We canʹt get no (life) satisfaction? Comment on Oswald and Wu (2010)

Oswald and Wu (2010; Science) recently reported life satisfaction ranks for residents of the 50 U.S. states. Their rankings were framed as measures of “well-being,” but were derived from responses to only a single survey item (“In general, how satisfied are you with your life?”). Here we contrast Oswald and Wuʹs subjective rankings with our (Pesta, McDaniel & Bertsch, 2010) objective measures of state well-being. Strangely, our global index of well-being correlates −.43 with Oswald and Wuʹs life satisfaction ranks (intelligence, itself, correlates −.33 with these ranks). We argue that Oswald and Wuʹs rankings are misleading for three reasons. First, their massive sample size created statistically significant effects with little practical value. Second, life satisfaction and well-being are different constructs. Third, Oswald and Wu adjusted their rankings for personal income differences across survey respondents. We see this as analogous to controlling for rainfall when attempting to measure precipitation. Accordingly, after re-adjusting life satisfaction by income, we show that Oswald and Wuʹs ranks fail to correlate with any objective sub-domain of well-being.