Nine Misconceptions About Social Security

To take a voguish example: Privatize Social Security? Any money saved would be eaten up in Wall Street transaction costs, including million-dollar salaries

A close look at Social Security

1. The Social Security Trust Fund is an accounting fiction.

The Social Security tax has been raising more money than is needed to pay for current benefits, in order to build up a surplus to help finance the retirement of the Baby Boom generation. All of this surplus is lent to the U.S. Treasury when the Social Security Trust Fund buys bonds from it. The money is then used to finance the federal deficit, just like any other money the government borrows. The bonds held by the fund pay the same interest as bonds held by the public. These bonds are every bit as real (or as much of a fiction) as the bonds held by banks, corporations, and individuals. Throughout U.S. history the federal government has always paid its debts. As a result, government bonds enjoy the highest credit ratings and are considered one of the safest assets in the world. Thus the fund has very real and secure assets.

It is true that the interest the government pays on these bonds is a drain on the Treasury, as will be the money paid by the government when the fund ultimately cashes in its bonds. But this drain has nothing to do with Social Security. If the Social Security Trust Fund were not currently building up a surplus, and lending the money to the government, the government would still be running a deficit of approximately $60 billion in its non-Social Security operations. It would then have to borrow this money from individuals like H. Ross Perot and Peter G. Peterson and to pay out more interest each year to the people it borrowed from. Therefore, the government's debt to the fund is simply a debt it would have incurred in any event. The government's other spending and tax policies, not Social Security, will be the cause if there is any problem in the future in paying off the bonds held by the fund.

The government bonds held by the Social Security Trust Fund will always be a comparatively small portion of the government's debt, and therefore a relatively minor burden. They will hit a peak of about 14.4 percent of gross domestic product in 2015, whereas the debt now held by individuals and corporations is about 47 percent of GDP. Therefore, at its peak the burden of interest payments to the fund will be less than a third as large as the interest burden the government now bears.

2.> The government uses overly optimistic numbers to convince people that Social Security will be there for them. The situation is much worse than the government admits.

Actually, Social Security projections are based on extremely pessimistic economic assumptions: that growth will average just 1.8 percent over the next twenty years, a lower rate than in any comparable period in U.S. history; that growth will slow even further in later years, until the rate is less than half the 2.6 percent of the past twenty years; that there will be no increase in immigration even when the economy experiences a labor shortage because of the retirement of the Baby Boom generation; and that this labor shortage will not lead to a rapid growth in wages. Both possibilities excluded in these projections—increased immigration and rapid wage growth—would increase the fund's revenues. These projections are genuinely a worst-case scenario.

3. The demographics of the Baby Boom will place an unbearable burden on the Social Security system.

Those who want to overhaul Social Security make their case with the following numbers: in 1960 there were more than five workers for each beneficiary; today there are 3.3 workers; by 2030 there will be only two workers for each beneficiary. At present the fund is running an annual surplus of more than $80 billion, approximately 20 percent as much as its current expenditures. This surplus will generate interest revenue to help support the system as the ratio of workers to beneficiaries continues to fall in the next century. Also, the fact that workers are becoming more productive year by year means that it will take fewer workers to support each retiree. The United States had 10.5 farm workers for every hundred people in 1929; it has fewer than 1.1 farm workers for every hundred people today. Yet the population is well fed, and we even export food. Rising farm productivity made this possible. Similarly, increases in worker productivity (which have been and should be reflected in higher incomes), however small compared with those of the past, will allow each retiree to be supported by an ever smaller number of workers.

In fact the demographics of the Baby Boom have very little to do with the long-range problems of Social Security. The main reason the fund will run into deficits in future years is that people are living longer. If people continue to retire at the same age but live longer, then a larger percentage of their lives will be spent in retirement. If people want to spend a larger portion of their lives in retirement, either they will have to accept lower incomes (reduced benefits) in their retirement years relative to those of their working years, or they will have to increase the portion of their incomes (higher taxes) that they put aside during their working years for retirement.

This is the main long-range problem facing Social Security. Current projections show that the annual deficit will be 5.71 percent of taxable payroll in 2070, long after the Baby Boom will have passed into history. But the annual deficit is expected to be only 4.44 percent of taxable payroll in 2035, when the worst crunch from retired Baby Boomers will be felt.

Examining just the change in the ratio of beneficiaries to workers overstates the burden that workers will face in the future. To assess the burden accurately it is necessary to examine the total number of dependents—beneficiaries and children—each worker will have to support. It is projected that this ratio will rise from 0.708 per worker at present to 0.795 in 2035. But even this number is well below the ratio of 0.946 that prevailed in 1965. And the fund's trustees project a lower birth rate, meaning that the increased costs of providing for a larger retired population will be largely offset by the reduction of expenses associated with caring for children.

