Poisson analysis of suicide in Japan using municipal data

Abstract

This article investigates the nexus between suicide and socio-economic factors in Japan using municipality-level data using Poisson-based regression methods. We demonstrate the advantage of this method over the OLS approach and show that our method yields reasonable results compared to OLS. Our results show that economic difficulty and social isolation are significantly associated with suicide rate, and accessibility to hospitals can reduce suicide rates.

Keywords:

• suicide
• Poisson regression
• count data
• socioeconomic factors

JEL Classification:

• I12
• I31
• C25

Notes

¹ Because the logarithm of zero is undefined, we add 10⁻⁷ to the suicide rates for municipalities with no suicides.