Determinants of Divorce over the Marital Life Course

Abstract

Using data from the young and mature women samples of the National Longitudinal Survey, this paper examines how the determinants of divorce (and separation) vary by the duration of marriage. In general, we find little evidence that the strength of previously identified predictors of divorce varies by marital duration. Variables such as race, wife's labor force participation, husband's employment, and urban residence seem to influence the probability of divorce, irrespective of the stage in the marital life course. The principal exception to this finding is the effect of wife's education, which appears to decrease the probability of divorce at early marital durations but to increase it at later durations. There is also suggestive evidence that the effects of home ownership and age at marriage may vary by marital duration. Demographic trends frequently shape research interests, and current trends are no exception. Increased attention from both demographers and sociologists is being focused on aging, the life course, and divorce. The source of interest in divorce is clearcut, since recent projections suggest that close to half of recent marriages will end in divorce (Preston and McDonald, 1979; Cherlin, 1981). Interests in aging and the life course probably stem at least in part from the aging of the population, with projections of up to 18 percent over age 65 by the year 2030 (Soldo, 1980). This concern may also be due to the aging of the baby boom generation, which has captured public attention, and which is now beginning to experience mid-life crises (Preston, 1984). While both areas of research are expanding at a prodigious rate, they have done so for the most part independently. Of course, demographers have long studied age/time-related determinants of divorce, such as age at marriage, marital duration, and life-cycle characteristics such as presence of children. Recent research (Morgan and Rindfuss, 1985; Thornton and Rodgers, 1984) has attempted to sort out the effects of these highly related variables. However, these analyses have focused almost solely on the age variables, and have not addressed the larger picture of how the divorce process may vary over the life course.