Premarital cohabitation and marital stability in West Germany. Kölner Zeitschrift für Soziologie und Sozialpsychologie 49

This paper investigates whether premarital cohabitation affects the likelihood of divorce in a subsequent marriage. The well-known “trial marriage “ hypothesis postulates that cohabitation should decrease the divorce rate because high-risk partners will separate before marrying. We use data from the West German Family Survey from 1988 to test this hypothesis. We find – contrary to the trial marriage hypothesis – that cohabitation increases the risk of a subsequent marriage nding in divorce. However, more sophisticated analyses indicate that this result is due to self-selection: High-risk persons sort themselves into cohabitation. Using a bivariate probit model, we can show that cohabitation – net of self-selection – actually decreases the risk of divorce. KEY WORDS: Cohabitation, Divorce, Self-Selection, Family Economics 1A change in marriage behavior that parallels the rise in divorce rates has occurred in most Western industrial nations. While it was still the exception at the beginning of the 1970s for potential marriage partners to found a joint household before marriage, premarital cohabitation has become the rule today in many countries (Blossfeld 1995: 20). In Germany, for instance, less than 10 percent of couples lived together before marriage prior to 1970. In c ntrast, more tha