Adapted from Chapter 2
Social Security Principles and Rationales
Like most public programs, Social Security is structured around principles mediated by politics.1 This mixture of principle and politics is fundamental to policymaking in a democratic society, and a crucial issue is how and when those principles are allowed to influence the political process. In our judgment, developing such principles is the first requirement for effectively reforming Social Security. The political process can judge proposals as "better" or "worse" only when it has a common benchmark against which to compare them. The principles laid out here serve as that benchmark.
These principles are from the theory of public finance as it has developed over the past two hundred years.2 The original designers of Social Security appealed to these principles in crafting the system and in selling it to the public. They also made compromises, often tried to blur distinctions among principles, and on occasion—for political and other reasons—made decisions that were inconsistent with all the principles. Even so, it's valuable to determine the extent to which principles did guide the original program design.
In theory, it does not matter for the reform debate what was intended in the 1930s, as long as there is agreement on the principles to be applied in the future. But in practice the question of how and why the system evolved does matter, in part because agreement about future reform is easier if the basic objectives of the past are recognized and related to modern circumstances.
Besides presenting the principles themselves, we examine here the role of these principles played in the original design, justification, and promotion of Social Security. We look, in particular, at the extent to which Social Security's founders and most ardent defenders publicly recognized these principles as both goals and justifications for the program, even when they did not use the same terms as we do today.
We draw especially on the reports of the two committees that drew up its initial blueprints: the Committee on Economic Security of 1935, and the Advisory Council on Social Security of 1937-38.3 The former produced the detailed social insurance plan that led to the original 1935 Social Security Act. It set up a contributory retirement system that began collecting taxes in 1937 and was scheduled to start paying out monthly benefits in 1942.4 The latter commission made recommendations that resulted in extensive amendments to the program, among which were bringing the starting date for monthly benefit payments forward to 1940. These 1939 amendments form the basis for the Social Security Old-Age and Survivors Insurance system as we know it today.5 Discussion of Social Security's goals and principles can also be found in a variety of other sources, such as speeches by President Franklin Roosevelt, debates in Congress, and the reports of the various Advisory Councils on Social Security.6
PRINCIPLES FOR JUDGING REFORM Four principles underlying the theory of public finance provide a useful framework for thinking about and evaluating Social Security and other public programs:
- Progressivity (sometimes called "vertical equity") holds that government ought to redistribute at least some resources from the better-off to those with lower incomes, greater health problems, and so forth. The goal is to reduce inequality in the society or to assure that people are able to meet certain basic needs. Progressive goals can be met by directing public expenditures to the needier or by adjusting tax burdens according to ability to pay.
- Individual equity holds that the taxes an individual pays government ought to be related to the benefits that individual receives, and vice versa. In the context of social insurance, it means that insurance protection should be related to premiums paid.
- Horizontal equity holds that people in equal circumstances ought to be treated equally. People with equal levels of economic well-being, according to this principle, should pay equal amounts of taxes.
- Economic efficiency holds that an economy ought to operate so that individuals can achieve the maximum well-being possible with their limited resources. Market signals and freedom of choice are often important in attaining this goal.7
These four principles sometimes overlap and sometimes conflict. They also intersect what economists call "opportunity cost"—the measurement of how efforts to achieve one goal inevitably reduce opportunities to do other things. (For instance, Social Security's founders used the concept of "social insurance" to try to resolve the conflict between individual equity and progressivity.)
Redistribution from the Better-Off to the Less Well-Off (Progressivity)
We do not have to examine the statements of the founders of Social Security to intuit that one motivation dominated all others: the desire to meet the needs of the elderly. If in the 1930s the elderly had no needs that were not shared by the rest of the population, there would have been no Social Security program—at least none that resembles the program we know. Social Security was meant to distribute resources to the elderly from the rest of the population because the elderly as a group were considered less well-off. In other words, it was meant to be progressive. Two main features were adopted to achieve this goal.
First, a new tax system collected from the better-off (workers) to finance benefits for the less well-off (retirees). Benefits were to be paid only to those who had contributed at least some amount to the system during their working years. But most participants who retired during the first several decades of the program, even up to today, received benefits that greatly exceeded their lifetime tax contributions. The immediate effect of this process was to redistribute income from later generations that, on average, were better-off to earlier generations that, on average, were worse-off. Because "retirement" was a rather rough measure of "need," of course, some individuals who were not particularly needy benefited too. In general, though, the belief then was that the older generations were needier than those that would follow them.
The second progressive feature of the original program was a benefit formula that provided a higher rate of return on the contributions of workers with low wages than for those with high wages. The formula in effect during the 1940s, for example, paid a basic monthly benefit equal to 40 percent of the first $50 of a worker's average monthly wages covered by Social Security taxation, plus 10 percent of the next $200.8 Under this formula, someone who had paid Social Security taxes on wages averaging $50 per month would receive a basic monthly benefit of $20. Someone who had contributed twice as much would receive only $5 more in monthly benefits. Only when the system had matured and people had contributed over their entire careers would the system in fact redistribute from richer to poorer within generations.
The Committee on Economic Security of 1935 made clear these needs-based progressivity goals by stressing that poverty among the elderly was the primary problem to be addressed.
[I]t should not be overlooked that old-age annuities are designed to prevent destitution and dependency. (Report of the Committee on Economic Security: 33) [Italics here and in subsequent quotations are added.]
