Social Security Privatization

Testimony of Michael Tanner
Director of Health and Welfare Studies, Cato Institute

Subcommittee on Social Security and Family Policy,
Senate Finance Committee

August 2, 1995

Mr. Chairman, distinguished Members of the Committee:

I am Michael Tanner, director of health and welfare studies at the Cato Institute and director of Cato's Project on Social Security Privatization. I want to thank the committee for the opportunity to testify on what may be one of the most important public policy issues facing this country at the 20th century draws to a close.

In less than two weeks, Social Security will celebrate its 60th anniversary. As it does so, it is an institution in profound crisis. According to a recent public opinion poll, more young Americans believe in UFOs than believe they will receive their Social Security benefits. The unfortunate fact is that, while their views on extraterrestrial visitation may be problematic, their opinion on Social Security may be perilously close to correct.

Recently, the government's own actuaries reported that the Social Security Trust Fund will go broke in 2030. However, this estimate itself may be unduly optimistic because the Social Security Trust Fund is really little more than a polite fiction. For years, the federal government has used the Trust Fund to disguise the actual size of the federal budget deficit, borrowing money from the Trust Fund to pay current operating expenses and replacing the money with government bonds. The real crisis starts, therefore, not when the trust funds run out, but when they peak and start to decline. At that point the trust funds must start turning in bonds to the federal government to obtain the cash needed to finance benefits. But the federal government has no cash or other assets to pay off these bonds. It can only obtain the cash by borrowing and running a bigger deficit, increasing taxes, or cutting other government spending.

Even if Social Security's financial difficulties can be fixed, the system remains a bad deal for most Americans, a situation that is growing worse for today's young workers. Payroll taxes are already so high that even if today's young workers receive the promised benefits, such benefits will amount to a low, below-market return for those taxes. Studies show that for most young workers such benefits would amount to a real return of one percent or less on the required taxes. For many, the real return would be zero or even negative. These workers can now get far higher returns and benefits through private savings, investment, and insurance.

In a forthcoming study for the Cato Institute, financial analyst William Shipman considers the potential investment return under a variety of scenarios. Mr. Shipman considered the examples of both high and low income wage earners born at three different dates (1930,1950, and 1970). Shipman then compared the social security benefits that the individual would receive with the potential return that the individual would have received if he or she had been allowed to invest an amount equivalent to the payroll tax in either stocks or bonds.

If, as is likely, the system's impending financial crisis forces reforms such as raising the retirement age, means-testing benefits, or increasing the payroll tax, Social Security will become an even worse investment for today's young workers.

The only viable alternative that will continue to guarantee that older Americans will be able to retire with dignity is to privatize the Social Security system.

What would a privatized system would look like? While it is not necessary at this point to go into all the details of how such a system would function, the logical alternative would be some form of mandatory savings program. For example, the 11.2% payroll tax that is the combined employer-employee contribution to OASDI, the Old-Age and Survivors Insurance and Disability Trust Fund portion of the Social Security program, could be redirected toward a Personal Retirement Account (PRA) that is chosen by the individual employee.

Under this scenario, Personal Retirement Accounts would operate similar to current Individual Retirement Accounts (IRAs). Individuals could not withdraw funds from their PRA prior to retirement, determined either by age or PRA balance requirements. PRA funds are the property of the individual. Upon death, remaining funds would become part of the individual's estate.

PRAs would be managed by the private investment industry in the same way as 401k plans or IRAs. Individuals would be free to choose the fund manager that best met their individual needs and could change managers whenever they wished. The government would establish regulations on portfolio risk to prevent speculation and protect consumers. Reinsurance mechanisms would be required to guarantee fund solvency. One way to protect against excess risk would be to have PRA fund balances reported in two categories. All funds up to a calculated minimum requirement would be designated as "Basic" fund balances. "Basic" fund balance limitations would be calculated by determining 110% of the present value of the actuarially-determined retirement annuity necessary to provide a real monthly income after retirement equivalent to the current national minimum wage. The future annuity cash flow could be discounted using the current 1-year T-Bill rate, providing an expected real rate of return without long-term inflationary expectations. "Basic" fund balances would be subject to asset allocation restrictions that would limit the risk to which they could be subjected. For example, there may be a limitation on how much of the portfolio could be placed in common stocks. Funds accumulated above the "Basic" fund balance would be reported as "Discretionary" fund balances. "Discretionary" PRA balances would not be subject to the asset allocation restrictions of "Basic" balances and would, therefore, be eligible for a wider range of investment options.

