Actual versus ‘natural’ rates of suicide: Evidence from the USA

Abstract

This paper employs a high dimensional variable selection technique to select a subset of suicide determinants from 167 potential factors, which are then used to estimate ‘natural’ suicide rates for US states by least squares dummy variables. Over the period 2005–2017, all states are found to have a non-zero and positive natural suicide rate, below their respective actual average rate. Higher actual rates suggest deterioration in socioeconomic conditions, inaccessible and unaffordable mental health care for certain sections of the population and inadequate implementation of measures to identify and reduce suicidal mortality. Evidence-informed policies aiming for zero suicide target could draw inspiration from exemplar states to direct resources towards states with greater relative differences between actual average and natural suicide rates.