The Committee noted that a large portion of the elderly were forced to depend on the uncertain goodwill of relatives, charitable and community organizations, and state and local governments—and that this often degrading dependency led to intergenerational conflict and tension.9 As a growing portion of the population survived to old age, moreover, these traditional means of support would be increasingly strained. In the context of the Great Depression—which robbed people of not only their jobs but also of their life savings—these problems were particularly immediate.
At least one-third of all our people, upon reaching old age, are dependent upon others for support ... For those now old, insecurity is doubly tragic, because they are beyond the productive period. Old age ... is a misfortune only if there is insufficient income to provide for the remaining years of life. With a rapidly increasing number and percentage of the aged, and the impairment and loss of savings, this country faces, in the next decades, an even greater old-age security problem than that with which it is already confronted. (Report of the Committee on Economic Security: 2)
Social Security's founders also argued that America's evolution from an agrarian society of small, close-knit communities to a modern, industrialized, and highly mobile society had made it more likely that elderly persons would become destitute. Workers in increasingly strenuous industrial jobs would have great physical difficulty working past a certain age, while traditional means of support were becoming less reliable. President Franklin Roosevelt expressed these sentiments on numerous occasions. We quote two:
Security was attained in the earlier days through the interdependence of members of families upon each other and of the families within a small community upon each other. The complexities of great communities and of organized industry make less real these simple means of security. Therefore, we are compelled to employ the active interest of the Nation as a whole through government in order to encourage a greater security for each individual who composes it. (Franklin D. Roosevelt, Message to Congress, June 8, 1934) The civilization of the past hundered years, with its startling industrial changes, has tended more and more to make life insecure. Young people have come to wonder what would be their lot when they came to old age. (Franklin D. Roosevelt, Presidential Statement Signing the Social Security Act, August 14, 1935)
In response to these perceived needs, the architects of the Social Security system sought to provide participants with retirement benefits adequate to secure a basic subsistence standard of living. "Social adequacy" (in other words, a minimum standard of living), however, could not be financed entirely by the contributions of low-income individuals. Nor would anyone retiring during the first few decades of the program have contributed for enough years to finance even a modest pension. Therefore, the system was crafted not only to give a higher return on the contributions of low-income workers, but also to transfer large amounts to the first generations of retirees who had, by definition, contributed little.
In a radio address to the nation discussing the purposes of the Social Security Act, President Roosevelt made the goal of a guaranteed minimum living standard clear:
This act does not offer anyone, either individually or collectively, an easy life—nor was it ever intended to do so. None of the sums of money paid out to individuals in assistance or insurance will spell anything approaching abundance. But they will furnish that minimum necessary to keep a foothold; and that is the kind of protection Americans want. (Franklin D. Roosevelt, Radio Address on the Third Anniversary of the Social Security Act, August 15, 1938)
The recommendations of the 1937-38 Advisory Council—responsible for many program features that remain today—typify the many statements emphasizing need as a reason for program benefits:
[T]he council has become increasingly impressed by the need to revise the existing old-age insurance program in the direction of fitting the structure of benefits more closely to the basic needs of our people, now and in the future .... Since it is the purpose of old-age insurance to prevent dependency in old age, the benefits payable under the program should, as soon as possible, be sufficient in amount to afford the aged recipient at least a minimum subsistence income. (Final Report of the 1937-38 Advisory Council on Social Security: 12-13)
Ensuring a Fair Return on Contributions (Individual Equity)
If the architects of Social Security had only a single goal—to alleviate poverty and dependency among the elderly—they could have stopped with the creation of a welfare program of Old Age Assistance. Instead, the United States followed the lead of several European nations and created a mandatory public annuity program designed to eventually cover the vast majority of workers. All participants who contributed to the system for a modest number of years and reached retirement would receive benefits.
The original architects likened the system to private pensions and annuities, under which benefits are clearly related to contributions.10 This analogy implies adherence to the principle of individual equity, which is close to the philosophical concept of "commutative justice." Individual equity means that parties to any exchange should each receive a fair return from the transaction—in other words "get their money's worth." The private market achieves this goal through voluntary transactions. Individuals engage in private transactions when they believe it is worth their while to do so. An exchange is deemed unfair or unjust when one party ends up losing because the other does not honor the terms of a contract, presents misleading information, or uses coercion.
In an annuity or pension system—which is basically a system that insures against the economic risks of old age—individual equity is measured by whether individuals receive an "actuarially fair" return on their contributions. Roughly, this means that, when each year of contributions and benefits is weighted by the probability that someone will be alive during that year and adjusted for accumulated interest, total expected lifetime benefits will equal total expected contributions.11 Individuals should be paid retirement benefits that, on average, equal their contributions plus a fair interest rate. Because an insurance system pools the risks of large numbers of people into an average risk, those who die young will receive less than an equal share of benefits, and those who live longer than average will receive more.
Applying the individual equity principle to government tax and expenditure policy is somewhat more controversial and complicated than its application to private programs, but it has a long tradition, dating back to at least Adam Smith.12 In public finance theory, this concept is related closely to the "benefit principle." Taxes and fees may be regarded as the price society pays for publicly provided goods and services. The benefit principle holds that the price government charges each individual for public goods should correspond to the amount of these goods he or she consumes; in other words, people should be taxed according to the benefits they receive from the government.