4. Future generations will experience declining living standards because of the government debt and the burden created by Social Security.

Projections indicate that workers' real wages will increase by approximately one percent a year. If Social Security benefits are left unchanged, in order to meet the fund's obligations it will be necessary to raise the Social Security tax by 0.1 percent a year (0.05 percent on the employer and 0.05 percent on the employee) for thirty-six years, beginning in 2010. This will be a total tax increase of 3.6 percent, approximately the same as the increase in Social Security taxes from 1977 to 1990.

Even with this schedule of tax increases, real wages after Social Security taxes are deducted will continue to rise. By 2046, when the tax increases are fully phased in, the average wage after Social Security taxes will be more than 45 percent higher than at present. As noted earlier, this is based on pessimistic projections about wage growth.

5. By 2030 federal spending on entitlement programs for the elderly will consume all the revenue collected by the government.

By far the greatest part of the projected increases in federal spending on entitlement programs for the elderly is attributable to a projected explosion in national health-care costs, both public and private. According to projections from the Health Care Financing Administration, average health-care spending for a family of four in 2030 will be more than 80 percent of the median family's before-tax income. If such an explosion in health-care costs actually occurred, the economy would be destroyed even if we eliminated entitlement programs altogether. If health-care costs in the public and private sectors are brought under control, the problems posed by demographic trends will be quite manageable.

It is very deceptive to combine other spending categories with health care; projected health-care costs by themselves will consume most of the budget. For example, projected federal spending on education, highways, and health care combined should be more than 80 percent of federal revenues in 2030; defense spending plus projected health-care spending should come close to 70 percent of federal revenues.

6. If Social Security were privatized, it would lead to a higher national saving rate and more growth.

By itself, privatizing Social Security would not create a penny of additional savings. All the privatization plans call for the government to continue to pay Social Security benefits to current recipients and those about to retire; therefore spending would be exactly the same after privatization as it was before privatization. Yet the government would no longer be collecting Social Security taxes. Each dollar an individual put into a private retirement account rather than paying it to the government in Social Security taxes would still be a dollar the government must borrow. Individuals would be saving more, but the government would have reduced its saving (increased its borrowing) by exactly the same amount. Most of the privatization schemes being put forward call for additional taxes and additional borrowing to finance a transition while benefits were being paid out under the old system. Any additions to national savings attributable to these plans would stem entirely from the tax increase. This tax increase would have the identical effect on national savings if it were not linked to privatizing Social Security. In other words, raising taxes is one way to increase national savings, and if we are willing to raise taxes, we need not privatize Social Security.

The fact that individuals might put their savings in the stock market or in other private assets, whereas the Social Security Trust Fund buys government bonds, doesn't affect the level of saving at all. If it did, the government could increase the level of saving in the economy by borrowing money and then investing it in the stock market, or by borrowing money and giving it to individuals with the requirement that they invest it in the stock market. If either step could increase the level of saving in the economy, the government should take it independent of any changes in the Social Security system.

In fact, all else being equal, if individuals invested the money they would otherwise pay out in Social Security taxes, less saving would result, because a large portion of this money would be siphoned off by the financial industry. Currently stock brokers, insurance companies, and other financial institutions charge their customers an average of more than one percent a year on the value of the money they hold. Thus if $1,000 is invested through a brokerage firm for forty years, the investor will have been charged in excess of $400 in fees on the original investment, plus an additional one percent a year on all gains. These fees are a big cost from the standpoint of the individual investor, and a complete waste from the standpoint of the economy as a whole. Meanwhile, the operating expenses of the Social Security system are less than $8.00 for every $1,000 paid out to beneficiaries.

It is easy to see why costs in the private financial sector are so much higher. The private sector pays hundreds of thousands of insurance agents and brokers to solicit business. It also incurs enormous costs in television, radio, newspaper, and magazine advertising. In addition, many executives and brokers in the financial industry receive huge salaries. Million-dollar salaries are not uncommon, and some executives earn salaries in the tens of millions. Privatization would add these expenses, which are currently absent from the Social Security system.

7. If people invest their money themselves, they will get a higher return than if they leave it with the government.

This may be true for some people, but it cannot be true on average, for much the same reasons as noted above. Some people may end up big winners by picking the right stocks, but if the national saving rate has not increased, the economy will not have increased its growth rate, and the economic pie will be no larger in the future with privatization than it would have been without it. Thus high returns for some must come at the expense of others. In fact, since the cost of operating a retirement system is so much greater through the financial markets than through the Social Security system, the average person will actually be worse off.

