Pay-As-You-Go Social Security

by Dan Cornwell

The Madison Institute
P.O.B. 5304, Madison, WI 53705-0304
December, 1997

Dan Cornwell is Professor Emeritus at the University of Wisconsin-Madison. Since retiring from the Department of Chemistry, he has devoted some of his time to the study of several national policy issues.


Some Background

As Americans re-think their Social Security system, the key decision for them to make is whether to have pay-as-you-go funding or advance funding.

In a pay-as-you-go system (approximated by Social Security today), money raised by a payroll tax on wages is paid out almost immediately to beneficiaries. The trust fund plays the role of a checking account, being only large enough to ensure a smooth cash flow and to even out fluctuations of the economy.

In advance-funded systems, the money collected from workers is diverted into financial markets, where it is held for an average of 30 years before the benefits are paid out. The major differences between the two funding methods are in the size of the trust fund, the time delay between collection and pay-out, and the predictability of benefits relative to the cost of living.

In other papers (see References) the author has made the case that pay-as-you-go funding is far superior to advance funding for a national pension program.

This Paper

The present notes are mainly excerpted from the earlier papers mentioned. The focus is on describing the steps which are needed to restore the current system to fiscal balance with complete pay-as-you-go funding, no increase in payroll tax rates, and no cut in the level of benefits.

Contrast With Other Approaches

Most others deal with the projected shortfall in Social Security funding by seeking revenues to restore fiscal balance. I see these solutions as doomed to fail because they ignore the basic causes of the problem.

Worst of all are the privatization plans. They achieve "full advance funding" by promising nothing.

The Campaign of the Privatizes

There are fortunes to be made if the nation can be persuaded to abandon pay-as-you-go and adopt an advance-funded, privatized system. Those who stand to reap these profits have waged an unrelenting campaign to destroy public confidence in the current system while touting an privatized alternative. No matter that the campaign has been marked by exaggerated claims and misleading comparisons. It has been effective.

Who Would Benefit from Privatization?

A system which mandates the flow of 10% of the nation's payroll annually into the stock and bond markets would be a bonanza for financial businesses and advertising firms. This is money which would otherwise go directly to beneficiaries, or be left in the hands of the workers who had earned it. The ceaseless churning of these funds over a 30-year average holding period would make for enormous profits for the finance industry.

The Fallacy of Advance Funding

Part of the appeal of advance funding rests on a false perception - that through a compulsory investment system each individual can provide for his or her own retirement and thereby reduce the burden on future workers. While the individual who has saved may be better off in retirement than one who has not, the same strategy cannot work on a national scale. The effect of a compulsory investment program would be to put more financial paper in the hands of all retirees at some future date. Doing this does not increase the total of goods and services available at that future time. Rather, the extra financial paper competes with wages of workers producing those goods and services. The effect is to increase the burden on these workers, not to decrease it.

Would Compulsory Investment Improve National Productivity?

What would be the effect on individuals and families of compelling workers to give up additional amounts from their wages for investment? Low- and middle-income workers, who now save little, would be the most affected. They would be forced to invest money which they might better spend for family needs such as education, child care, health care, housing, and even (at the lowest income level) for nutrition. Diverting money from these needs does not improve national productivity.

Pay-As-You-Go Social Security

The Goal

The goal of Social Security is to provide insurance to every worker and immediate family against loss of income due to disability, old age, or death. The challenge is to do this, despite the aging of the population, without unduly burdening younger workers.

To meet this challenge efficiently, the system must be simple, universal, and focused.

Two Problems

Social Security faces two problems. The first is demographic--the aging of the population and the population surge known as the baby boom. As will be demonstrated below, the solution to this problem is clear.

More serious is the political problem. Those who stand to reap enormous profits by adoption of alternative schemes have waged an unrelenting campaign to destroy public confidence in the system. This campaign has fed on and added to widespread public confusion about the operation of the current system.

In truth, the current system is efficient and fundamentally sound. With changes in some of its rules, but not in its basic structure, it can perform better than any alternative which has been proposed.


