is that the government cannot be trusted to invest in the private capital markets without meddling with them as well. Neither argument is persuasive.
To give the first argument its due, privatization can be seen as an attempt--indirect, to be sure--to shore up confidence in the system. Americans who have "ownership" of an individual or personal security account might view their investment as more secure than a claim on the Social Security system. If so, this surely has more to do with the drumbeat of criticism in the media than any reasonable judgment of the program. The Social Security system avoids inflation risks, bankruptcy risks, and market risks. It has been running for 60 years without ever missing a payment. It continues to have the overwhelming support of the American populace, and Americans say that they are quite willing to pay some additional taxes to ensure the financial soundness of the system into the distant future. Nevertheless, it is conceivable that Americans will come to prefer risky over non-risky investments. Nor can we fully discount the Lake Wobegon effect--overoptimism among young workers that their lifetime earnings will be above average. If so, they would be increasingly susceptible to the argument that Social Security provides an inferior return on their contributions.
The real risk, however, is that partial privatization will lead to inexorable pressure for full privatization. Investment of some Social Security funds in equities, rather than Treasury securities, will of course improve the investment performance of Social Security. But if this investment is done in a privatized form, it will appear that the improvement has come through privatization of accounts rather than from a simple shift in investment holdings. And, because most workers tend to ignore the life insurance, dependents' benefits, and inflation protection that are a part of the Social Security pension package, this argument may be persuasive.
Even more importantly, workers may ignore the crucial protection that social insurance provides to everyone against low average lifetime earnings, poor performance of their individual investments, or against higher taxes or intrafamily transfers to support those who do have these experiences. The less stake that American workers believe themselves to have in the collective provision of retirement benefits through Social Security, the more likely political support for the system is to erode. Partial privatization in this scenario would be destabilizing rather than anchoring. It could lead to the very destruction of the economic security that "reform" was supposed to preserve.
The argument about government "meddling" in capital markets is even less compelling. Maintaining the soundness of and confidence in financial markets by massive governmental "meddling" is one of the great success stories of American public policy since the 1930s. Investing Social Security funds in equity securities will not roil capital markets, so long as investments are limited to broad "index" funds, these funds are managed solely in the interest of beneficiaries, and government is a passive shareholder. These constraints are not difficult to construct, as the experience of the Federal Employees Thrift Plan, the Tennessee Valley Authority and Federal Reserve Board's defined benefit retirement programs, and many state retirement funds illustrate.
In short, workers would not, on average, be better off investing privately for retirement than having those investments made through Social Security. Because private investing exposes beneficiaries directly to temporal fluctuations in the financial markets, privatizing accounts will make many worse off. There is no reason, other than ideological antipathy to government money management, for Americans to prefer the PSA to the Ball approach. There are obvious reasons for private money managers to prefer PSAs.
Critics of Social Security make much of the supposed burden that retirees place on the working young. By now, the image of the affluent old enjoying a secure retirement on the backs of hard-pressed wage earners is an established cliché. But the most recent complete data on the income of the aged for 1994 reveal that 56 percent of persons over 65 would be below the poverty line without their Social Security payments. Three-quarters of all recipients have total income, including their Social Security benefits, under $25,000 per year. Fifty percent have income under $15,000 per year. Families with an income of $50,000 or more represent only 9 percent of Social Security beneficiaries. Most elderly are not rich.
Reform of Social Security really poses two distributional issues-intergenerational fairness, and fairness within age cohorts. The intergenerational equity issue is mostly a distraction. The first generation of pensioners indeed enjoyed a windfall, but that is history. Both proposals aim to put Social Security pensions into long-term actuarial balance. Given that the burdens on current and future generations under the two schemes will be equivalent, the real issue is equity within generations. Here an ideological chasm separates the two proposals.
The Ball approach maintains the worker insurance model of equity that currently undergirds Social Security. Distributional fairness in this social insurance model is straightforward, but not always well understood. First, workers are insured against a lifetime of relatively low-wage work by a guarantee of a minimally adequate pension in old age. Second, recognizing that the level of wages includes some combination of personal circumstance and effort, the size of the pension increases with a worker's level of contribution. But it redistributes by giving lower-wage workers a better "return" on their lifetime earnings.
In other words, everyone signs up at the beginning of their working lives for a system that makes two promises: a minimally adequate retirement income for all workers, and a guarantee of higher returns (in absolute dollar amounts) to those who make higher contributions over their working lives.
The privatization approach proposes significant distributional changes. Under the PSA model, the pensioner is viewed as an investor. Supposedly, higher-wage workers who save more (and those who make more fortunate investments) are fully entitled to their better retirement situation. As a matter of individual dessert, the "investor" notion of fairness seemingly rewards individual prudence and self-denial--the decision to give up current consumption as a hedge against an uncertain future. Yet a mandatory requirement to save a fixed percentage of wages rewards neither prudence nor self-sacrifice. The saver, after all, did not choose to save. And sacrifice is inversely related to affluence.
The fairness of this shift is even more doubtful when the context of America's overall retirement policy is considered. Tax policy already offers greater subsidies to the retirement savings of higher earners than lower earners. The home mortgage interest deduction and the nontaxability of IRA, Keogh, 401(k), and defined contribution plans provide much more assistance for wealth accumulation to high earners than to low earners. The current structure of Social Security pensions somewhat reduces this imbalance. A shift to PSAs would eliminate an important equalizing feature of the overall retirement system. Given that "personal circumstance" as a determinant of lifetime earnings includes being born black or white, male or female, able-bodied or impaired, or into a rich or poor family, the unfairness of this approach seems manifest.