This formulation involves some obvious complications. First, how can one determine how much of a government-provided good or service each individual consumes, or how to value it in the absence of a price system? Fees for parkland use present an easy case. National defense, the criminal justice system, and environmental protection, in contrast, exemplify public programs or goods whose benefits cannot be allocated easily to individuals. The problem is somewhat reduced in the case of Social Security since the cash benefits going to each participant can be measured.
A second problem, which is extremely relevant to Social Security, is that strictly applying the benefit principle makes any redistribution by government to reduce inequality or ameliorate need impossible. Some public finance theorists have dealt with this issue by dividing government functions into separate "distributive" and "allocative" branches. The first redistributes income and wealth to achieve whatever degree of "distributive justice" (i.e. fairness) society desires. The second tries to satisfy public desires not met by the private market acting alone. To the extent possible, the allocative branch should be guided by the benefit principle and allow private choices to help determine which goods and services get produced and consumed. (Musgrave 1959).
The architects of Social Security combined the allocative and distributive functions of government into a single program. To satisfy a perceived public demand for a safe, secure, and universally available retirement savings vehicle, the government established a mandatory annuity program that would pay benefits related to contributions (more precisely, to earnings on which taxes have been paid). This annuity was woven together with redistribution into a single benefit formula designed as a compromise between progressivity and individual equity.
Another obvious difference between Social Security and the private insurance and pension model is that the program was designed to pay out benefits before adequate "contributions" had been built up. Essentially, benefits of current retirees were to be paid for partially or wholly out of the contributions of current workers ("pay-as-you-go"). Although a partial reserve was set up by investing surplus Social Security taxes in government bonds, these reserves couldn't begin to cover accrued liabilities. Moreover, such reserves often amounted to little more than IOUs that would have to be redeemed by future taxpayers. When workers eventually retired, their benefits, in turn, would depend on the ability and willingness of the next generations of workers to pay Social Security and other taxes.13
This was a clear divergence from the private annuity model, but it did not by itself keep the system from meeting individual equity standards. In theory, so long as each successive generation of workers paid enough taxes, participants could receive fair returns on their contributions. (In the 1930s many private pensions were not backed up by funds or savings either.14)
The founders of the Social Security system repeatedly emphasized the contributory and annuity aspects of the program, at least for the employee's share of the taxes. For example:
Contributions by the employees represent a self-respecting method through which workers make their own provision for old age. (Report of the Committee on Economic Security 1935: 33)
Social Security's founders also envisioned a strong individual equity component to the program so participants would "get their money's worth" out of the program:
[M]embers shall not receive less than the actuarial equivalent of their own contribution. (Report of the Committee on Economic Security of 1935: 31) [W]orkers who enter the system after the maximum contribution rate has become effective will receive annuities which have been paid for entirely by their own contributions and the matching contributions of their employers. (Report of the Committee on Economic Security of 1935: 31)
This promise could not be kept for everyone since the plan was also redistributive and since worker-to-retiree ratios would fall significantly over time. The conflict between the two principles was sidestepped during Social Security's first few decades through intergenerational redistribution and through further legislated increases in benefits that were not tied to past contributions. As it turned out, the system went well beyond providing a fair return on the contributions of the first few generations of participants. No participants—not even the highest earners—were required during Social Security's youth and adolescence to make contributions that came close to covering the value of benefits received.
But this situation could not persist forever. The proportion of the population receiving benefits would grow dramatically, and those who had contributed to the system throughout their careers would finally begin to retire. As a result, the system would have to impose ever-increasing taxes (a) to provide everyone with an actuarially fair return on individual contributions, and (b) to supplement the benefits of low-income persons. Eventually these tax rates would reach a political limit and some tradeoff between progressivity and individual equity would have to occur within each generation. This, indeed, is exactly what is starting to happen now.
The conflict between progressivity and individual equity was recognized to a degree by the 1935 and 1937-38 Councils. At times they argued in individual equity terms only with respect to the employee contribution, leaving the employer contribution available for redistributive purposes (an argument economists don't buy since the employer contribution is paid by employees through lower cash wages).
At other times the two Councils recognized the conflict by proposing to support the system in the future with contributions or interest payments from government revenues. This idea would have preserved the letter but not the spirit of an actuarially fair return on contributions. It would have allowed later working populations to pay lower Social Security payroll tax rates, which in turn would provide them an actuarially fair return on their Social Security payroll taxes—but not on all the taxes they would have to pay to support the system. In the end, the President and Congress decided to forgo general government financing and rely exclusively on payroll taxes.
Equal Treatment of Equals (Horizontal Equity)
The principle of equal treatment of equals pervades government policy, though it is so intuitively obvious that it is often left implicit. If you and I commit equal crimes and are equal in all other circumstances, then we should face the same penalty. If you and I have equal income and are equal in all other respects, then we should pay the same income tax. If our homes in the same community are of the same value, and if there are no other economic differences between us, we should pay the same amount of property tax. To the extent that there is redistribution toward the needy, those with equal needs and in equal circumstances ought to receive equal assistance. And so forth.
Anyone who would deny the importance of this principle should be reminded that the stimulus for many reforms in tax and expenditure policy, civil rights, and other areas has been the outcry when horizontal equity is violated—the sense that equals were not being treated equally under the law.
One beauty of the horizontal equity principle is its lack of conflict with other principles. It is the only one, for example, that never conflicts with progressivity. You and I may disagree on how progressive we would like government to be, but we can still agree that, whatever the level of progressivity, two persons in equal circumstances should be treated equally under the law.