The plan described here is funded entirely on a pay-as-you-go basis, to take full advantage of the inherent merits of this funding method.

In fairness to future workers, the payroll tax rate is not increased. The plan is designed to maintain a fair and economically sustainable ratio of workers to retirees. In fairness to future retirees, the level of benefits is not cut.

How Pay-As-You-Go Works

The trust fund plays the role of a checking account, assuring a steady cash flow. Typically, the sum on deposit is about equal to one year's pay-out.

The Requirement for Fiscal Balance

The requirement for long-term fiscal balance is that, on average,

Annual pay-in = Annual pay-out


Since Social Security benefits are determined by a formula which ties them to the current national average annual wage, the key to assuring fiscal balance at a fixed payroll tax rate is to maintain a stable ratio of workers to retirees.

Social Security Today: Close To Pay-As-You-Go

Despite modest surpluses in recent years, Social Security is essentially a pay-as-you-go program. The trust fund balance at the start of 1997 was about 1.5 times the anticipated pay-out for that year.

By way of contrast, an advance-funded system would require trust fund levels to be roughly 30 times as large as one year's pay-out. These huge reserve funds would be invested in financial markets. All this is money which a pay-as-you-go system would leave in the hands of individuals.

Advantages of Pay-As-You-Go

1. Built-in inflation protection. Inflation is not a problem, because pay-in and pay-out are in the same dollars. No investment-based system can provide such accurate inflation protection, because of the long time lag between pay-in and pay-out.

2. Predictability. Since income to the system comes from a tax on wages, benefits can track the current level of wages. Since wages represent the actual cost of living at the current standard of living, it is assured that benefits can track the current cost of living.

In an advance-funded system, retirement income depends on the value of an investment portfolio at the time of retirement. The portfolio could be worth twice the value of Social Security benefits, or it could be worth nothing. With advance funding, there are no guarantees.

3. Stability of income to the system. The payroll tax assures a steady and reliable stream of income to support benefits. No investment-based system can match this stability and also track the cost of living.

4. Low administrative cost. The simplicity of the pay-as-you-go system means very low operating costs. For advance-funded plans, costs would be from 10 to 50 times as large, reducing benefits or requiring higher payroll tax rates. Privatized systems would entail additional costs for advertising, public relations, lobbying, and profits.

How Are We Doing?

Projected pay-in as a percentage of pay-out extended to 2075 is shown in the following graph:

The prediction is that, if no adjustments are made, revenues will be sufficient to cover only about 70% of benefits beyond 2030. The creek is not running dry, but something needs to be done.

Although the rapid fall in the graph between 2000 and 2030 is an effect of the baby boom retirement, the fundamental long-term problem is the continued trend beyond 2050 (by which time most of the baby boom generation will have passed from the scene). Two factors cause this long-term trend:

  • Increased life spans combined with a fixed retirement age mean a steady decline in the worker-retiree ratio.
  • There continues to be a gradual shift of compensation for work to formsnefits) not subject to the payroll tax. This erosion of the payroll tax base further lowers revenues.

The Approach to be Used

The approach outlined below is first to deal directly with the two causes of the long-term trend, then to address several issues of equity. The level of benefits is maintained. The payroll tax rate is not increased.

The proposal is designed to support an approximately constant fraction of the adult population. This fraction would be slightly smaller than it now is, still allowing full benefits to begin at an age when most people are still healthy and able to work. It would mean a life expectancy at retirement of over 15 years, increasing after 2030. By keeping the payroll tax rate constant, it avoids increasing the burden on future workers.

This proposal differs from most (perhaps all) others, which instead of dealing directly with the causes of the long-term shortfall, focus on raising funds to support an ever-increasing fraction of the adult population in retirement. Such an approach is economically unrealistic.

Outline of a Pay-As-You-Go Solution

To counter long-term trends:

  • Gradually increase the age at which full retirement benefits become available. Replace incentives to retire early with incentives to continue working. These adjustments only recognize the striking increases in longevity and improvements in health of people in their 60s which have taken place over recent decades.
  • Avoid further erosion of the payroll tax base. The value of new fringe benefits should be counted as compensation and included in wages subject to the payroll tax.

