Marital separation and home ownership: a longitudinal analysis of life course dependencies

Philipp Lersch, University of Bremen
Sergi Vidal, University of Bremen

This paper sheds light on the interdependence between marital separation and home ownership over the life course. Past research showed that separations decrease the chances of being and becoming a home owner. This affects the subsequent life courses of the ex-partners and their children negatively, because home ownership provides a number of benefits that increase the resources and life chances of owners. We advance past literature in three ways: First, we assume that partners are aware of the risk of separation and that investments in home ownership are related to partnership choices in the past and expectations about the future and vice versa. More committed couples will select into home ownership, and those couples that are home owners are less likely to separate. By means of simultaneous equation estimation predicting both, the home ownership status and the risk of union dissolution, we allow partnership decisions to be endogenous in home ownership choices and the other way around. Second, we consider biographical dependencies and include the retrospective marital history in our analysis. Thereby, we can show how past separations affect the present housing situation. Third, we investigate two institutional contexts, i.e. Britain and Germany, as couple-specific investments are very sensitive to variations at the institutional level, above all regarding family policies. All results are presented separately for men and women to allow for gendered effects of institutions and life course dependencies. Longitudinal survey data from the British Household Panel Study and the German Socio Economic Panel Survey for the years 1991-2008 is used. Results show that separation clearly affects home ownership adversely, also when controlled for selection effects. Our findings indicate ambiguous gender effects. In addition, we find national differences in housing outcomes.