Another beauty is that horizontal equity promotes economic efficiency (discussed immediately below). A tax policy that tends to favor one type of income or expenditure over another, for example, not only causes people with equal economic resources to pay different amounts of tax. It also creates an incentive to engage more in the favored activity and less in the taxed activity than economic efficiency would dictate.
Horizontal equity is applied in the Social Security system through such features as equal assessment of payroll taxes on those with equal earnings (up to the earnings maximum), and payment of equal benefits to those born in the same year, with equal earnings histories, and of the same family type—admittedly narrow definitions of equals.
Achieving Maximum Benefit to Society from Available Resources (Economic Efficiency)
Economic efficiency in its purest version requires that everyone who could have been made "better off" without anyone else being made "worse off" should be. More generally, it means society ought to get the most out of its limited resources. One of the most fundamental precepts of economic theory is that if consumers and firms are left to pursue their own self-interest and make exchanges as they choose, the operation of price and incentive systems in a free market will tend to lean towards an economically efficient outcome.
In some cases, however, the decisions of individuals acting separately in the private market achieve outcomes that are less efficient than what these same individuals acting together as a society could achieve. In such cases, government intervention may be appropriate. Government may also intervene in the market to achieve other socially desirable goals, such as progressive redistribution to the needy. Any such government effort should, of course, minimize any unintentional losses of efficiency. For example, very high marginal tax rates should be avoided when possible. Moreover, programs that offer recipients choices are often more efficient than those that provide benefits in-kind.
"Efficiency" concerns can be found in the early Social Security literature on incentives to work. The effect of both tax rates and Social Security benefits on the efficiency of labor markets was emphasized mainly through arguments that tax rates should be kept low to prevent detrimental effects on the economy. The tying of benefits to covered earnings was also regarded as a way to preserve economic incentives—by offsetting the disincentive of additional payroll taxes with the prospect of additional benefits.
The council believes that such a method of encouragement of self-help and self-reliance in securing protection in old age is essentially in harmony with individual incentive within a democratic society ... It is only through the encouragement of individual incentive, through the principle of paying benefits in relation to past wages and employment, that a sound and lasting basis for security can be afforded. (Advisory Council on Social Security of 1937-38: 11)
As the system evolved, other aspects of labor market efficiency were occasionally addressed. The financial penalties of working past normal retirement age, for instance, have been gradually reduced, partly because they curb the labor supply of older workers.
Opportunity Cost
The concept of "opportunity cost," a favorite of economists, is a useful way of comparing alternative uses of resources. For every dollar the government spends providing public annuities to the elderly, for example, there is one less dollar to be spent on other goals or left in the taxpayer's pocket. If government is to be progressive (aiming help where the need is greatest) and efficient (spending each dollar where it will produce the best result), it must be able to measure and make tradeoffs among competing programs and priorities. And these tradeoffs must be made on a continuing basis if programs are to adapt to inevitable demographic and societal changes.
Social Security's founders recognized the concept of opportunity costs when they cautioned that aid to the elderly, though important, should not use up so many resources that other societal needs are neglected.
The council is also aware of the great financial costs, particularly in the future, involved in an insurance program. The pattern cannot be larger than the cloth; the degree of security afforded must be limited by the national income and the proportion of that income properly available for any specific purpose. Old-age insurance is only one element in the whole structure of governmental social services. The protection of the aged must not be at the expense of adequate protection of dependent children, the sick, the disabled, or the unemployed; or at the cost of impairing such essential services as education and public health or of lowering of the standard of living of the working population....(Final Report of the 1937-38 Advisory Council on Social Security: 11-12)
The need to weigh aid to the elderly against other social goals, in fact, played a role in the decision to finance Social Security through a separate, earmarked payroll tax:
The council believes that the contributory insurance method safeguards not only the wage earner but the public as well. By this method benefits have a reasonable relation to wages previously earned, and costs may be kept in control relative to tax collections. Through careful planning, the continued payment of benefits can be assured without undue diversion of funds needed for other governmental services. (Final Report of the 1937-38 Advisory Council on Social Security: 11)
The founders also clearly intended to leave future generations with some choice over what share of national income they would devote to Social Security as opposed to other needs.
No benefits should be promised or implied which cannot be safely financed not only in the early years of the program but when workers now young will be old. (Final Report of the 1937-38 Advisory Council on Social Security: 26)
THE OVERARCHING CONCEPT OF SOCIAL INSURANCE
As the designers and advocates of the Social Security system struggled with compromises among conflicting principles—particularly between progressive redistribution on the one hand, and individual equity and economic efficiency on the other—they sometimes tried to justify the entire system through a broader conception of individual equity and efficiency known as "social insurance." This concept also enabled them to move beyond a welfare-like system and to downplay explicit redistribution.
The "social insurance" argument is both ambiguous and revealing. Insurance of any kind by definition involves some after-the-fact redistribution from those who do not experience the insured event to those who do. Redistributive government policies could and did come to be defined as insurance against a wide variety of risks. If the insured event is "an impoverished old age" or "loss of earnings," society's transfers to the poor or retired elderly become forms of insurance payments. Since any individual's lifetime income is subject to some uncertainty and risk, redistribution can be regarded as insurance against doing poorly in the lifetime-income distribution sweepstakes.15
People who were completely ignorant of the economic circumstances they would face over a lifetime would probably view the "risk reduction" provided by social insurance as a valuable commodity indeed.16 In a private market for insurance, however, premiums must be sufficient to cover the risks for which insurance payments are made. Often, people are more informed than insurers about conditions that make their risk different from the average risk. If the insurer assesses premiums according to average risk, those knowing their risks are low may well decide not to buy insurance. If only high-risk persons are left in the insurance pool, premiums must be raised to pay for the protection. This is the problem of adverse selection, which bedevils the current health insurance market when the healthy opt out of insurance coverage or band together to buy cheaper plans.
