Objectives
This report presents national estimates of the probabilities of
marital and cohabitation outcomes for women 15–44 years of age in 1995, by a
wide variety of individual-and community-level characteristics. The life-table analysis
in this report takes a life cycle approach to estimate the probabilities that:
+ a woman will marry for the first time,
+ an intact first cohabitation will make the transition to marriage,
+ a first cohabitation will end in separation,
+ a first marriage will end in separation or divorce,
+ a disrupted first marriage will be followed by a new cohabitation,
+ a separation from first marriage will result in divorce,
+ a divorce from first marriage will be followed by remarriage, and
+ a second marriage will end in separation or divorce.
Methods
The life-table estimates presented here are based on a nationally
representative sample of women 15–44 years of age in the
Results
The analyses show that various individual and community-level characteristics are related to the marital and cohabitational outcomes examined in this report. The results consistently demonstrate that the cohabitations and marriages of non-Hispanic black women are less stable than those of non-Hispanic white women. An analysis of trends over time suggests that differences by race/ethnicity are becoming more pronounced in recent years. Racial differences observed are associated with individual characteristics and with the characteristics of the communities in which the women live.
Discussion:
In each comparison of racial/ethnic subgroups, the results consistently suggest that the unions of non-Hispanic black women are less stable than those of non-Hispanic white or Hispanic women. Black women are less likely to marry by age 30 and less likely to make the transition from cohabitation to marriage, and their cohabitations are more likely to disrupt than those of other women. The first marriages of black women disrupt faster than the first marriages of other women. Black women are less likely to enter a cohabitation after the dissolution of the first marriage. The separations of black women are less likely to make the transition to divorce, and the interval between divorce and remarriage is longer for black women. The data suggest that the remarriages of black women disrupt faster than the remarriages of other women. The trend analysis suggests that, at least for some of these marital outcomes, the differences by race are increasing over recent decades. The differences between white and Hispanic women are smaller. Some researchers have suggested that these differences may be related to higher rates of unemployment, incarceration, and mortality among the black population, their lower levels of educational attainment and earnings, their previous experiences as children of unmarried or less-educated parents, and higher rates of poverty and lack of job opportunities in the communities in which they live (13,19,20). The findings in this report suggest that individual characteristics such as race/ethnicity may not be the sole determinants of marital and cohabitation success. The neighborhoods in which people live may be important, and differences in marital and cohabitation outcomes between white and black women may depend to some extent on the community environments in which the women live. Both non-Hispanic white women and non-Hispanic black women who live in neighborhoods with high levels of poverty, receipt of welfare and unemployment, and low levels of income and education are more likely to experience separation and divorce. Black women live disproportionately in low-SES neighborhoods. Whether the lower marital success of black women is due to their disproportionate prevalence in low-SES neighborhoods, to individual characteristics, or to other factors, is a question for further study.
This report presents analyses of eight cohabitation and marital outcomes by a wide variety of individual and community characteristics. This presentation is not meant to represent a definitive explanation of any of the outcomes presented here. Rather, the intention is to provide benchmark statistics by a wide variety of characteristics, and to encourage researchers to consider these factors when studying marital outcomes. Of particular note is the presentation of statistics on the probability of postmarital cohabitation. The authors know of no other study that analyzes this facet of the marital life course with nationally representative data.
The outcomes analyzed in this report deserve further study using multivariate statistical techniques. These techniques allow the statistician to determine whether characteristics used in this report are less important or more important when other characteristics are controlled. Hazards models are one method of controlling for multiple characteristics simultaneously (47). For example, a hazards model could show the effect of income on the chances of marital disruption, controlling for education, race/ethnicity, age, and other characteristics. Another possibility for further study is multilevel modeling, to control for the effects of the community (or neighborhood) environment (54). This report suggests that community characteristics are important for a full understanding of the outcomes analyzed in this report.
This report has focused on changes in the relationships between spouses (or between cohabiting partners) and has not dealt, in detail, with the effects of children on these outcomes. Although we have tried to examine the presence of children as a characteristic that may be associated with a marital or cohabitation outcome, this is a topic that deserves closer scrutiny. In many of our analyses, the measurement of children at the beginning of the interval is less than satisfactory. For example, in the analysis of first marriage disruption, it would be better to know how many children were present in the family at the time of the disruption, rather than at the time the marriage began, but because many of the first marriages had not disrupted by interview (were censored), it did not make sense to code the presence of children at the time of disruption. Although it makes sense to measure the number of children at the beginning of the interval when studying the transitions from separation to divorce, separation to postmarital cohabitation, and divorce to remarriage, it is less than satisfactory for the earlier outcomes. In addition, characteristics of the children themselves could be studied, such as their age, gender, and whether they are biological, adopted, or step-children. To do so, however, would require an analysis specifically designed to measure the effects of the characteristics of children on these marital outcomes.
Finally, another avenue of further research is non-coresidential families (i.e., families that do not live together). The term ‘‘fragile families’’ has been used to describe families that do not share a single residence and are at higher risk of poverty and family instability than married-couple families (55). This report focuses on cohabitation and marriage, both of which are shared-residence relationships, and does not analyze alternative forms of family life. The stability of such families and the effects of individual and community characteristics on that stability, is an important topic, especially for understanding low-income families.
Although the statistics presented in this report are descriptive in nature, it is possible to draw some conclusions about the characteristics of individuals and communities that may contribute to the stability of cohabitations and marriages. Cohabitations and marriages tend to demonstrate more stability if the woman was older at the time the cohabitation or marriage began, if her family income is higher, if she has a religious affiliation or reports that her religion is important to her, if she was raised through childhood in a two-parent intact family, if she had never been forced to have intercourse, if she had no children at the start of the cohabitation or marriage, if her first birth was more than 7 months after the beginning of the cohabitation or marriage, if she has never suffered GAD, if she is the same race/ethnicity as her husband, or if she lives in communities with higher median family income, lower male unemployment, less poverty, less receipt of welfare, and more adults who are college-educated. Some of these characteristics show stronger effects for the stability of marriage than for the stability of cohabitation and some of the effects vary by race/ethnicity. With the exception of controlling for race/ ethnicity, none of these characteristics were tested in a multivariate context, and it may be that some of the effects are spurious, reflecting the effects of other variables. Researchers are encouraged to consider analysis of these outcomes with a multivariate approach, and to include controls for race/ethnicity and community context in the analysis of these outcomes.