Widowhood: Economic Issues

Widowed older women, on average, report lower incomes and are more likely to be poor than are other groups of elderly persons. This is true in the United States and in other countries as well, though the difference in the United States is greater overall than in other developed countries. In the United States, over 48 percent of the poor elderly are widows, even though widowed women account for only 26 percent of all persons age sixty-five and older. Compared to the slightly more than 4 percent of couples age sixty-five and older who are poor, about 20 percent of widowed women are poor. This poverty rate is considerably lower than the 50 percent of widows who were poor in 1970, reflecting gains in earnings for both men and women, as well as improvements in pension and Social Security benefits. Nevertheless, on average, married women in the United States experience a decline in income when their husbands die. Although widowers (men whose wives have died) are somewhat more likely to be poor than are married couples, data that follow couples over time do not show a decline in average economic well-being for men when wives die.

The decline in economic well-being upon widowhood is somewhat of a puzzle. There exists a well-functioning life insurance market that sells products that insure against the loss of income upon widowhood. In addition, legislation has increased the rights of spouses to spouses pension benefits. While the timing of death is uncertain for a given individual, death probabilities can be predicted with considerable accuracy, permitting the estimation of probabilities and length of widowhood. Information that is readily available on Social Security benefits, and requirements that pensions provide annual reports on workers’ accrued benefits, would seem to provide the information necessary for couples to protect against any loss of income upon one spouse’s death.

One explanation for the difference between the economic status of married couples and widows is the association between death probabilities and economic status. Poorer men are more likely to die than higher income men. This is partly because individuals with chronic health problems generally have lower lifetime earnings (and lower retirement income) and die at younger ages than healthier individuals. For these individuals, low wages and early death are both due to long-term health problems. On the other hand, individuals who work in lower-paying jobs may be engaged in more hazardous tasks, have no employer-provided health insurance, and be less able to pay out-of-pocket for health care. Their low earnings are a cause of poorer health and consequent higher mortality. Whatever the reason for the association between lower earnings and poorer health, the lower income of widows can be attributed in part to widows being drawn from couples who were economically worse off when married than were women of the same age whose husbands are alive. This, however, is only a partial explanation. Husbands in higher-income couples do die and, on average, a decline in economic status is measured for their widows as well.

For some couples the out-of-pocket costs of health care associated with a spouse’s terminal illness or nursing home stay can diminish resources. However, there is little evidence that uncovered health care expenditures are a major contributor to higher poverty rates among widows. The concern that some spouses were left impoverished by payments to long-term care facilities for the care of the other spouse led to the 1988 amendments to the Social Security Act, which permits a community spouse to retain assets and incomes above the limit that previously had been allowed for nursing home reimbursement under the Medicaid program. Currently, the institutionalized spouse’s eligibility for Medicaid reimbursement of long-term care costs is assessed after a specified share of the couple’s income and assets (approximately half, up to a maximum amount) is allocated to the community spouse. The remaining share is allocated to the institutionalized spouse, and it is only this share that must be spent-down in order for the institutionalized spouse to be deemed Medicaid-eligible. While these provisions are important to some couples, their influence on widows’ well-being is limited. Married men have a low probability of nursing home entry. That a similar decline in economic resources is not observed for married men when wives die suggests that the expenses of a terminal illness per se are not a major contributor to the lower economic status of widows.

There is stronger evidence that widows, in contrast to widowers, are more likely to suffer income declines due to the loss of the deceased spouse’s Social Security and pension benefits. Why this happens, despite the availability of survivor benefits from Social Security and employer-provided pensions, is described below. While couples could insure against the loss of this income through the purchase of life insurance, research shows that many elderly couples are underinsured against probable income loss, and that life insurance is most likely to be purchased by relatively wealthy and healthy couples. The limited evidence available on asset change upon widowhood suggests that some decline in assets occurs upon widowhood. While it is expected that the growth in financial resources owned by women will reduce the financial consequences of widowhood, the persistent association between health, low income, and low assets means that the wives in couples with low lifetime earnings may remain vulnerable to economic insecurity as widows.

A final source of measured income differences between married and widowed women deserves mention. In surveys, some individuals identified as widows may not, in fact, have experienced the death of a marriage partner. This occurs for two reasons. First, it is known that divorced women whose former husbands die often describe themselves as widowed, an identification encouraged in part by the availability of divorced survivor benefits from Social Security. Second, it is also likely that women long separated from their husbands or in common-law marriages may term themselves widows after those partners die. The lower incomes and assets of these individuals may be due to their ineligibility for the benefits provided to widows by laws and programs. Their classification as widows not only exaggerates economic differences between widows and married couples, but also obscures the reasons these women are worse off. This situation may deter policy aimed at improving the economic position of divorced, separated, and never-married elderly women.