Now let's take the case of future impoverishment. Normally, by the time people have enough money to pay for such insurance, they already have a pretty good idea what some of their social risks are. Those who consider poverty an unlikely fate will not purchase insurance against it. For society as a whole to provide all its members insurance against that risk, government must do the job.
Franklin Roosevelt, in particular, used this social insurance emphasis to sell the program, framing the problems and needs of the elderly as "hazards" against which prudent citizens should be insured. This was a particularly persuasive argument in the wake of the Great Depression.
People want ... some safeguard against misfortunes which cannot be wholly eliminated in this man-made world of ours. (Franklin D. Roosevelt, Message to Congress, June 8, 1934) We can never insure one hundred percent of the population against one hundred percent of the hazards and vicissitudes of life, but we have tried to frame a law which will give some measure of protection to the average citizen and to his family against the loss of a job and against poverty-ridden old age. (Franklin D. Roosevelt, Presidential Statement Signing the Social Security Act, August 14, 1935)
There is one big difficulty with the social insurance argument, as developed so far, when applied to the Social Security program. It justifies benefits only to those who have had bad luck. Why should social insurance include annuity payments to all retired participants, regardless of need?
RATIONALES FOR A BROAD PUBLIC ANNUITY PROGRAM
Even those who would abolish the current Social Security system typically agree that the government should provide some assistance to elderly persons who are truly needy.17 Their criticism focuses on the mandatory public annuity aspect of Social Security. Such critics argue that the system requires all workers to devote a specific portion of their income to the purchase of a government-sponsored annuity and denies them the choice of consuming with that money or investing it privately—thus limiting individual choice and reducing economic efficiency. If individual equity is the goal, why doesn't the government simply provide assistance targeted at the needy elderly and leave pensions and annuities for the rest of the population to private markets?
Undeniably, Social Security sacrifices consumer choice and individual freedom; any government program does. The real question is whether there are persuasive reasons for such a tradeoff. Building an annuity element into the original program design has three possible justifications: the desire to stimulate popular support, the need to ensure government efficiency when private markets fail, and the requirement to avoid the efficiency and equity problems inherent in a program targeted exclusively to the poor. Once a pay-as-you-go program is in full operation there is a fourth justification. It is impossible to change from a public annuity program to private insurance without running into major equity problems.
Need for Popular Support
Although the fundamental purpose of Social Security was to prevent destitution among the elderly, its architects also wanted the program to receive enthusiastic and sustained popular support. They believed that a universal annuity component was necessary to gain public support for the redistributive efforts.18 By making it easier to think of Social Security as "insurance" rather than a "handout"—social insurance rather than welfare—they made the redistributional component of the program more palatable to the American public.19
These rationales are expressed clearly in the words of Wilbur J. Cohen, a former Secretary of Health, Education, and Welfare who played an important role in the evolution of the Social Security system for almost five decades:
... I am convinced that, in the United States, a program that deals only with the poor will end up being a poor program. There is every evidence that this is true. Ever since the Elizabethan Poor Law of 1601, programs only for the poor have been lousy, no good, poor programs. And a program that is only for the poor—one that has nothing in it for the middle income and the upper income—is, in the long run, a program the American public won't support. This is why I think one must try to find a way to link the interests of all classes in these programs. (Cohen and Friedman 1972: 55)
The contributory annuity aspect of the program was also intended to help ensure its long-term political survival by giving participants a sense of "entitlement" to their promised benefits. Some benefit would be available to all workers as an "earned right" that was related to working and to paying Social Security taxes. Franklin Roosevelt, in an oft-quoted response to a close advisor, summarized the argument:
... I guess you're right on the economics, but those taxes were never a problem of economics. They are politics all the way through. We put those payroll contributions there so as to give the contributors a legal, moral, and political right to collect their pensions... With those taxes in there, no damn politician can ever scrap my Social Security program. (Schlesinger 1958: 308-309)
Many of Social Security's advocates also felt very strongly that a contributory annuity approach was necessary to preserve the "dignity" and "self-respect" of recipients.20
A final political rationale behind the annuity concept was that it attracted the support of taxpayers who would have to pay for the transfers required in the early years of the program. As noted, because large numbers of elderly in the 1930s had standards of living below what society deemed acceptable even at that time, one important goal was to pay at least subsistence-level benefits to those retiring in the near future. Such benefits required substantial subsidies from younger generations of workers through their tax contributions.21 Promising a quid pro quo in the form of universal annuities for all taxpayers made these transfers much easier to achieve politically. As it would turn out, almost everyone of voting age at time of enactment received more than a fair return—benefits in excess of what their taxes could possibly cover.
Social Security's opponents have denounced this strategy as fooling the community into voting for a program it would have been against had it known the full redistributive truth.22 Whatever one may think of the ethics of the strategy, it certainly worked. According to public opinion polls, Social Security remains immensely popular among all age groups.23 This popularity, of course, is about to be tested in a period when many will pay in more than they receive.
