Past, Present, and Future Patterns of First Union Formation in the United States

Hsien-Hen Lu, Columbia University
Neil G. Bennett, City University of New York
David E. Bloom, Harvard University

Until now, many scholars have documented current and historical patterns and trends in cohabitation and marriage, without any attempt to project future such patterns and trends in first union. The current study will extend our knowledge by applying the parametric models in the context of first union formation. Our goals are the following: to summarize in terms of descriptive statistics the trends over birth cohorts in the age at the formation of first union; to document differences among subgroups in these trends; to contrast these patterns and trends with those found for the formation of first marriages; and, to compare model fit among various semi-parametric and parametric models in order to find the most appropriate model suited to capture the patterns of first union over time. The benefit of applying parametric models is that we will be able to project the remaining first union experience among relatively young birth cohorts.

Presented in Session 105: Innovations in Population and Household Forecasting