The dynamics of bargaining in partnerships

Christian Schmitt, Universität Rostock and German Institute for Economic Research (DIW Berlin)
Gosta Esping-Andersen, Universitat Pompeu Fabra

The literature on couple bargaining is theoretically rich, but empirical applications are few and far between. We challenge the widespread theoretical perspective of income based indicators being the only valid measures of bargaining power. The questions we pose are, a) whether the established perspective on income as prime bargaining resource is adequate to model couple negotiations, and b) which alternative indicators might be competing representations of bargaining power. We apply this bargaining nexus to two outcomes. First, we hone in on housework, i.e. on tasks that are unattractive and best avoided. Second, we investigate divorce propensities, i.e. an outcome that mirrors bargaining failure. Comparing income to alternative indicators of bargaining power, we consider the role of affection, represented by the ability to satisfy and please one's partner, and physical attractiveness. This approach has the aim of identifying, on one hand, the relative salience of ‘looks’ versus ‘money’; and, on the other hand, the relative salience of ‘money’ versus ‘making him or her happy’? And how do these relativities evolve over the life course of partnerships? We expect to find gender differences in how these indicators of bargaining power become influential. Previous research suggests that income is more important for the male earner, while attractiveness is primarily a female asset. Dynamically, one would expect that the relative salience of the partner’s ‘looks’ will fade as the partnership ages. In parallel, the male’s earnings-based bargaining power will weaken the more the female partner contributes to household income. In this case, his capacity to ‘make her happy’ should gain prominence. We estimate outcomes based on different measures of bargaining power, using data from the German Socio-Economic-Panel Study (SOEP), applying event history methods on divorce propensities. For the analyses of the division of household labor, we implement a panel-econometric random effects model.