A Mixture Model for Nuptiality Data with Long-Term Survivors

Paraskevi Peristera, Stockholm University
Gebrenegus Ghilagaber, Stockholm University

The tacit assumption in the analysis of duration data with censored observations is that censoring time is independent of event time. This implies that the individuals who have not experienced the event of interest by the end of the study do not differ in any systematic manner from those who have experienced the event. This assumption is violated in many situations. For instance, in the analysis of data on family formation, individuals with a tendency to remain single over long periods may be overrepresented among the censored observations. In this work, we use a mixture model where parameters of a binary logistic regression model are jointly estimated with those of a continuous intensity model for family formation. The model allows for incorporation of a frailty-term for unobserved heterogeneity. Preliminary results show that failure to account for long term survivors may yield misleading results and plague the purpose of the analysis.