In addition, the government could maintain a safety net, guaranteeing a minimum pension benefit. The minimum pension could be set to a benchmark such as the minimum wage. If upon retirement the balance in an individual's PRA is insufficient to provide an actuarially-determined retirement annuity which would provide a real monthly income equal to the minimum wage, the government would provide a supplement sufficient to bring the individual's monthly income up to the level of the minimum wage. Given historic rates of return from the capital markets, even a minimum wage earner will receive more than this minimum from the new system if he or she participates their entire life. Therefore, in the absence of a major financial collapse, the safety net would be required for few aside from the disabled and others outside the workforce.

Those presently in the workforce would have the option of remaining in the current Social Security system or switching to the new private system. Individuals entering the workforce after implementation of the private system would be required to participate in the new system. Thus, the current system would eventually be phased out.

It is important to realize that the idea of privatizing Social Security is not completely untested. Chile's Social Security system predated ours, having started in 1926. By 1981, Chile faced the same difficulties as presented by the U.S. Social Security system today. In response, Chile privatized its system.

The new Chilean system which went into effect on May 1, 1981, is a true "defined contribution" pension plan with mandatory contributions of 10% of earnings for program participants. The pension available from the system is simply that which is actuarially computed from the accumulated contributions.

When the new system began, those in the old system were given the option of switching to the new. After 1982, all new employees were required to join the new system. As of 1992, approximately 90-95% of all persons under the old system had shifted.

Contributions to the system are paid entirely by the employee, with no employer payroll tax to support it. At the initiation of the system, however, all employers were required to give a wage increase of 18% to all employees, approximating the increased cost to the worker but less than the reduced cost to the employer of the new system.

Pension funds are invested in security portfolios administered by private organizations known as "Adminstradoras de Fondos de Pensiones" (administrators of pension funds, or AFPs). Twenty-one AFPs, which compete with each other on the basis of investment returns and service, are closely regulated, complying with government mandated financial and investment requirements. Each worker chooses the AFP in which he wants to participate, and may transfer fund balances at his own discretion up to four times per year. Like any other mutual fund, the AFP invests fund balances in a portfolio of securities, and charges the portfolio an administrative fee for its services. Fees are a combination of a flat monthly percentage plus a percentage of earnings, and the AFP fee charges are well publicized so that individual workers may consider the charges in their choice of funds. Fees average 1% of total wages, down from more than 2% since the system was started. Several of the funds, in fact, are owned and operated by U.S. investment firms. Provida, with 25% of the system's assets and the largest AFP, is 42% owned by New York-based Bankers Trust (acquired as part of a $45 million debt-for-equity swap in 1986), and Santa Maria, the second-largest AFP, is 51% owned by Aetna Life & Casualty of Hartford, Connecticut.

AFP asset allocation, however, is strictly regulated by the government. Portfolios must consist of no less than 50% investment in government obligations, "agency" issues of other government-guaranteed securities leaving no more than 50% of the portfolio that may be invested in private-sector securities. Common stocks may comprise a maximum of 30% of the portfolio (with no more than 7% of the total in any one company and no more than a 7% stake in any particular company). Finally, only stocks on a government-approved list may be purchased. No foreign securities have made the list.

The entire system provides for automatic market indexation by translating contributions into investment units. Investment unit value is calculated similarly to a mutual fun Net Asset Value (NAV), taking the total current value (in pesos) of the total funds of the AFP divided by the total number of investment units of all members at a point in time.

