Introduction: The position papers presented during the last few weeks focus on why reform is important, why it should happen now, and on structural alternatives for the Social Security system. This paper will focus on special issues affecting women.

Why special focus on women: Women are 60% of Social Security beneficiaries, and all of the discussion affects women. However, there are some special issues requiring thought, which we would like to raise here. They are:

• Women live longer and many of them live alone in their final years
• Women are much more likely to have periods of caregiving interspersed between periods of work
• Women provide much more of the caregiving
• Women as a group earn less than men - at any time - and over their lives, as a result of family and work patterns

All of this adds to lower Social Security benefits and too many older women in poverty and near poverty.

Link of work, family and benefits under the current system: Social Security benefits are based on a combination of an individual's work history and their family history. Benefits are paid either on the basis of personal work history, or a spouse's work history, but not both. The greater amount is paid. If benefits are based on a spouse's work history, they are equal to 50% of the spouse's benefit while both are alive, and then 100% is paid to the survivor after the death of the spouse. A person who gets divorced gets the same benefits as a spouse if they were married for at least ten years before divorce. This works pretty well for individuals who worked for a full career and for others who were long-term homemakers, but it may not work well for someone who was a homemaker for a number of years and worked for a number of years. Sometimes it works well for a divorced spouse, and sometimes it does not.

Single earner vs. dual earner couples: The benefits work very differently for single and dual earner couples. Single earner couples get both a worker benefit and spouse benefit. Dual earner couples will get two worker benefits and no spouse benefits unless the combination of a worker benefit for the higher earning work and a spouse benefit for the lower earning worker gives them a higher benefit. This can lead to serious inequities. The following is an example of a couple both age 65 and earning a total of $34,200 in 1998. One is a single earner family and one is a dual earner family each earning $17,100. The following table shows their contributions to social security, their benefits in 1998 if both retired, and the widow's benefits if they retired and the husband died right away.

On retirement the single earner family gets $275 per month more, and the widow gets $408 per month more.

Reform options focusing on this issue: Changes to widow's benefits are important even if there are no other changes made to the current system. One of the elements of some reform options calls for reducing the spouse benefit to 33% while both are alive, and then increasing the widow's benefit to 75% of the combined benefit of husband and wife. This change would go a long way to solving this inequity. It would, however, reduce the benefit of single earner couples by 11%.

Individual accounts and widow benefits: Moving to individual accounts does not address the issues around widows, and it may make them worse. Here are some of the considerations:

• If benefits can be paid as a lump sum, there is a danger that widow's will be left without resources. The two earner family issues are different, but do not go away. The risks are greatest in the single earner family.
• If benefits must be paid as an annuity, then protection of widows depends on the form of the annuity. The two earner family issues need to be evaluated to see how to make this work well regardless of the earnings split.
• There are different issues with regard to pre-retirement survivor benefits. These are most often payable to women, and would need to be provided in addition to the individual accounts. The account balances could be set up so that they could be transferred to a surviving spouse, but in many cases this would be very inadequate.

Divorce: Benefits on divorce under the current system work well for some beneficiaries and poorly for others. Current benefits seem consistent with the general idea of lifetime alimony. There are several points where current benefits seem illogical in the absence of long-term alimony.

• It makes no sense to pay a benefit, which increases when a former spouse dies. This gives divorced spouses an incentive to harm their former spouses.
• The benefit is a disincentive to remarriage.
• There is a discontinuity at ten years of marriage.
• People with mixed periods of work and divorce may well lose out.

In the event individual accounts are used, then a method of sharing account balances would be needed on divorce. Similar issues arise with regard to pension benefits. One view is to allow the courts to split accounts. At the other extreme is to simply require a 50/50 split. These issues are complex. There are a number of pros and cons around benefit sharing.

My position: Regardless of the reform option being studied, it is important that it be analyzed to determine the effect on people in different family situations, at different earnings levels, with different patterns of work and caregiving. This is particularly important in order for us to understand how women may be affected at different income levels.