The current Social Security system balances equity and adequacy; i.e.,
benefits are paid partially on the basis of perceived need and partially on
level of earnings. There is insurance, which results in redistribution from higher to lower income workers and from single individuals and dual income couples to single earner couples. Benefits include retirement income, survivor benefits and disability
benefits. There is insurance or redistribution to those who collect some of these benefits from those who do not. There is also redistribution between generations.
What Is Meant by Privatization?Privatization is a term that has been used to describe a number of different types of changes and proposals. These proposals combine the design elements of Social Security systems in different ways. The elements of any Social Security system will include:
Death and disability benefits can be included, so that the system provides
not only for retirement, but also for survivors in the event of premature
death, and for disabled individuals. Privatized systems may include these
benefits in addition to the defined contribution accounts. These benefits
would be funded by a tax in addition to the contribution to the defined
contribution account. The design of such benefits under a privatized system
would probably be different from the current system, and there would be
many alternatives for design. The current U.S. system includes death and
disability benefits as well as retirement benefits.
Implications for WomenPrivatization will hurt some women and help others. It is important to remember that women are in many different family situations and income levels and to look at each issue separately. Changing the basic design from defined benefit under the current system to a simple defined contribution plan will allocate more dollars to higher earners and fewer to lower earners. Because women on average are lower earners, this will tend to hurt more women than it helps, but it will help higher earning women. This could be modified by using a defined contribution formula that has a progressive allocation so that a higher percentage of the lower tiers of earnings are allocated to the accounts. That would make the benefit allocation more similar to the current plan.
The method of allocation of benefits to family members and among families is very important to women, who more often are the lower earner in the family or who may not be working. It is unlikely that a defined contribution system would include spouse benefits similar to the current program. Spouse benefits could be provided by an additional tax. It is likely that lower earning or nonworking spouses would lose out under a defined contribution system. Earnings sharing has been suggested as a way to solve this problem. Under such an arrangement, the earnings of both spouses are shared equally by the partners. Earnings sharing under either privatization or a defined benefit system will help women in the aggregate. However, it will not help high-earning women with spouses with lower or no earnings.
Prefunding is mandatory under a defined contribution plan and possible under a defined benefit plan. Where there is prefunding, part of the benefit is provided by investment earnings and part by contributions, so that the plan appears to be more efficient than a pay-as-you-go plan. The higher the rate of return, the greater the savings. However, an argument for prefunding should not be taken as an argument for defined contribution and individual accounts. It should also be noted that this saving in retirement cost from investment earnings is not free. If we assume that the funds would be saved elsewhere if not saved in the Social Security system, then prefunding in the system can be compared to saving outside. It is a good deal if the earnings in the system are higher than in the alternative savings vehicle, but not otherwise. There are also economic issues linked to prefunding and privatization. In order for the system to work, the workers in each year need to be saving enough to buy the assets from retirees who want to redeem them to use the money for living expenses. This can be a major issue in a broad social system.
It should be noted that it is possible to have a benefit design that looks like individual accounts within a defined benefit plan. In such a structure, there are notional accounts that assign benefit credits to each individual, and investment income is credited according to a formula. Cash balance plans in the private sector are an example of such a structure. Prefunding would be optional (just as it is now) for a defined benefit Social Security system that included notional accounts.
Death and disability benefits under the current system are important. It is
very possible that they would disappear or be reduced under a privatized
system. However, if equal death and disability benefits are provided under
the two systems, privatization has no impact in this area if the two systems
are equally efficient. To measure the impact one needs to look at the design
of these benefits. The same is true with regard to minimum benefits.
The last comment relates to expenses. Privatized arrangements often involve individual insurance contracts, and are likely to have higher expense charges. This issue is much more important for lower income individuals.
Comments on a StudyThe Benefits of Social Security Privatization for Women, by Ekaterina Shirley and Peter Speigler, concludes that privatization would offer tangible financial benefits to women. The authors conclude that, "Under a properly designed system of individually owned, privately invested accounts, with a provision for earnings sharing between spouses, women would likely be significantly better off than under the current system."
In the study by Shirley and Speigler:
1. Full earnings sharing is assumed.
2. Higher returns are assumed as a result of having a fully-funded program invested in the private capital markets. A real rate of return on capital of 6.2 percent is assumed for defined contribution accounts, while the study discounts current law Social Security benefits at 2.3 percent.
3. The cost and implications of transition are ignored.
4. It is assumed that death and disability benefits will be replaced.
I am very pleased to see this fine study and to enter into the dialogue that it generates. However, the results of the study do not in fact demonstrate that privatization will help women.
My comments are as follows:
1. Earnings sharing will help many women, regardless of the use of defined contribution accounts. However, earnings sharing redistributes money between couples and, as discussed below, must be considered carefully.
