Background: The notion that increasing prevalence of cohabitation relative to marriage, and increasing age at first marriage are part of a broader shift in societal norms -- a second demographic transition -- is now well supported by studies focused on US and European populations. Recent research points to the similarly high prevalence of cohabitation in Latin America as perhaps signaling the diffusion of modern ideals and norms about union formation. In Central America this is unlikely to be the case given the long history and enduring acceptance of cohabitation that is unrelated to modern ideals. While there are studies that have documented this history and current prevalence, there is no research examining the intersecting life course pathways from adolescence through early adulthood that lead to marriage or cohabitation. This is not surprising given that available data for Central American countries are not ideally suited to studying the process.

Methods: We use retrospective questions from large, nationally representative Central American surveys (Guatemala, Honduras, and Nicaragua) to establish the timing of marriage or cohabitation and events that are closely tied to union formation. We utilize additive causespecific hazard models, and predicted transition probabilities based on selected covariate pathways, to study the competing risks of exiting from the status of never in union.

Results: Our results identify sexual activity and pregnancy as the primary drivers of union formation and indicate that education serves as a protective factor against union formation. We also find distinct differences among countries and a strong indication that cohabitations are less stable unions.