Understanding Unobserved Propensities of Suicide in the United States : A Hierarchial Model with Spatially Correlated Random Effects

This paper investigates the underlying causes of suicide. In contrast to previous literature, we use data from the United States at the county level. Our primary methodology is a two-level Bayesian hierarchical model with spatially correlated random effects. Our results show that the significant effects of observable factors on suicides found by earlier research may partially stem from excluding small area effects and time trends. Without controlling for these area and time effects, the true contribution of unobserved propensities and time trends can be hidden within observable factors. Most importantly, we find that a lot can be learned from unobserved yet persistent propensity toward suicide captured by the spatially correlated county specific random effects. We argue that resources should be allocated to counties with high suicide rates, but also counties with low raw suicide rates but high unobserved propensities of suicide