Understanding Poverty Among Elderly Divorced Women

Published: 2002
Project ID: UM02-08


Until recent years, very few elderly people had ever experienced divorce. Most had been married to one person their entire adult lives. This pattern is changing rapidly. The sharp rise in divorce between 1965 and 1975, with persistently high rates ever since, implies that over half of marriages will end within 20 years. The birth cohorts that were the first to experience the rapid increase in divorce is also the leading edge of the baby boom, born primarily in the 1940s and 1950s. These cohorts are just now beginning to hit retirement ages. While 12 percent of today’s 67-year-old women are currently divorced, within just 15 years this will rise to 20 percent. And a much larger share of women are ever-divorced – they divorced but then remarried. These trends alone warrant a close investigation of the economic well-being of elderly divorcees. However, combining these trends in divorce with the fact that elderly divorced women are 5 times as likely to be poor than elderly married women, and one-third more likely to be poor than widows, it is surprising to us that only a handful of studies have examined this population. The goal of this study is to begin to establish the stylized facts regarding the economic status of divorcees and describe some of the difficulties they face. While this paper is only the beginning of a series of studies that are needed to address a wide array of issues, it reaches some important conclusions: Despite high poverty rates, divorced women are no less educated than married women; This similarity in education persisted over the past 35 years despite the fact that during the same time period the gap in poverty between these two groups increased from 2:1 (in 1967) to 4:1 (in 2001). Labor market earnings are a particularly important source of income for elderly divorced women; Divorced women’s level of labor force participation is comparable to that of married men; however, retirement of divorced women experiences no spike at ages 62and 65, as is the case for married men; Divorced women are quite distinct from separated women, with the latter much less educated, lower income, and minority; grouping these two populations together, as is often done, is a mistake; Initial analyses suggest that the reduction in the requirement on length of marriage from 20 to 10 years to receive Social Security divorcee benefits had little or no effect on alleviating divorcee poverty. A large number of questions remain unanswered, but given the speed at which the cohorts with high prevalence of divorce are approaching old age, the time is now to address these questions.