Abstract
In 2020, close to 46,000 people died from
suicide in the U.S. Globally, rates of suicide have declined some in the last
20 years − not so for the U.S. In the last two decades, the U.S. has seen more
than a 25% increase in age-adjusted suicide rates. Recent reviews of the sociologically grounded
ecological studies of suicide find jurisdictional divorce and unemployment
rates to be key suicide risk factors.
However, a new vein of this literature is beginning to scrutinize
long-established ecological links based on faulty statistical methodologies
that previously ignored variable non-stationarity and the lack of series
cointegration. The purpose herein is to fully dissect the tenuous ecological
relationships between U.S. annual divorce rates, unemployment rates and suicide
rates using a 53 year non-stationary panel of all 50 states and the District of
Columbia. Results suggest no statistically imperious association, short-run or
long-run, between suicide rates in the U.S. and the long-established risk
factors divorce and unemployment rates. Implications of this dissection
advocate for a shift in research focus to the individual − controlling for
idiosyncratic specific-effects and key social processes.
JEL classification numbers: B55, C18, C49.
Keywords: Suicide rates,
Cointegration, Ecological analysis, Non-stationary panel.