After analyzing several comprehensive data sets, a team of U.K. and U.S. researchers have found a dark paradox among U.S. states: those that have the highest individual reports of subjective well-being (i.e. happiness) also tend to have the highest suicide rates.

The paper (PDF here) draws its data from the Behavioral Response Factor Surveillance System, which measured subjective well-being among 1.3 million Americans, and each state's published suicide rate, among other sources. After adjusting for population differences, the researchers found a "very strong correlation" between the states that have been classified as the happiest populations and the published suicide rates (they also found similar correlations between the happiest countries and high suicide rates).

Here are some of examples teased out of the press release of the happy states that were found to have high suicide rates:

  • Utah: ranked 1st in life-satisfaction and has the 9th highest suicide rate.
  • Hawaii: 2nd in life-satisfaction and has the 5th highest suicide rate.

Meanwhile, states classified as unhappier tended to have lower suicide rates, examples:

  • New Jersey: 47th in life-satisfaction, and 47th in suicide rates.
  • New York: ranked 45th in life-satisfaction, and has the lowest suicide rate of all states.

The full list of states and their adjusted suicide and satisfaction rankings are included in the full PDF version of the study.

But what's the reason behind this fascinating finding? The research team explained their theory in the Discussion section this way:

Discontented people in a happy place may feel particularly harshly treated by life. Those dark contrasts may in turn increase the risk of suicide...If humans are subject to mood swings, the lows of life may thus be most tolerable in an environment in which other humans are unhappy.  Whether such relative comparisons work by producing discord due to unmet aspirations...or reflect a real inability to integrate into broader society and gain access to key supports, remains to be understood.

The research has been peer-reviewed and accepted for publication in the Journal of Economic Behavior & Organization, but it is not yet in its final form.