All Commentary
Thursday, September 1, 2005

Social Security and the Insurance Illusion

Social Security Is a Tax and a Stream of Welfare Payments

In 1937, shortly after Franklin Roosevelt threatened to destroy the independence of the Supreme Court by “packing” it with ideological cronies, the Court came to heel and handed down verdicts in three cases affirming that the Social Security Act was, unlike several structurally similar pieces of pre-intimidation New Deal legislation, in accord with the U.S. Constitution.1

Wilbur Cohen, a ubiquitous figure in the history of Social Security, provides a window into the administration’s state of mind on the Court’s momentous decision. At the time of the ruling Cohen was an assistant to Social Security board chairman Robert Altmeyer and recalls gliding down the Supreme Court steps that day “in a glow of ecstasy. . . . When I got back to the office I received Mr. Altmeyer’s approval to send out a memo to the staff stating that because of the decision, we could now call the old age benefits program ‘old age insurance.’ ”

Why? Because insurance sounds a lot better to voters than a tax and stream of welfare payments, which is what Social Security is. Because, as Cohen explained it, “The American public was and still is insurance-minded and opposed to welfare, the ‘dole’ and ‘handouts.’ ” The Brain Trust knew about the importance of “framing” decades before Berkeley linguist George Lakoff did.

The irony, or hypocrisy, of Cohen’s ecstatic rush to reframe is that, executive intimidation aside, the government won the Social Security Act cases by arguing that Social Security is not insurance, but just a plain old tax on wages, falling under Congress’s taxing power, and an entirely separate and unconnected welfare program, falling under the “general welfare” provision. The Act was scrupulously drafted to ensure that the tax and the government transfers would not appear to have anything to do with each other. And the program is never described therein as “insurance.” The 1960 Flemming v. Nestor decision reaffirms that paying the tax creates no entitlement to benefits.

Nevertheless, FDR pushed hard for a dedicated payroll tax specifically so it would be connected in voters’ minds to their benefits in the way the premiums are connected to insurance payments—to create the illusion of property, contract, and legal, moral, and political entitlement. As FDR infamously declared, the dedicated payroll taxes “are political 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.”2

In anticipation of a constitutional challenge, Social Security officials went out of their way to purge their informational materials of insurance language. A 1937 pamphlet, written shortly before the Supreme Court decisions, described the program accurately and with a minimum of manipulative art: “The United States Government will send checks every month to retired workers . . . under the old-age benefits plan . . . The same law that provides these benefits for you and other workers sets up certain new taxes to be paid to the United States government.”

The 1938 pamphlet, published after the decisions, shows the insurance-framing project once again in full flower: “Your [Social Security] card shows that you have an insurance account with the U.S. Government—Federal old age and survivors insurance. This is a national insurance plan for all workers in commerce and industry . . . . [T]axes are like the premium on any other kind of insurance.”3

From the program’s inception the lack of a legally binding entitlement to benefits was deliberately obscured. By framing the program as insurance, it was possible to make benefits seem earned rather than part of a socially stigmatized “dole.” Accordingly, during “fireside chats” and public speeches, Roosevelt told American workers that they have an “insurance policy” with the government, that “the insurance policy . . . is bought” with a payroll tax “premium” and is “far more favorable . . . than any policy that any private insurance company could afford to issue.”4 He told Congress that the “old-age insurance system” has created “individual accounts” for millions of workers who may be “likened to the policy holders of a private insurance company.”5

Roosevelt’s framing strategy calls to mind an old story attributed to another formidable American president, Abraham Lincoln, that goes something like this:
Abe asks, “How many legs does a cow have?”
The bemused reply: “Four.”
“If we also call its tail a ‘leg’ how many legs does it have?”
“No, calling a tail a ‘leg’ doesn’t make it a leg!”

Nor does calling Social Security “insurance” make it insurance. Nevertheless, “To challenge the insurance analogy or resist using the terms,” writes Social Security scholar Martha Derthick, “was to show oneself an enemy of the program.”6 And this continues to be true today. Jonathan Chait of The New Republic writes, “Privatizers portray Social Security as a kind of low-performing 401(k) plan. But the program was never intended as a personal retirement plan. It’s a form of social insurance, designed to spread risks throughout the population.”7 But is it really? Is Social Security a leg because we call it a leg? Let us challenge the insurance analogy and risk showing ourselves enemies of the program.