Correcting for Market Failure
A second rationale for a universal public annuity program was the idea that the private market was failing to accommodate the needs not only of the poor but also of the middle class. Thus, government annuities might be regarded as an efficient response to a private market failure.24 Put another way, insurance is often desirable to deal with a variety of risks, not just those related to poverty. Where private insurance markets are believed to provide inadequate protection, as in the case of insurance of bank deposits, government insurance may step in on a more universal basis.25
Prior to the Great Depression of 1929, a fairly large and robust life insurance market had developed in the United States, and private retirement annuities were available though not common (Weaver 1983).26 Some companies were also beginning to institute employer-sponsored pension plans. By 1930, about 15 percent of all workers were covered by such plans. These plans were not very safe or reliable, however. In particular, employees often lost benefit eligibility when they changed jobs and, since funds were rarely set aside separately for the plans, declining profits—not to mention bankruptcy—left many employees with no pensions and no recourse.27
Even so, the American public was reluctant to follow the lead of other industrial countries in backing a government-sponsored social security system. This mood changed, of course, when bank failures and widespread loss of private savings accompanying the Great Depression made people throughout the income distribution feel economically insecure. Providing a universal public annuity system that would be secure and resistant to economic fluctuations thus became a powerful rationale for a government-guaranteed (and mandated) method of "saving" for retirement.
Another argument held that individuals might want to commit themselves to saving for retirement to avoid irrational but tempting behavior. To the extent that people believe that they would fail to save sufficiently for retirement on their own, they may feel that a government-forced saving program is in their own best interests. "Ulysses paternalism" is the inspired name Alan Blinder applies to this rationale, named after "the ancient mariner of mythology who tied himself to the mast so as not to be lured by the Sirens' song."28 Ulysses was in a somewhat different circumstance, though. Being tied to the mast was his own choice.
Yet another argument is that most Americans seem to want a social security system more because of what it does for their parents than for what it does for themselves. No private system can help most of us provide adequately for our parents without incurring some risks associated with our own financial and employment situations. The government can. It can also adjust our premiums according to how much we earn, and at the same time mute intergenerational resentment between dependents and providers within a family.
Once individuals value Social Security as a system for dealing with the needs of their parents, however, the individual equity case for providing a quid pro quo for the contributions of current workers is weakened. The problem is that a quid pro quo has already been obtained. Paying premiums that are used fully to insure your parents hardly allows you to decide that those premiums also entitle you to insurance. That is double counting. Your children are free to decide in their own way how much to provide to you, just as you were for your parents.
Avoiding the Inequities and Inefficiencies Associated with
"Means Testing"
Can benefit programs be designed for the poor alone without running into equity and incentive problems for those higher up in the income distribution? In practice, this is impossible.
Suppose that a person saves for retirement and that these savings are invested and earn a positive return. The earnings on this saving are already subject to normal income taxation. In a means-tested system, such earnings would also reduce Social Security benefits. Together the various benefit reductions and taxes may significantly reduce the net earnings obtained from saving. The only way to scale down this disincentive to save is to pare benefits for the truly poor or cut back the rate at which benefits are reduced as earnings increase under the means test. The latter extends the benefits to higher income groups and by definition makes the program more universal. Social Security was, in fact, designed originally so that retirement benefits would be unaffected by savings.29
The disincentive to save is not the only problem with means testing. The other is the temptation for those earning above the benefit cutoff point to change their behavior to end up qualifying for benefits. When contributions are made over a lifetime, but assessment of need is made at retirement, it is easy to game the system. As middle- or high-income earners approach retirement, they can simply spend all their wealth or give it to their children (as the elderly often do today to become eligible for means-tested Medicaid nursing home benefits).
The horizontal equity violations in a poor-only program are, thus, severe. Suppose taxpayers A, B, and C have equal lifetime incomes, but only A and B save for retirement, and B gives his savings to his children. Taxpayer A is rewarded for her prudent behavior by being forced to transfer money to both B and C. This is patently unfair. Efficiency violations are severe also. By saving more, the taxpayer decreases the transfer that she will receive at retirement; by working more, she pays more taxes to support others but not to provide for her own retirement.
If, in contrast, everyone at all income levels is forced to contribute in their pre-retirement years, all will have borne part of the cost of their own retirement. Analogous arguments are often made in favor of requiring every car driver to buy automobile insurance or everyone to get health insurance. The case here is one of horizontal equity and efficiency, not progressivity. By reducing the number who could have paid for their own retirement but did not—economists call them free riders—mandated insurance at all income levels can reduce the net amount of redistribution and the amount of net taxes (in excess of benefits) needed for redistribution.30 Surprisingly, then, universality sometimes brings a system closer to the efficiency goal than can means-testing, even though means-testing results in a smaller program.
The Problem of Dissolving a Pay-as-you-go System
The final rationale for a universal system deals with changing a program that is already firmly established and covers most of the population. Once a mandatory government annuity system is set up on a pay-as-you-go basis, privatizing the entire annuity component is virtually impossible without violating some of the principles discussed here.31 Recall, in particular, that the Social Security system paid large transfers to the first generations of retirees. Given this payout, if promises of individual equity were to be honored, later generations would have to transfer income to the generation that paid for the early transfers—and the process would have to continue forever.32
As the system now stands, several trillion dollars worth of liabilities to current and future retirees have been incurred for which funds have not been set aside. Past contributions have already been spent; promises based on those contributions can be met only by taxing current and future workers. Since current retirees' benefits are paid for by current workers, converting contributions into private investment alternatives would require either withdrawing benefit promises already made to current retirees (and for which contributions had already been made) or requiring the current generation of workers to pay twice for only one benefit—that is, to pay for both themselves and previous generations of workers whose benefits were not funded. Either option would violate some notions of both horizontal and individual equity.
