The anatomy of Social Security and Medicare.

• Author: Browning, Edgar K.
• Position: Essay

Social Security in the United States is the largest government expenditure program in the world, with expenditures of $823 billion in 2005. That figure includes only the cash retirement benefits paid. Social Security's companion policy that covers the elderly's medical expenses, Medicare, had expenditures of another $299 billion. Taken together, the federally financed retirement benefits amounted to $822 billion.

Despite its vast size and its effects on almost all Americans each year of their lives, the Social Security system is probably the most poorly understood government policy of all. The jargon alone is incomprehensible: we hear of unfunded liabilities, infinite and seventy-five-year time horizons, average indexed monthly earnings and primary insurance amounts, trust funds and lockboxes, wage indexing versus price indexing of benefits, bendpoints in the benefit formula, carve-outs and add-ons, replacement rates, covered and uncovered earnings, and so on. Probably no policy, however, is more important to understand because its consequences for the way we live and for how well we live are monumental.

In this article, I focus on the fundamental issues surrounding the design of a system to provide retirement benefits to the elderly. I begin by explaining how Social Security works.

Social Security Basics

The retirement benefits provided to the elderly by Social Security (and Medicare) each year are financed by taxes on workers' earnings. The benefits received by those persons now retired do not come from a fund they accumulated by paying taxes during their working years. All the taxes they paid in earlier years were spent on providing benefits to those retired then. Similarly, the taxes paid by workers today do not go into a fund to finance their own retirement; they are spent in providing retirement benefits to those retired now. Social Security is thus an income-transfer program, with income transferred each year from workers to retirees.

But, you may ask, what about the trust funds we hear about all the time? Doesn't the money go into the trust funds? It is true that there are trust funds, but they play a minor, even negligible, role in the operation of Social Security. They are temporarily playing a somewhat larger role right now than they did in the past or will in the future, but it is a minor factor even today. (In 2005, eighty-four cents of each dollar in workers' taxes were paid out immediately as retirement benefits; only sixteen cents went into the trust funds.) I discuss the trust funds later, but we can get a clearer understanding of how Social Security operates by first ignoring them.

A retirement system that finances retirees' benefits by taxing younger workers' earnings (as Social Security does) is said to be run on a pay-as-you-go (PAYGO) basis. Because no fund is being accumulated on behalf of the taxpayers, this arrangement is also sometimes called an unfunded system.

A PAYGO system bears an eerie resemblance to a Ponzi scheme (also known as a pyramid scheme), named after Charles Ponzi, who apparently first utilized this scheme to swindle investors. In 1920, Ponzi began to borrow money from investors, promising them a return of 50 percent after only forty-five days. (On an annual compound basis, this is comparable to a return of 2,500 percent!) He paid off the early investors by using the funds provided by later investors (as Social Security paid off early retirees by taxing later retirees--that is, younger workers); he did nothing with the funds to generate such fantastic returns (as Social Security does not invest the taxes paid by workers). Like most pyramid schemes, Ponzi's system collapsed, leaving the later investors with nothing because their funds had been used to pay off the earlier investors. The entire swindle lasted less than a year. Ponzi pleaded guilty to mail fraud and spent four years in jail. Since that time, pyramid schemes have been illegal.

That is, unless they are run by the federal government, as with Social Security. But differences between privately operated Ponzi schemes and publicly operated PAYGO retirement programs enable Social Security to be a viable system for providing retirement benefits. Notably, the government can force current and future workers to "invest" (by collecting taxes from them), so that retirees can always be assured of incoming funds supplied by younger workers.

Saying that a PAYGO system is a feasible way to provide retirement benefits is not the same as saying it is desirable. Let's consider how it affects people over their lifetimes. To understand how it operates in the simplest way, imagine that people live only two years, working in the first year and retired in the second. When working, they pay a tax on their earnings of 10 percent, which finances retirement benefits for those retired. What can you expect to get from such a system when you retire?

When you retire, you get the proceeds of the 10 percent tax on the earnings of those then working. If their earnings are greater than the earnings you had when you paid the 10 percent tax, you will get back more in retirement than you paid in taxes when you were a worker. All retirees can potentially get back more than they paid in taxes as long as workers' earnings are growing over time and the number of workers per retiree does not change.

This relationship suggests a link between the growth in earnings and the returns that PAYGO Social Security can generate. Indeed, more elaborate calculations show that Social Security can provide over the long term on average an implicit (annual) rate of return on taxes paid that equals the annual rate of growth in (taxable) earnings. So all generations of workers can get back more than they paid in as long as earnings rise over time. As it is sometimes expressed, Social Security allows people to "share in the growth of the economy" because economic growth and earnings growth are approximately the same thing.

What you can expect to get back (or, more precisely, what people on average can expect to get back) is therefore tied to the growth in earnings. Over the past fifty years or so, the annual growth in total real earnings has averaged about 2.5 percent. For reasons to be discussed later, future growth is expected to be somewhat lower, perhaps 1.5 to 2.0 percent per year. Therefore, the implicit real rate of return on taxes paid into Social Security will likely average around 1.5 percent or so in the foreseeable future. In the next section, I compare that return to the returns available from other ways of providing for retirement.

All of the rates of return (or interest rates) discussed in this article are real rates; in other words, they are adjusted for inflation. If the nominal (or monetary) rate of return is 10 percent, but prices are rising by 10 percent a year, then the real rate of return is zero--because the purchasing power of the $1.10 you get back one year after investing $1.00 has the same purchasing power as the $1.00 you started with. Real returns determine living standards, and therefore the growth rates in earnings I cited previously are expressed in real terms--that is, in terms of dollars of constant purchasing power.

One other matter deserves emphasis. What I have explained is that over the long haul a PAYGO system offers an implicit real rate of return equal to the rate of earnings growth. "The long haul" here refers to people who spend their entire lifetimes under the full-blown system, paying taxes throughout all their working years and then receiving retirement benefits. In a PAYGO system, some people will do much better than this long-haul return. They are persons who retire in the early years of the system's operation and who did not pay taxes for all their working years or who paid taxes when rates were lower. To see this effect, imagine that we start a PAYGO system with a 10 percent tax rate this year. People who are retired this year will receive the tax revenues as retirement benefits (remember, it is PAYGO), even though they paid no taxes at all during their working years because the system did not exist before this year. They get a really sweet deal--benefits at no cost to themselves--and clearly fare much better than young workers just starting out, who will pay taxes and receive only a return equal to the rate of growth of earnings. People who are near retirement age also do extremely well because they will pay taxes for only a small proportion of their working years.

Therefore, we expect those who retire in the early years of a PAYGO system to fare much better (receive a much higher rate of return) than those who retire later. They receive windfall gains, as did the early participants in Charles Ponzi's scheme who were paid off before the operation collapsed. In contrast to Ponzi's scheme, however, Social Security, as we have seen, can continue to pay later generations of retirees an implicit rate of return equal to the rate of earnings growth.

The windfall gains that early retirees receive are one of the major differences between a PAYGO system of providing retirement benefits and a funded system. In a funded system, what you get back is strictly linked to what you put in; if you put nothing in, you get nothing back. Early generations of retirees do not automatically get much higher returns than later generations in a funded system, as they invariably do in a PAYGO system.

Is Social Security a Good Deal?

We have seen that not all generations of retirees will receive the same rate of return on their Social Security taxes. Generations that retire earlier do better than later generations. In contrast to my earlier heuristic discussion, however, Social Security did not emerge "full blown" after its enactment in 1935. Benefits were first paid in 1940, and not all the elderly received benefits because of the specific legislative restrictions on eligibility. In fact, in 1950 only one of six people older than...