Minimum retirement ages are 65 for men and 60 for women. Participants may, however, retire earlier if the pensions payable is at least 50% of their average earnings over the previous 10 years and 100% of the legal minimum monthly wage. Three alternative methods for determining the pension value are available at the participant's discretion:

1. The accumulated contributions may be used to purchase a life annuity from a private insurance company. Annuities ;must be government approved and must include survivor benefits for dependents.



2. The retiree may elect to receive a pension paid from the AFP directly. It is calculated using the life expectancy of the family group applied to the balance remaining in the account, which continues to earn income based on the AFP's performance.



3. A partial withdrawal may be used to purchase a private annuity with the remaining paid out directly from the AFP.

Perhaps the most innovative feature was the means by which the Chilean government sought to provide for transition to the new system. The government issues "bonos de reconocimiento"(recognition bonds), which effectively recognize the value of the obligation incurred by the government (the taxpayers) to those who have participated in the old system.

"Bonos" are available to any worker who had at least 12 months of contributions to or coverage under the old system in the 60 months prior to the start of the new system. The calculation of the "bonos" due an individual system participant is technically complex, but provides the financial mechanism for the transition to the new system. An alternative method of calculation allows anyone who contributed to the old system after July, 1970, to receive value for the participation. "Bonos" are essentially government bonds that pay 4% annual interest and add to the accumulated contribution value of the AFPs at the time of retirement. Interest on the bonds is paid out of the government's general revenue fund and is in no way supported by the new pension system.

Finally, a minimum retirement pension is payable to individuals with at least 20 years of contributions to the old and new systems combined. Disability cases have a two year contribution requirement. The minimum pension is set at 85% of the government-mandated monthly minimum wage, but does not apply to workers in the "informal" labor market who have never contributed to a plan. Disability and survivor benefits are not paid from the 10% contribution to the AFP. An additional required contribution (variable by AFP and averaging about 1.5%) is collected by the AFPs and paid to private insurance companies to purchase private insurance coverage for the group of workers contributing to that AFP.

The success of Chile's public pension privatization can be measured in many ways. Whereas in the late 1970s there were virtually no savings, now the cumulative assets managed by AFPs are about $23 billion or roughly 41 percent of GDP. During the past decade Chile's Real GDP growth has averaged over 6 percent, more than double that of the U.S. And for the five years ending 1994 the annualized total return of the Chilean stock market was 48.6 percent versus 8.7 percent for the U.S.

But most important, beneficiaries are receiving much higher benefits. Since the privatized system became fully operational, the average rate of return on investment has been 13 percent per year. As a result, the typical retiree is receiving a benefit equal to nearly 80 percent of his average annual income over the last 10 years of his working life, almost double the U.S. replacement value. Chile's reforms are seen as such a huge economic and political success that countries throughout Latin America, including Argentina, Peru, and Columbia, are beginning to implement similar changes.

Obviously the Chilean model cannot be directly imported to the United States. There are many differences between the two countries economies and cultures. In addition, there are areas where the Cato Institute believes the Chileans were to restrictive or made other errors. However, the Chilean experience shows that the privatization of Social Security can be carried out successfully.

The most difficult question for any proposed privatization of Social Security is the issue of the transition. Put quite simply, regardless of what system we choose for the future, we must continue benefits to today's recipients.

At the same time, however, we should understand that the design of a new system has nothing to do with the liabilities that (rightly or wrongly) have been accrued in the past. The government's obligation to current (and even future) retirees is unchanged by a decision to privatize the system. What does change is the willingness to acknowledge currently unfunded liabilities. The commitments entered into by the federal government as a result of spending current Social Security receipts are what financial economists call a sunk cost. The liability has already accrued and exists whether we privatize the system or not. In the future the government, if it is to honor its commitments, will be forced to either tax or borrow additional funds from the private sector to finance the cash outflows necessary to meet these obligations.