To improve equity:

  • Make Social Security pension benefits subject to income taxation on the same basis as other government and private pension income. This means including in taxable income all amounts received in excess of previously-taxed contributions by the employee.
  • Extend mandatory Social Security coverage to all newly hired state and local government employees.

Effect of the Proposed Adjustments

The graph below shows the estimated effect of adopting the above proposals. More than half of the change comes from increasing the retirement age.

The surpluses before 2020 would be far more than enough to carry the system through the slight remaining deficits of the baby boom retirement years. The long-term decline after 2030 is replaced by an upward trend.

The surpluses of around 10% in each of the first 20 years can play a significant role in facilitating reform:

  • Part of the surpluses could be used to pay the transition costs of introducing "earnings-sharing" or other systems for improving equity for couples, divorced spouses, and surviving spouses.
  • A modest reserve could be accumulated to cover unexpected costs associated with reform or to make up for errors in projections.

The total surplus is about $800 billion (1997 dollars) for the period 2000-2020.

Longevity and the Normal Retirement Age

The Normal Retirement Age (NRA) for Social Security is the earliest age at which full retirement benefits are available. It has been 65 since the beginning of Social Security. Under current law, it will rise gradually to 67 and then remain constant.

Increasing the NRA is often portrayed as a cut in benefits, but this view is misleading. It is more accurate to say that keeping the NRA fixed while longevity is rising amounts to increasing benefits. Life expectancy at the NRA will have increased from about 12 years in 1935 to 18 by 2050. Providing income to retirees for an average of 18 years as compared with 12 represents a 50% increase in the total benefit to a retiree.

Such increases are not sustainable. What the system can do, indefinitely, is to support a given fraction of the total population. This result can be achieved by gradually increasing the NRA to keep pace with longevity.

The graph below shows projected life expectancy at the current-law NRA (full columns, including the white caps). If the proposal of this paper were adopted, the NRA would gradually rise to 69 in 2030 and then be indexed to life expectancy (black columns). Life expectancy at retirement would still be more than 15 years and would continue to rise after 2030.

Don't Cut Cost-of-Living Adjustments (COLAs)

The following graph shows the effect on purchasing power of a 0.5% reduction in COLAs for someone who lives to age 95 after retiring at 65.

If it is felt to be necessary to cut benefits, a far better way is to reduce the initial benefit but then provide COLAs sufficient to maintain its purchasing power. The following graph shows the effect of using this approach to achieve the same savings for the retirement system

The CPI Has Not Overstated TRUE Changes in the Cost of Living

Does the Consumer Price Index (CPI) overstate increases in the cost of living? Consider the historical evidence.

The following graph compares the rate of growth of the national average annual wage with the CPI since 1960. Data are grouped in five-year periods to damp year-to-year fluctuations. It is clear that the CPI has lagged behind the average wage.

The average annual wage, after deduction of the small amount saved, represents the amount actually spent from wage income to purchase the products which define each year's standard of living. Thus, the average annual wage is an excellent measure of the true cost of living at each year's standard of living. Clearly, the CPI has not overstated the yearly changes in this cost.

Social Security is an income insurance program supported by a tax on current wages. Benefits should track current wages.

Earn More by Investing the Trust Fund in Stocks?

For a pay-as-you-go system, the trust fund is relatively small (see p. 3) and it doesn't matter very much how the fund is invested, as long as it is liquid and safe. To the extent that it matters, the trust fund now gets market rates for a practically risk-free investment. For 1996, the interest earned was 7.6%.

Higher returns for the trust fund would make a real difference only if the fund were increased several-fold so that the system became advance-funded to a significant extent. As emphasized elsewhere, advance funding is not a good strategy for Social Security.

Direct Comparison with Compulsory Investment Plans

Plans which mandate a significant level of investment in individual accounts generally require a payroll tax increase of about 1.5%. The pay-as-you-go plan recommended here requires an increase in the retirement age of about 2 years by the middle of the next century, but does not require a payroll tax increase. Consider the trade-off:

With the pay-as-you-go plan, the worker still has the option to invest the additional money privately to provide for earlier retirement but is not compelled to do so. Spending for other needs might well contribute more to future well-being of the family than saving for earlier retirement.


