Modeling American Marriage Patterns
This paper investigates the application of the three-parameter, Coale-McNeil marriage model and some related hyper-parameterized specifications to data on the first marriage patterns of American women. Because the model is parametric, it can be used to estimate the parameters of the marriage process, free of censoring bias, for cohorts that have yet to complete their first marriage experience. Empirical evidence from three surveys is reported on the ability of the model to replicate and project observed marriage behavior. The results indicate that the model can be a useful tool for analyzing cohort marriage data and that recent cohorts are showing relatively strong proclivities to both delay and forego marriage. Consistent with earlier work, the results also indicate that education is a powerful covariate of the timing of first marriage and that race is a powerful covariate of its incidence.
Journal of the American Statistical Association, Vol 85, No. 412, December 1990.