| dc.description.abstract | In this paper we explore the changes in the relationship between female educational attainment and the risk of union disruption in seventeen countries: Austria, Estonia, Finland, Flanders (Belgium), France, West-Germany, Greece, Hungary, Italy, Latvia, Lithuania, Norway, Poland, Spain, Sweden, Switzerland, and the United States. We start from the hypothesis presented by William J. Goode (1962; 1979; 1993), stating that in the Western countries, the initially positive relationship between social class and divorce would gradually change during the modernization process and waning of barriers to divorce, so that eventually there would either be no relationship between the two, or that the lower classes would divorce more. We expand the examination to all unions - not just marriages - due to the increasing importance of non-marital cohabitation in many of our countries. We run separate models for all unions. We first analyse the data within each of the seventeen countries with discrete-time event-history analyses. We find important variation across countries in the relationship between education and union disruption, and find that the relationship has become more negative in five countries. Second, we use multi-level models for event-histories in discrete time to examine the macro-level correlates of this variation. We report that a higher level of employment in service sector and higher percentages of economically active women are associated with a more negative relationship between education and union disruption. Overall, we find support - although not unanimous - for Goode’s hypothesis, and conclude that the waning of social, and economic barriers to union disruption increases the risk of union disruption relatively more among the less educated. | en |