Other Rationales
In addition to the rationales that we have discussed, all of which have some merit, there were a few that were either based on economic propositions that were plain wrong or that are less relevant today than they were in the 1930s.
The most conspicuous incorrect proposition was that there were only so many jobs in the economy. Encouraging the retirement of older (or "superannuated") individuals was believed, commonly and incorrectly, to open up jobs to younger individuals. For example, Senator Robert Wagner, in the debates surrounding the bills that would become the original Social Security Act of 1935, argued that "The incentive to the retirement of superannuated workers will improve efficiency standards, will make new places for the strong and eager, and will increase the productivity of the young by removing from their shoulders the uneven burden of caring for the old."33 Although not everyone who helped design the Social Security system held this view, a rather strict "earnings test" was established as a strong financial incentive to retire.34 (This was finally removed in 2000 for those aged 65 or older). Inadequate recognition was given to the simple economic proposition that an older worker both earns income and produces output. The spending of that income creates demand for additional output to be produced by other workers. When older workers produce output, they also provide a bigger economic pie to be shared.
One concern that is less relevant today is the need to design a Social Security system to deal with increasing insecurities created by industrial changes and the inability of older workers to perform strenuous physical labor. Since Franklin Roosevelt used this argument, the economy has moved from an industrial age to a post-industrial technological and information age; in the process, many manufacturing jobs for men have been replaced by less strenuous service sector jobs for both men and women. If President Roosevelt's argument justified earlier retirement in the 1930s, it justifies later retirement today.
IMPLICATIONS FOR SOCIAL SECURITY REFORM
Reform of the Social Security system is inevitable, both because the system is financially imbalanced and because, as a contributory system based partly on need, it must adjust to the evolving resources and needs of the population as we move into the 21st century.
Even so, the four major principles of progressivity, individual equity, horizontal equity, and economic efficiency, are valuable in sorting through and ranking proposals for reform. We can also learn valuable lessons from the way these principles were applied and compromises made along the way. Using these principles as a lens allows us to sort reform proposals into three major categories. "Right" ways increase program consistency with principles balanced or weighted in coherent ways. "Wrong" ways reduce the consistency of the program with at least one principle without increasing consistency with any of the others. A middle group cannot be unambiguously categorized as better or worse than the current system in terms of the principles laid out here.
Even within "right" ways individuals will still rank proposals differently depending upon the weights they give different principles—in particular, the value they place on overall progressivity. This does not mean that there is no meeting ground. No reform that both makes the system regressive and fails to provide any return on contributions, for instance, can be viewed as a legitimate compromise between progressivity and individual equity. Similarly, while in practice the equal treatment of equals requires some judgment as to whether equals are measured by current income, lifetime income, wealth, or a combination, a system that ignores horizontal equity has little merit. As an exaggerated example, to determine benefits simply on the basis of race is "wrong" because it violates the principle of horizontal equity and cannot be justified as consistent with any other principle. Many features of both the current system and proposals to change it fail to meet these criteria in similar ways.
Notes
1. See Derthick (1979) for a detailed discussion of the political dimensions of Social Security's development.
2. See for example, chapter 11 in Musgrave and Musgrave (1984). In point of fact, these principles usually have been applied more rigorously in tax policy than in expenditure policy analysis. Because taxes and expenditures are integrated so closely within Social Security, we apply the principles consistently across both tax and expenditure functions.
3. These two reports, as well as the speeches by President Roosevelt quoted here, were compiled recently in a single volume, The Report of the Committee on Economic Security of 1935 and Other Basic Documents Relating to the Development of the Social Security Act—50th Anniversary Edition. (1985). When referring to these reports and speeches in the text, however, we will cite the original sources and page numbers, which are preserved in the 50th Anniversary publication.
4. Some lump sum benefits were paid out as early as 1937.
5. Robert M. Ball, who served as Commissioner of Social Security from 1962 to 1973, has noted that "In a very real sense, the 1939 Act more than the old-age benefit provisions of 1935 formed the Social Security structure we know today." (Report of the Committee on Economic Security of 1935..., 1985, p.162)
6. Seminal discussions of the development of and fundamental goals and rationales for the Social Security system are included in Ball (1978), Myers (1993), Altmeyer (1966), Cohen (1983), and the reports of various Advisory Councils on Social Security (for example: 1939, 1949, and 1964).
7. For a detailed discussion of equity and efficiency principles in government policy, see Musgrave (1959 and 1990), Musgrave and Musgrave (1984), and Head (1992). Literature on the Social Security system has often focused on the tradeoff among basic principles. In particular, it has focused on the compromise between "individual equity" and "social adequacy." See, for example, chapter 2 of Myers (1993) and chapter 10 of Derthick (1979). In our discussion, we retain the commonly used "individual equity" terminology, but treat "social adequacy" as a subset of the broader principle of "progressivity."
8. U.S. Social Security Administration (1990: 4). A credit of 1 additional percent for every year with at least $200 in wages also applied.