Still proponents of privatization bear the responsibility for suggesting funding mechanisms for the transition. The reality is that the transition will probably involve some combination of four approaches.

The first of these is a partial default. Any change in future benefits amounts to a partial default. This could range from such mild options as raising the retirement age, reducing COLAs, or means-testing benefits to "writing off" obligations for individuals under a certain age who opt into the private system. For example, any individual under the age of thirty who chooses the private system may receive no credit for past contributions to Social Security.

The second solution to the problem of unfunded liabilities is one that provides for the recognition of the present value of those liabilities in the form of government bonds to be issued to current system participants and taxpayers. Once we have decided on the extent of the limited defaults the system will tolerate, it is not a difficult calculation to determine the moral (if not legal) stake each working American currently has in the implied promise of the current Social Security system to each of us. The system currently calculates a figure known as a "Primary Insurance Amount" (PIA) based on a review of the taxpayer's average monthly earnings from employment covered by the program. "The PIA is the benefit for a single retired worker who starts receiving his monthly Social Security check at the normal retirement age." Normal retirement age is now 65, but will rise to 66 in 2008 and to 67 in 2027 (and could, as above, rise further with further system defaults). Benefit computations are based on earnings during the 35 years of highest covered earnings up to age 62 (or the worker's age when he or she applies for benefits, whichever is later), and the wages in each year of the earnings record before age 60 are multiplied by an index factor to take into account the growth in national average earnings since that year. The result is the individual's "average indexed monthly earnings" (AIME), which is then multiplied by percentages that are weighted to favor low-income earners to finally determine the Social Security benefit.

The AIME can be used to calculate for each American worker today his or her expected retirement benefit given tax "contributions" to the system to date. Current retirees' benefits are, of course, already determined. The present value of the actuarially-calculated annuity due each system participant may then be easily calculated discounting at the T-Bond rate, and each system participant can be issued zero-coupon T-Bonds maturing at their projected retirement date. The bonds would be placed in each individual's PRA.

It is important that these zero-coupon Treasury securities then be allowed, in turn, to trade on the secondary market. Within the limitations already described for Basic fund balances, both current retirees and prospective retirees should immediately begin to personally manage their PRAs according to their own risk preferences, thus increasing the diversification benefits of individual PRA portfolios and maximizing personal liberty.

A third method of financing the transition would be continue a small portion of the current payroll tax. For example, workers could be allowed to invest 10 percent points of the current 12.2 percent OASDI payroll tax, with 2.2 percentage points continuing to fund a portion of current benefits.

Finally, Congress could identify additional spending cuts and

In conclusion, we must realize that Social Security is an unfunded pay-as-you-go system, fundamentally flawed and analogous in design to illegal pyramid schemes. Government accounting creates the illusion of a trust fund, but in fact the government spends excess receipts immediately. The liabilities already created are unrecognized by the government accounting system, but represent sunk costs that cannot be recovered. Only adjustments in spending patterns can pay for those commitments. The choice remaining is between continuing to support a bankrupt system, or building a financially sound structure for the future.

Only private pensions with individual property rights to accumulated fund balances can create a secure pension system. Evidence of such a system's effectiveness is available from the example of Chile, which privatized its system in 1981. The plan has been a success but stops short of full privatization. Various plans have been proposed for the U.S., but each suffers the effects of compromise with central-planning approaches.

A plan that achieves the dual objectives of security and personal liberty would divert current OASDI payments to private Personal Retirement Accounts, similar to Individual Retirement Accounts (IRAs), managed by the financial securities industry. Modern risk-management methods should be used to minimize risk for the portion of the account necessary to finance minimum retirement needs. Personal risk preferences should be allowed to guide the investment of fund balances in excess of the minimum.

Transition to a new system requires a recognition of current intergenerational commitments and makes choices that minimize transactions costs as we liquidate obligations to ourselves and integrate system liabilities into a privatized financial structure.

Thank you, I look forward to answering you questions.