Investigating the dynamic interdependency between poverty and marital separation

This paper investigates the links between poverty and marital separation. In order to account for both unobserved heterogeneity and complex dynamics, the paper builds and estimates a dynamic nonlinear panel model with correlated random effects. The main finding is that a substantial portion of the link between poverty and separation appears to derive spuriously from unobserved factors that affect both outcomes. However, even after accounting for dynamics and unobserved heterogeneity, there remains a strong association between the arrival of poverty and marital separation. The results appear to suggest that poverty and separation often arrive concurrently, though that finding might stem from the relatively course periodicity of large micro household panel surveys. To that end, this paper presents a case for household surveys to adopt more frequent recordings of information.