Monday, Nov 3
Melvin Stephens, Estimating Program Benefits
Review originally in Journal of Marriage and Family, pp. 1088-1089, Volume 70, Number 4, November 2008. Reproduced with permission of the reviewer.
Do family characteristics at birth foretell union formation decisions over 20 years later?
Cohabitation research started with the question of whether married couples who cohabited before marriage had a higher or lower risk of divorce compared to married couples who did not cohabit. One of the most replicated findings in social science is that cohabitors have a higher risk of divorce than noncohabitors. Those early studies provided the impetus to delve further into understanding "why."
In their book entitled Marriage and Cohabitation, sociologists Arland Thornton, William Axinn, and Yu Xie examine intergenerational factors that predict whether young adults will cohabit or marry directly. A compelling reason for understanding union decision making is because "decisions made during the transition to adulthood have a particular long-lasting influence on the remainder of the life course because they set individuals on paths that are sometimes difficult to change" (p. 13). Targeting an academic audience, they present results from the Intergenerational Panel Study of Parents and Children (IPS). This longitudinal community study of 950 families spanned 31 years from 1962 to 1993 and included White women who gave birth in 1961 in the Detroit area and their child born in 1961. Mothers were interviewed eight times over the course of the study, and the target child was interviewed three times as a young adult from age 18 to 31. The authors outline direct associations and more complex indirect associations between family factors and the union choices of the younger generation.
The authors examine 45 maternal and young adult factors as predictors of whether the target child will cohabit or marry. Maternal factors, arranged temporally from before the birth of the target child through the young adulthood of the child are grouped into the following domains: family immigration/farm background, parental socioeconomic standing, maternal marital experience, parental childbearing, religion, family organization, and maternal attitudes. Young adult factors include (a) sexual experience before age 18; (b) attitudes at age 18 about religion, sex, education, and marriage; and (c) education and expected earnings after age 18. The authors examine a comprehensive array of sociodemographic and social psychological variables. This study does not examine genetic or physiological attributes or clinical characteristics such as depressive symptoms or alcohol or drug use.
The authors carefully examine each domain in temporal order chapter by chapter and present a table in Appendix A with all of the variables integrated into a single model. There are many results and more complex development of the data, but for the sake of parsimony I highlight some of the significant direct associations from Appendix A. Young adults were more likely to cohabit (a) when their mothers were widowed, more gregarious, and had more egalitarian sex role attitudes and (b) when the young adults had sex before age 16, were less religious, had lower school grades, and were less likely to be 1088 Journal of Marriage and Family 70 (November 2008): 1088-1091 enrolled in school full time after age 18. Young adults were more likely to marry when their mothers had more children and lower educational expectations for their child. Young adults were also more likely to marry (a) when at age 18 they were more religious, preferred a younger age of marriage, and had lower career aspirations and (b) when after age 18 they had more years of education and expected higher earnings. Different patterns of young adult education and career aspirations predicted earlier or later marriage.
Clearly the authors have made a significant contribution to the cohabitation literature with such a comprehensive array of predictors of union formation measured longitudinally with a large sample. Meticulously, the authors present an enormous amount of data with clear organization and prose. Variables with direct associations that preceded union formation decisions temporally are interpreted as causal factors. Particularly compelling is the power of variables present before or at the time of the birth of the target children to predict the timing and choice of cohabitation or marriage. The authors help the reader with data tables and summary and concluding statements throughout the book and are careful to point out the limitations of their data. The inclusion of more heuristic models throughout would have helped the reader to grasp conceptually each step in the series of analyses more easily. Those who study union formation and cohabitation more specifically will find this book informative. If the major overarching question in cohabitation research is "why marriages of cohabitors are more vulnerable," then this book provides insight into that issue by examining the intergenerational strengths and vulnerabilities that partners bring into their romantic relationships.
CATHERINE L. COHAN, The Pennsylvania State University