The imposition of substantial market risks on lower earners is also objectionable. The lower one's earnings over a lifetime, the more Social Security pensions matter to one's retirement security. And, other things being equal, the more Social Security provides one's bedrock protection against destitution in old age, the less likely one would be to prefer having that protection subject to market risks of the sort contemplated by the PSA proposal. After all, if one's investments went south under the PSA plan, one would be left with a guaranteed benefit of only $410 per month in 1994 dollars--less than current supplemental security income payments and much less than the poverty threshold.
The PSA scheme also trades a portion of Social Security's protections--survivor benefits--for ownership of the PSA, which passes to one's heirs at death. Security for younger workers and lower-wage workers' families is again being traded for increased benefits to higher-wage workers, and particularly to the survivors of those who do not outlive the value of their PSAs. This is not a trivial trade. Social Security currently provides life insurance valued at $1.3 trillion, more than all private life insurance policies currently in force in the United States. In short, the PSA proposal piles market and other risks on families who are poorly positioned to bear them.
The biggest winners in a shift to a PSA scheme would be families with two moderate- to high-wage workers. Each worker will have an individual account with no cap on expected returns. The PSA system thus would reinforce the distributional shift toward high-wage, dual-earner families that is occurring in the economy generally. Yet, while this is distributionally perverse in one sense, the PSA corrects what some view as a major inequity in current arrangements. Because spousal benefits now tend to smooth out returns to one- and two-earner families, some people currently view the FICA contributions of a family's second worker as "wasted." PSAs would eliminate this "unfairness."
While this position is superficially plausible, it is not compelling. First, the "loss" to high-earning couples is not nearly so great as Social Security's pension payment schedules suggest. High income is highly correlated with longevity. High earners will on average collect pensions for a considerably longer period than lower earners. Moreover, reforming the system to make it more favorable to two-earner families has no necessary connection to privatizing it. The same thing can be done, if we want to do so, through simple accounting changes.
Most importantly, this sort of group-by-group analysis of who gets what, on average, misses the basic point of social insurance. At birth, none of us knows for sure where we will end up socially or economically. We may be low- or high-wage earners. We may marry or not. Our spouses may work or not. Our working spouses may be low- or high-wage earners. We may incur misfortune.
The current scheme ensures us that however things turn out in life, our retirement benefit will not be too awful, and our dependents will be provided for if we expire before retirement. Should we be high earners with high-earning spouses, we will receive relatively low returns on our FICA taxes viewed strictly in terms of the size of our Social Security pension checks. But we will have insured ourselves against destitution and against the risk of having to pay higher taxes to support the destitute. This binding together of the interests of low and high earners in a common enterprise, before they know for certain who they are, is the essence of social insurance.
ALTERNATIVE AGENDAS AND RED HERRINGS
Some advocate privatization as a necessary means for increasing national savings, which they see as key to increased prosperity. These claims should have no bearing on the current Social Security debate for three reasons.
First, the Ball plan and PSA proposal are indistinguishable with respect to savings rates. If we believe, somehow, that investing trust funds in capital markets is good for savings and investment, the Ball plan does the job just as well as PSAs. Second, neither plan necessarily increases national savings. Both take money that is now invested in government securities and put it into private securities. But, if the government continues to tax and spend at the same rate, it will simply have to borrow at the same rate elsewhere. National savings can increase only if overall private savings goes up at a rate not equaled by government borrowing. Changing the form of Social Security trust fund holdings will not do the trick, nor will relocating those holdings into private security accounts.
Third, the capacity of the economy to pay projected benefits in 50 or 75 years is a function of the health of the future economy, not simply of the private savings rate. And the rate of economic growth reflects a variety of factors--public and private investment, technology, employment, human capital, business profits--not just the rate of personal savings.
The latest attack on Social Security is only a variation on familiar themes. Opponents today attack Social Security's "unfunded liabilities" and "trust fund fakery"--two of the Concord Coalition's favorite phrases. But conservative counterparts in 1939 demanded that the program be put on an unfunded, pay-as-you-go basis because government was not to be trusted with management of retirement-fund reserves. In 1983 and again in 1996, Social Security opponents claimed the system was going bankrupt--yet in 1989 they also attacked it for running huge surpluses. If bankruptcy and surplus are both crises, something other than financial prudence has to be driving these concerns.
Beneath the surface of a technocratic debate about Social Security's tax rates, benefit schedules, and long-term financial projections, there is a deep ideological divide between opponents and defenders of social insurance. Defenders, like us, see necessary adjustments as a natural and inevitable evolution. Opponents see these occasions as opportunities for radical revision--times for convincing the American public that social insurance is an unfair, unsustainable sham.
Privatizing reforms can be a legitimate means to public objectives where government failure is likely, and where private markets reliably produce superior results. We can fairly debate whether private contractors, or vouchers, introduce beneficial competition in many public programs, without compromising their purposes. Privatizing Social Security, by contrast, is a contradiction in terms. It not only fragments our most successful and universal program, but changes the entire dynamics of the program in order to fit the imperatives of private markets. By itself, the market can provide investment vehicles, but it cannot supply social insurance.