9. For discussions of the problems associated with old age dependency prior to the establishment of the social security system, see Achenbaum (1978) and Fischer (1977).
10. Benefits were actually based upon earnings subject to tax, not the amount of tax paid. This usually resulted in a de facto positive relationship between taxes and benefits within each cohort. Between cohorts, however, using the wage base allowed greater net transfers to be made over time to the earlier generations of retirees. Those who contributed to the system at a 3 percent tax rate, for instance, would be treated the same as those who later contributed at a 6 percent tax rate.
11. Adjustments must be made for profits, costs of administration, and marketing, although this may be reflected in the interest rate. Interest relates the value of money over time and represents the "opportunity cost" of forgoing consumption today. When someone buys insurance, moreover, their decision is based on the gains they achieve because of the protection from risk, as well as the expected monetary value of the insurance. Thus, private insurance is often purchased when the expected monetary value of benefits is less than the cost. Otherwise the wages and profits of the insurance company could never be earned. A more detailed explanation of "actuarial fairness" is presented in chapter 4.
12. See Musgrave (1959), chapter 4, and Head (1992) for detailed discussions of the benefit principle and its intellectual origins.
13. As noted by Robert J. Myers (1988), "for about the first 20 years after its inception in 1935, OASI was financed on a partial reserve system, although moving toward a current cost (or pay-as-you-go) basis."
14. It wasn't until the Employee Retirement Income Security Act of 1974 (ERISA) that many of today's current funding standards were established for defined benefit plans. Private annuities purchased through insurance companies, however, met stricter insurance standards. Even today, many private defined benefit plans are not adequately funded, partly because of exceptions to ERISA standards. In the Depression, of course, many private pensions failed for the very reason that future pension liabilities were not funded.
15. A formal economic argument along these lines has been made by Varian (1980).
16. See Rawls (1971), chapter 4, for a related discussion of the "veil of ignorance" construct in theories of justice. In this type of philosophical inquiry, the citizen is asked what distribution of income might be chosen if he or she were ignorant of his or her economic circumstances.
17. These include Milton Friedman (1962) and Peter Ferrara (1980).
18. For a detailed discussion of the role of political factors in the development of Social Security, see Derthick (1979).
19. Robert Ball, Commissioner of Social Security for three presidents, expresses great concern for this difference in attitude toward welfare and social insurance. He notes the "negative attitude toward welfare ... growing out of the punitive and paternalistic poor-law tradition (Ball, 1978: 339)" ... "Public assistance is not paid because of an economic service, as wages and social insurance are (p. 342)"... "Programs designed solely for the poor do not get the same sustained interest and support as programs that serve us all. Whenever the budget is tight, it is the programs for the poor that are likely to suffer (p. 347)."
20. For evidence of this, see Derthick (1979). An additional rationale, cited commonly in Europe but less frequently in the United States, was the desire to build "solidarity" among people of different income classes in order to foster social cohesion.
21. Waiting a long period of time before paying out significant funds would also have created large budget surpluses at a time when some believed that a deficit stimulus was necessary to help the depressed economy.
22. Milton Friedman has written "in essence what this argument says is that the community can be fooled into voting for a measure that it opposes by presenting the measure in a false guise (Friedman, 1962: 185)."
23. See, for example, Yankelovich, Skelly, and White, Inc. (1985) for polling data on the popularity of the system.
24. For related discussions of "market failure" rationales behind social insurance, see Musgrave and Musgrave (1984), Meyer (1987), and Blinder (1988).
25. Social Security also works as countercyclical policy in the sense that payments continue even when taxes temporarily go down because of a recession. The 1949 Advisory Council seems to have given some attention to this issue. The same could be true, of course, for a well-funded and insured system of private pensions.
26. Weaver (1983) further discusses the relationship between this private sector development and the lack of a political consensus for Social Security, even among labor unions with higher priorities, before the Depression.
27. See Achenbaum (1986: 15).
28. See Blinder (1988: 29). Also see Elster (1979) for origins of the idea of "Ulyssees Paternalism."
29. In 1983 Congress first inserted legislation to subject benefits of high-income retirees to income taxation. Because income from savings is included in the test of whether benefits are taxable or not, one's level of private savings now has a modest effect on one's net after-tax level of benefits.
30. In effect, the net tax rate—the tax rate less the benefits derived from those taxes—can also be reduced.
31. For a more detailed discussion of the problems associated with privatization and other radical reform proposals, see Meyer (1987).
32. Note that the start-up of a private pension plan can also involve transfers among generations and increase the amount of consumption of older persons relative to younger persons.
33. U.S. Congress, Senate, Congressional Record, 74th Cong., 1st sess., 1935, 79, pt. 9, p. 9286. Quoted in Graebner (1980: 185).
34. Graebner (1980) makes a strong case that the retiring of older workers was commonly seen as a means of opening up jobs for younger workers, and that this played an important role in political support for Social Security. Robert Ball and Robert Myers, who have been instrumental in the development of the Social Security system through much of its history, contend that this idea did not play a major role in the adoption or maintenance of Social Security or the earnings test. Indeed, many of Social Security's founders and advocates were opposed to this argument, and cited other rationales for maintaining the earnings test. Nonetheless, it is clear that many people, including Congressmen, business and labor leaders, and some members of the Council on Economic Security, did subscribe to this argument. Their political support was important to the passage of the Social Security Act.