Transfer of Risk

Well, what is insurance, anyway? Insurance is a device for guaranteeing an individual against loss by transferring a risk from the insured individual to the insurer. In private insurance the agreement between the insured and the insurer is a legal contract, a “policy,” which sets out the terms and conditions of coverage. The fee paid by the insured individual is the “premium.” Reimbursement of losses incurred by the insured through the incidence of an event covered by a policy is paid from a fund constituted by the premium payments of many individuals exposed to a similar risk of loss. This fund is sometimes called the “risk pool.” Premiums are determined by actuarial principles sensitive to the probability of the occurrence of the insurable event and the likely cost of the loss should the event occur.8

Presumably, social insurance, to be worthy of the name, should function roughly like private insurance. One of the core components of private insurance is the legal entitlement to benefits based on contractual agreement, and this component is conspicuously missing from Social Security. Another key difference—the difference that transforms mere insurance into social insurance—is that the risk pool for social insurance is the general public instead of merely those individuals who voluntarily decide to hold policies and pay the associated premiums. So, for example, for the disability insurance component of Social Security, all workers pay “premiums” in the form of payroll taxes, and all receive predetermined benefits when the “insurable event” occurs—namely, when a medical condition prevents the worker from performing his job.

However, the main element of Social Security, assistance for senior citizens, differs significantly from the disability component. It is altogether baffling how the prospect of reaching a certain age, or voluntarily withdrawing from the labor force, constitutes a “risk” of loss. Achieving the age of 62 or 65 simply does not carry with it a significantly heightened risk of poverty, nor does retirement. On the contrary, old age is correlated with wealth, the average 70-year-old being rather better off than the average 25-year-old. Becoming older and retiring from the work force is not a risk to insure against, but a near-inevitability to prepare for. A loss might occur on a birthday, but a birthday is not a reimbursable loss.

In his 1910 book Social Insurance: A Program for Reform, the first systematic American work on the topic, Columbia University economics professor Henry Rogers Seager laid out his criteria for determining which events should and should not trigger coverage by social insurance.

If the need is one the wage earner clearly foresees as certain to arise, then I should be the last person to wish to relieve him of responsibility for meeting it . . . But the future needs we are considering are not of this sort. Many wage earners go through life without being the victims of industrial accidents, without serious illness, never lacking for work, and not living long enough to become superannuated. These are all risks to which wage earners are exposed, not certain needs which they can clearly foresee.9

When Seager wrote these words, life expectancy at birth was about 51 years. A 25-year-old in 1910 could expect to expire just before reaching 65, today’s age of full Social Security eligibility. It was not unreasonable, then, to consider living well past that age, “living long enough to become superannuated,” as an unforeseeable risk to which one is exposed and may wish to be insured against. Similarly, when Social Security became law in 1935, life expectancy barely exceeded the age of eligibility.10

Today, a representative 25-year-old can expect to make it to his 80th birthday.11 The need to prepare for the interim between retirement and death is now one “the wage earner clearly foresees as certain to arise,” or, at least, foresees as very likely to arise. Under these conditions, no honest proponent of social insurance should “wish to relieve him of the responsibility of meeting it.” The phenomenal rise in life expectancy over the course of the twentieth century has simply removed retirement from the category of risk, and has therefore rendered the idea of old-age insurance obsolete.

Mutuality Lacking

Additionally, the retirement component of Social Security lacks the feature of mutuality that is at the center of the notion of a risk pool. Because all retirees—Bill Gates and Warren Buffett included receive benefits regardless of their financial condition, it can’t be that the lucky “winners” are indemnifying the unlucky “losers.” Instead, the Social Security is a demographically unsustainable system of intergenerational transfers from current workers to current retirees. It is this feature of Social Security that has led it to be characterized as a “chain letter,” or a “Ponzi scheme.”

