Gary Becker's theory of the family has had a tremendous impact on the understanding of marriage, yet it does not explain everything. This has been attributed, for example, to the fact that there is an element of randomness (irrationality?) in marriage. But adding randomness is not sufficient. For example, one predition of the model is that as female labor participation increases, the divorce rate increases because marriage conditions have changed. But now that female labor force participation has stabilized for about a generation, i.e. it is in a steady state, the divorce rate should go back down. Yet it remained high. What about going beyond Economics and using some insights from Sociology?
Take Signe Hald Andersen and Lars Gårn Hansen, who assume that marriage does not only happen on productivity matches but also on partner specialization matches. By that they mean that like-minded people, especially with respect to female labor market participation, divorce less. But this is not sufficient: such matching is subconscious, it does not follow a rational decision, but it is consistent with it. Concretely, this means that people match on these preferences without knowing it, and when conditions change (female wages rise), divorce rates increase, but less than without partner specialization matches, and do not have to decrease. The authors then demonstrate this with an agent based model.