The data presented here show what changes are needed in order to make the current pay-as-you-go system work with the assurance of long-term fiscal stability.

A public pay-as-you-go system can provide income insurance benefits which are tied to current wage levels and are practically immune from inflation. Benefits are secure and predictable. The simplicity of the system leads to remarkably low administrative costs. No investment-based system can match any of these advantages.

To be fair across generations, the system must operate with an approximately constant payroll tax rate. This requires maintaining a stable ratio of workers to retirees. As life spans increase, people must expect to work longer.

Compulsory investment plans promise earlier retirement, but with no guarantees. These plans require mandatory contributions of about 1.5% of wages beyond the current payroll tax rate for building up the investment funds. The pay-as-you-go plan proposed here would leave this money in the hands of individuals, who would still have the option to invest privately for earlier retirement but would not be compelled to do so.


All calculations and graphs are based on data taken from the references listed below under "Sources for Data".

The book by Steuerle and Bakija is an excellent background reference on Social Security.

Pay-as-you-go proposals which differ considerably from the present proposal are given in the papers by Eisner, Myers, and Weisbrot. Eisner would balance Social Security by means of an increase in income tax rates or, alternatively, through use of incentives for voluntary contributions to the trust fund. The proposal of Myers is much closer to that of the present paper, although it would raise the payroll tax rate and require investment except by low-income workers. Weisbrot supports continuance of the present system with a gradual increase in payroll tax rate.

The Advisory Council report contains data for a "PAYGO" proposal which simply raises the tax rate to achieve fiscal balance.

Papers by John Mueller present economic analyses leading to the conclusion that it would be a costly mistake to end pay-as-you-go Social Security.

Sources for Data

Bipartisan Commission on Entitlement and Tax Reform, "Staff Report on Entitlement Reform Options" (Washington, DC, December, 1994).

1994-96 Advisory Council on Social Security, "Volume I: Findings and Recommendations; Volume II: Reports of the Technical Panels" (Washington, DC, January, 1997).

Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds, "1997 Annual Report" (Washington, DC, 1997).

Other References

R. V. Burkhauser, K. A. Couch, and J. W. Phillips, "Who Takes Early Social Security Benefits? The Economic and Health Characteristics of Early Beneficiaries". The Gerontologist, 36, 789-799 (1996).

Dan Cornwell, "Social Security: The Case for Pay-As-You-Go." (The Madison Institute, Madison, WI, October, 1997). This paper presents the case for pay-as-you-go and against advance funding and privatization.

Dan Cornwell, "Pay-As-You-Go: The Way to Go With Social Security" (The Madison Institute, Madison, WI, October, 1997). This paper presents the results of further computations showing the fiscal balance of the proposal of the preceding paper.

Robert Eisner, "Don't Sock the Elderly, Help Them: Old Age is Hard Enough", Elder Law Journal, 5, 181-193 (spring, 1997).

K. G. Manton, L. Corder, and E. Stallard, "Chronic Disability Trends in Elderly United States Populations: 1982-1994". Proc. Natl. Acad. Sci. USA, 94, 2593-2598 (1997).

John Mueller, "Three New Papers on 'Privatizing' Social Security, One Conclusion: Bad Idea", reports prepared for the National Committee to Preserve Social Security and Medicare, Washington, DC (1997).

Robert J. Myers, "The Proposals of the Advisory Council on Social Security: Not Only Undesirable, but Impossible of Enactment", paper presented at the Bowles Symposium, Georgia State University, Atlanta, GA, September 26-27, 1996.

C. E. Steuerle and J. M. Bakija, "Retooling Social Security for the 21st Century" (Urban Institute Press, Washington, DC 1994).

Mark Weisbrot, "Unequal Sacrifice. The Impact of Changes Proposed by the Advisory Council on Social Security", The Preamble Center for Public Policy (January 1997).