Now, if one wishes to play fast and loose with the language of insurance, one might argue that any government program that protects individuals from the suffering of poverty is a kind of insurance. But then the category has become so slack that almost anything gets in. Any means-tested anti-poverty program qualifies. Similarly, if an income-supplementing stream of government Social Security checks counts as insurance, then it is difficult to see how an income-supplementing stream of personal retirement account annuity checks fails to perform the same function. However, the entire point of the insurance frame was to cast Social Security as something unique and distinctive. “Social insurance” is supposed to differ from mere welfare. And contemporary status quo-ists like Chait deploy insurance rhetoric to distinguish Social Security from a system of personal retirement accounts.

In a 1973 debate with Wilbur Cohen, Milton Friedman argued that “social security is not in any meaningful sense an insurance program in which individual payments purchase equivalent actuarial benefits.” It is a tax plus a program of government transfers “in which all sorts of considerations other than the amount paid determine the amount received.”12

Arguing that a tail is a leg when you call it a “leg,” Cohen maintained that Social Security must be insurance because the government calls it insurance, and that, furthermore, he once wrote an article in the Encyclopedia Britannica stating that it is. In any case, Cohen argued, rhetoric has its virtues. “I believe in rhetoric,” Cohen said, “because it makes a lot of things palatable that might be unpalatable to economists.”13 But it is not only economists, Cohen goes on to clarify, to whom “a lot of things” might seem unpalatable:“Let me emphasize that the reason why [welfare] programs don’t get appropriations, don’t get support from the taxpayer, is simply that they do not appeal to the middle class, middle income person.”14

“In the United States,” Cohen argued, “a program that deals only with the poor will end up being a poor program.” This belief—that the elderly poor will eat only cat food unless American voters are under the influence of the noble lie of insurance—is the key to Social Security as we know it.

It is not well known that under Social Security over 90 percent of payroll tax revenue goes back to the income bracket from where it came.15 Welfare liberals often suppose that this largely pointless churn is necessary to preserve the illusion of an insurance—like connection between “federal insurance contributions” and Social Security benefits, under cover of which a relatively small quantity of residual redistribution to the elderly poor may fly. Social Security is a program that turns “the skies black with criss-crossing dollars,” to use William Buckley’s vivid phrase, in order to generate a sense of false entitlement and political support for welfare transfers that, so the argument goes, un-deluded voters would not otherwise support.


1. On the legal history of Social Security, see Larry DeWitt, “The 1937 Supreme Court Rulings on the Social Security Act,”U.S. Social Security Administration, 1999,
2. Quoted in Martha Derthick, Policymaking for Social Security (Washington: Brookings Institution, 1979), p. 199.
3. Quoted in Jerry R. Cates, Insuring Inequality (Ann Arbor: University of Michigan Press, 1983), pp. 32-33.
4. Franklin D. Roosevelt, “Madison Square Garden Speech,” October 31, 1936.
5. Franklin D. Roosevelt, “A Message Transmitting to the Congress a Report of the Social Security Board Recommending Certain Improvements in the Law,” January 16, 1939,
6. Derthick, p. 199.
7. Jonathan Chait,“Blocking Move,” The New Republic, March 21, 2005, p. 21.
8. For standard definitions of insurance see, for example, “Insurance,” The Columbia Encyclopedia, Sixth Edition (New York: Columbia University Press, 2004); or “Insurance,” Wikipedia: The Free Encyclopedia, (retrieved May 10, 2005).
9. Henry Rogers Seager, Social Security: A Program for Reform, (New York: Macmillan, 1910), p. 21.
10. Elizabeth Arias, “United States Life Tables, 2002,” National Vital Statistics Reports (Washington, D.C.: U.S. Department of Health and Human Services), vol. 53, no. 6, November 10, 2004.
11. Ibid.
12. Wilbur J. Cohen and Milton Friedman, Social Security: Universal or Selective? (Washington, D.C.: American Enterprise Institute, 1972), p. 26.
13. Ibid., p. 54.
14. Ibid.
15. Jeffrey Liebman, “Redistribution in the Current U.S. Social Security System,” in Martin Feldstein and Jeffrey Liebman, eds., Distributional Aspects of Social Security and Social Security Reform (Chicago: University of Chicago Press, 2002), p. 4.

  • Will Wilkinson is vice president for policy at the Niskanen Center, overseeing the Center’s research and publications. Before coming to the Niskanen Center, Mr. Wilkinson was U.S. politics correspondent for The Economist.