University at Buffalo - The State University of New York

Published: February 17, 2005

Isaac Ehrlich is UB Distinguished Professor and Chair of the Department of Economics, Melvin H. Baker Professor of American Enterprise and research associate of the National Bureau of Economic Research. His recently completed paper on Social Security, coauthored with Jinyoung Kim, assistant professor of economics, will be published in the NBER's working paper series, which is widely read by economists and policy analysts.

Let's start with the basics. What is Social Security and how is it funded?
Social Security was enacted in 1935 to cover three major programs: old-age insurance, unemployment benefits and means-tested old-age assistance (Supplementary Social Insurance, or SSI, today). What we focus on now, however, is the Old Age and Survivors Insurance (OASI) part of the system, initiated in 1939.

This essentially pension program has functioned as a defined-benefits, pay-as-you-go (PAYG) system. In this setup, contributions by current generations of workers are used to pay set benefits to retired workers, but there is no full-funding provision. In 1940, when benefits started being paid out, the tax rate for this program was 1 percent for employees and employers, or a total of 2 percent of earnings, and maximum annual taxable earnings were $3,000. Annual benefits were $270. At these levels, the program was accumulating surpluses.

Since then, the program has undergone a series of amendments. The most important ones are the expansion of the OASI program to include Disability Benefits (DI) in 1956, allowance for early retirements of workers since the early 1960s, introduction of a separate tax to fund Medicare in 1966 and allowance for cost of living adjustments (COLAs) in 1972, which became automatic after 1975.

The program's financial viability was seriously shaken following the stagflation period of the 1970s. To solve the problem, taxes were increased and the normal retirement age was raised gradually from 65 to 67, beginning in 2000. As we now know, that has not solved the problem.

How does the system work now?
Social Security still operates as a PAYG, defined-benefits system. Since benefits are defined by the political system, taxes need to be adjusted periodically to assure long-term solvency. The combined rate paid by workers and employees has risen over time from 2 percent to 12.4 percent (not counting Medicare), and maximum taxable earnings rose to $90,000 in 2005.

The program has been successful in achieving its basic mission—improving the real incomes of the elderly. But especially since the 1972 reforms, this has caused a major income transfer from the younger to the older generations. High price and wage inflation in the 1970s caused significant increases in adjusted retirement benefits, while per-capita real income was stagnant.

President Bush maintains Social Security is going broke? What does he mean by that?
The Social Security Administration (SSA) estimates that the cash flow of the system will turn negative in the year 2018 and the system will be increasingly in the red throughout the next 75 years. While the system has generated annual surpluses since 1984, these have been used by the government to finance current deficits, for which the government has been issuing bonds, or IOUs. Some commentators argue that this means that the system will not go broke until the IOUs accumulated in the trust fund are exhausted, which is supposed to happen in 2042 by the latest projections. Absent change, the system will go bankrupt about that time.

Do you agree that Social Security is in a crisis situation?
Here is the problem in a nutshell: The financial viability of the PAYG system critically depends on the "worker-support ratio," i.e., the number of contributing workers per one retired beneficiary. That ratio initially was 16 to 1. It now is 3.3 to 1, and is projected to fall to 2 to 1 in about 25 years. The problem stems from long-term demographic trends that are common in other developed countries as well—falling birth rates and growing life expectancies. In the next two decades, the number of retirees will climb 70 percent, partly as a result of retiring baby boomers. The system as presently constituted cannot remain solvent in the long term.

Is this a crisis situation? You bet it is. First, the real crisis will arrive closer to 2018 than to 2042. This is because the system's trust fund reserves—government bonds—are paper commitments, not real resources. In 2042, for example, Social Security taxes may cover only 70 percent of current mandated benefits. To fully cover the benefits, the government will have to start using general tax revenues, which would necessitate significant spending cuts or issuing costly public debt. Alternatively, the SSA can cut Social Security benefits, push up the age of retirement to 70, or raise Social Security taxes. There are several problems with these solutions:

• They were tried as part of the 1984 reforms and have provided only a Band-Aid solution because of the continuing fall in the worker-support ratio.

• The PAYG is providing increasingly bad rates of return to workers. The system yields on average only about 1-2 percent return to current beneficiaries. If further reductions in net benefits occur, this means that young people entering the labor force can expect to receive negligible, or even negative, rates of return on their Social Security contributions.

• The PAYG system also has had some unintended consequences that do not serve the economy well, and actually have magnified the impending financial shortfalls. For example, the system provides an incentive for many workers to take early retirement. In 1960, when the early-retirement provision was introduced, just 10 percent of workers retired at age 62. By contrast, in 2002 56 percent retired at age 62. Studies by economists, including my own NBER working paper with Jinyoung Kim, have shown that Social Security also has contributed to declining fertility, lower private savings and even slower productivity growth.

• There is an even bigger crisis looming in the Medicare system, which is expected to go into the red around 2019. The SSA assesses the present value of future Medicare obligations at $23 trillion over the next 75 years, about twice as high as that of Social Security. Together, the costs of these programs can reach more than 20 percent of GDP. Moreover, the problem will get increasingly worse the more reform is delayed.