Some advocates of government-mandated saving plans argue that individual investors can get real returns of seven percent on money invested in the stock market (the historical rate of return), and that this would ensure a comfortable retirement for everyone. Certainly people have gotten far better returns in the market in the past few years, but if Social Security projections are accurate, such rates of return cannot be sustained. Profits can rise only as fast as the economy grows (unless wages fall as a share of national income, which no one is projecting). If stock prices maintain a fixed relationship to profits, then stock prices will grow at the same rate as the economy. The total return will therefore approximate the ratio of dividends to the stock price (currently about three percent) plus the rate of economic growth (two percent over the next ten years, but projected to fall to 1.2 percent in the middle of the next century). This means that the returns people can expect from investing in the stock market will be five percent in the near future and 4.2 percent later in the next century. For stock prices to rise enough to maintain a real return of seven percent, price-to-earnings ratios would have to exceed 400:1 by 2070.

8. The Consumer Price Index overstates the true rate of increase in the cost of living. Social Security recipients are therefore getting a huge bonanza each year, because their checks are adjusted in accordance with the CPI.

There is considerable dispute about the accuracy of the CPI. The Boskin Commission, which was appointed by the Senate Finance Committee to examine the CPI, stated in its final report, in 1996, that overall the CPI had been overstating the cost of living by 1.3 percentage points a year. However, the Bureau of Labor Statistics found that the CPI understated the cost of living when compared with an index that measured the cost of living for the elderly. This is because the elderly spend an unusually large share of their income on health care and housing, which have risen relatively rapidly in price. Questions remain about the accuracy of the CPI, and they cannot be resolved without further research.

However, one point is clear. If the CPI has been overstating inflation, then future generations will be much better off than we imagined. If inflation has been overstated, then real wage growth must have been understated, since real wage growth is actual wage growth minus the rate of inflation. If we accept the Boskin Commission's midrange estimate of CPI overstatement, average real wages in 2030 will be more than $54,000 (measured in today's dollars). If the commission's high-end estimate is right, average wages will be nearly $65,000. By 2050 average wages will be at least $82,000 and possibly as much as $108,000.

Another implication of a CPI that overstates inflation is that people were much poorer in the recent past than is generally recognized. This conclusion is inescapable: if the rate of inflation is lower than indicated by the CPI, then real wages and living standards have been rising faster than is indicated by calculations that use the CPI. If wages and living standards have been rising faster than we thought, then past levels must have been lower. Projecting backward, the Boskin Commission's estimate of the overstatement of the CPI gives a range for the median family income in 1960 of $15,000 to $18,000 (in today's dollars)—or 95 to 110 percent of income at the current poverty level.

If the Boskin Commission's evaluation of the CPI is accepted, any assessment of generational equity looks very bad from the standpoint of the elderly: they lived most of their lives in or near poverty. And the future looks extremely bright for the young. Average annual wages in 1960, when today's seventy-three-year-olds were thirty-five, was between $10,006 and $11,902 in today's dollars. Average annual wages in 2030, when today's newborns are thirty-two, will be between $54,000 and $65,000 in today's dollars. Such numbers make it hard to justify cutting Social Security for the elderly in order to enrich future generations, on the grounds of generational equity.

9. Social Security gives tens of billions of dollars each year to senior citizens who don't need it. This money could be better used to support poor children.

Most of the elderly are not very well off. Their median household income is only about $18,000. However, even if they were better off, it would be hard to justify taking away their Social Security on either moral or economic grounds.

Social Security is a social-insurance program, not a welfare program. People pay into it during their working lives. They have a right to expect something in return, just as they expect interest payments when they buy a government bond. Social Security is already progressive: the rate of return on tax payments is much lower for the wealthy than for the poor. This progressivity is enhanced by the fact that Social Security income is taxable for middle- and high-income retirees but not for low-income retirees. If benefits for higher-income retirees were cut back further, those people would be receiving virtually no return for the taxes they paid in. This would be certain to undermine support for the program.

From an economic standpoint, means testing or any other way of denying benefits to the wealthy would be foolish, because it would give people a great incentive to hide income and thereby pass the means test. There are many ways this could be done. Parents could pass most of their assets on to their children and then continue to collect full benefits. People could move their money into assets that don't yield an annual income, such as land or some kinds of stock. Most of the income of retirees is from accumulated assets, which makes it much easier to hide than wage income. Means testing would in effect place a very high marginal tax rate on senior citizens, giving them a strong incentive to find ways to evade taxes. It may be desirable to get more revenue from the wealthy, but means testing for Social Security makes about as much sense as means testing for interest on government bonds.