2. The investment return can be attributed to prefunding rather than to use of defined contribution accounts per se. A real rate of return of 6.2 percent over a lifetime is extremely high, particularly for individually chosen investments. It also ignores that retirement can occur at any time, and that if funds must be withdrawn, it can be at a poor time in market cycles.
3. In order to determine whether defined contribution accounts will help women, it is necessary to separate out the implications of prefunding and of earnings sharing. It is my contention that a defined benefit system with earnings sharing and prefunding could be designed to have far greater advantages for women than the system proposed in the study.
4. Transition cost must be considered. Much of the current tax is used to pay benefits to current retirees, and to ignore transition, and assume that all funds generated by Social Security taxes are available to pay benefits to future retirees, is an unfair comparison.
It should also be noted that other studies of privatization have reached different conclusions. The Institute for Women's Policy Research in The Impact of Social Security Reform on Women states: "The PSA plan would entail great risks for low and moderate-income elderly women who are no longer married. Our results indicate that these are the women who are currently most likely to become poor; the PSA plan is likely to exacerbate poverty for these women." They indicate that the IA plan is more favorable for women than the PSA but that it still would harm some of the most vulnerable older women. (The IA and PSA are two of the plans proposed by the last Social Security Advisory Council, which included elements of privatization.)
Comments on Earnings SharingEarnings sharing will redistribute funds between couples and goes a long way to solve some of the inequities with regard to two-earner families. However, because it will reduce benefits to single-earner couples or cost more, it will probably not be acceptable politically. Some subsidy for single earner couples could still be included through designing an added benefit on top of the shared benefit, but this would be quite confusing and add more cost. Some of the other considerations with regard to earnings sharing are as follows:
1. This is not a suitable method to use in building a wage history for disability or preretirement death benefits. If earnings sharing is used to build retirement wage history, then the basis for determining death and disability benefits needs to be decoupled from the method of determining retirement benefits. This would be necessary because under an earnings shared approach with a substantial difference in earnings between the couple, each would still get equal credits for their Social Security history. if the higher earner were disabled and the benefit based on credits defined by half the couple's earnings, then the benefit would be inadequate. There are similar problems if the higher earner dies before retirement.
2. There are also potential problems caused when the higher earner retires first. If the higher earner retires first, but if benefits are based not on that person's earnings, but on half the earnings of the couple, the benefit will be inadequate. Conversely if the lower earner retires first, the benefit would be larger than intended.
3. Earnings sharing could be a disincentive to marriage.
In order for earnings sharing to work, it will need to be more carefully examined, and solutions to these issues developed.
Alternative to Earnings SharingAnother alternative is the proposal made by the recent Advisory Council. Under this proposal, the spouse's benefit would be reduced to 331/3 percent of the earner's benefits, and the benefit payable after the death of either spouse after retirement is increased to 75 percent of the combined benefit of the couple. This would help solve the problem of inadequate widow's benefits and reduce the inequity between single and dual earner families.
SummaryIn summary, I believe that the use of defined contribution accounts may well increase the number of elderly women in poverty and near-poverty. Unless there is earnings sharing or some other system to revisit the distribution of benefits within the family, a good minimum benefit and additional widow's benefits, defined contribution approaches are likely to be particularly difficult for widows. As indicated above, there are other issues that must be considered with earnings sharing. Social Security is the major source of support for the 50 percent of the elderly with the lowest incomes, and these people are very likely to be hurt.
Defined contribution accounts will help higher income women in two-earner families. They could get substantially greater benefits.
Earnings sharing combined with defined contribution accounts can be used to remove inequities between single and dual earner families. Use of defined contribution accounts will create a new set of problems for homemakers, and will leave them much more vulnerable. Earnings sharing is one solution to these issues. Defined contribution accounts will not solve problems of women who have a career mixed between work and homemaking. In fact, they might make these problems worse. Earnings sharing can mitigate them, either with or without defined contribution accounts, but as noted above, it would need to be accompanied by other changes.
Defined contribution approaches without disability benefits also hurt the disabled. With such benefits, it depends on the design of the benefits. Depending on whether annuitization is required, there may be increased challenges for widows.
Generally, Social Security involves insurance, and defined contribution approaches will hurt those who are given greater benefits relative to taxes and help those who have lower benefits relative to taxes. Defined contribution approaches will undo the social adequacy features of Social Security. On balance, I see defined contribution approaches as hurting women, particularly the most vulnerable.
Anna M. Rappaport is a principal and consultant with William M. Mercer Inc. and immediate past president of the Society of Actuaries.
This article is adapted from a paper prepared for distribution to the participants in the January 23 Americans Discuss Social Security teleconference. The views represented are solely the views of the author.
From the Enrolled Actuaries Report, February 1999, page 1