How is the president proposing to reform Social Security?
The idea, long advocated by many economists, is to convert the PAYG defined-benefits system, at least in part, to a fully funded defined contribution system by allowing workers to shift a portion of their Social Security taxes—up to 4 percent of wages—to personally owned saving accounts (PSAs) managed by private pension funds and accumulating in proportion to the returns on their contributions. This will offset the amount of benefits the worker would be entitled to receive from the traditional system, although by how much has not yet been made clear. Little also is known about whether the shift would entail other changes in current taxes and benefits, as the system will remain, in part, a PAYG, defined-benefits system.

A major problem that needs to be overcome to enable a partial shift to PSAs, entails large transition costs ($1-2 trillion) since current and many future beneficiaries who would not be able to shift to PSAs still would be entitled to Social Security benefits on past contributions, while at the same time new worker contributions would be partially diverted to PSAs. According to some reports, the president may propose to pay for transition costs by indexing the rise of future benefits to inflation, rather than wages. While the president is said to resist any hike in the payroll tax rate, if the proposal includes raising the maximum taxable earnings to, say, $200,000, this will, in effect, mean a tax increase. It also is not clear how other current Social Security provisions, such as disability insurance, will be handled.

Isn't privatization risky—doesn't it take money out of the trust fund and subjects people's nest egg to the volatility of the stock market?
The proposal does not amount to full privatization, but rather to personalizing Social Security since the shift does not eliminate the PAYG system and workers will even have the option of remaining in the present system. Surely, a partly privatized system entails risks for those who wish to take advantage of it, but there are many reasons to expect that the risk would be tolerable and well compensated by higher rates of return.

In the first place, financial economists have estimated that there has never been any continuous 35-year period in the history of the U.S. stock market in which market indices, such as the S&P 500, have yielded less than 6.5 percent inflation-adjusted rate of return. From 1926 to 2004, the stock market actually has provided on average an annual return of just over 10 percent.

The risk can be contained through diversification requirements allowing only for index stock funds and a combination of stocks, commercial bonds and government bonds. These are options available to workers in social security systems that have fully privatized, such as in Chile and Hong Kong, or in the government-managed Central Provident Fund in Singapore. The Chilean system has yielded a 12.7 percent real rate of return from 1981 to 1995, and 6.5 percent real return from 1996 to 2004. Despite initial skepticism and resistance, it is hard to find any workers in Chile today who wish to return to the old system. Partial privatization also has taken place in Sweden, Poland, Argentina, Bolivia and Britain, and remains popular. These personalized saving accounts work very much the same as 401K plans or the pension systems used by members of Congress and federal employees.

Not reforming the current system entails its own risks—that of reduced benefits and declining net returns on past contributions because of projected declines in contributions. Indeed, the U.S. in 1983, the U.K. in 1986 and 1994, Germany and Italy in 1992, and France in 1993 already have significantly scaled back projected benefits for both men and women.

Do you agree with the president's ideas? If not, what do you propose?
I see a definite need to move away from a PAYG defined-benefits system toward a fully funded, defined contribution system for two basic reasons:

• We need to overcome the inherent financial weakness of the present system in view of a significant upward trend in the aging of the population, which was not anticipated at the time the system was created.

• A move toward a defined contribution system that relies on private market investment options achieves this objective because it is fully funded, and thus unaffected by demographic changes. The shift will be especially important for young workers entering the labor market who otherwise will see their net pension benefits significantly eroding.

More important, however, a shift to a mandatory savings program escapes most of the adverse incentive effects generated by the present system. The defined-benefits PAYG provisions imply that there is only a weak link between what a worker contributes individually to the program and the defined benefits the worker is entitled to receive. This increases the incentive to retire early. Also, while a balanced PAYG system depends on maintaining a high rate of aggregate fertility and a productive labor force, individual workers' retirement benefits are fixed, regardless of how many children they bear and educate, or whether they have any children at all. These incentive effects lower fertility, savings and economic growth. In contrast, PSAs controlled by individual workers increase the link between contributions and accumulated benefits and do not affect the incentive to work and invest in children.

I would like to see, however, a stronger emphasis on regulating the proposed personalized system to minimize the possibility that private insurance companies can abuse it. The government needs to require pension funds to create insurance pools that alleviate the danger of failure and make their operation conditional on achieving minimum returns, based on industry standards. Moreover, the government needs to remain an insurer of last resort. I also would like to see a plan to pay for transition costs that better spreads the burden of these costs across all future generations, rather than having the present generations of workers bear the brunt of these costs.

There are those who argue that Social Security is a moral, as well as a financial issue—that the richest country in the world should be able to take care of its citizens in their final years, particularly those at the lower end of the economic scale. How do you respond?
I do not disagree with these concerns, but we need to face reality: The PAYG system has worked relatively well for past generations of retirees, but as currently structured, it is untenable. It provides benefits to the elderly at an increasing cost to younger generations who would, therefore, be denied a fair rate of return on their contributions. Social Security should continue to fulfill the other objectives of the present system, including disability insurance and the Supplementary Social Insurance program, providing means-tested, old-age assistance to workers unable to attain a sufficient level of retirement income. But these latter programs should be financed by general revenue-they are in essence welfare programs, not a retirement-benefit program. Separating these objectives would allow us to support each one more efficiently than bundling them together. In short, to save Social Security there is no escape from reforming it.