Jonathan Rottenberg, PhD

Jonathan Rottenberg Ph.D.

Charting the Depths

State Happiness Rankings Reveal Americans' Happiness Insecurity

The unbearable misery of being New York.

Posted Dec 22, 2009

New data reveal that Louisiana, Hawaii, and Florida are the happiest states whereas Michigan, Connecticut, and New York are the most miserable.

The media have glommed on to the state-by-state happiness rankings that are incidental findings in a new study to be published in Friday's Science by economists Andrew J. Oswald of the University of Warwick in England and Stephen Wu of Hamilton College in Clinton, N.Y

There is gloating from the winners. Perhaps not surprisingly, the strongest reactions are from the "losers." See strong gut-level reactions from New York, here here here and here  

Before there is a mass exodus from Utica to Baton Rouge, a few cautions are in order.

1. Oswald and Wu transformed the raw data to produce their rankings. The Science article was a reanalysis of CDC data from 1.3 million Americans collected between 2005 and 2008. In this reanalysis, the authors focused on one of the questions, which was an item about overall life satisfaction, rated on a 1 to 4 scale. Importantly, the authors did not simply compare average responses on this item between people in different states. Rather the authors tried to statistically control for a variety of other variables that differed by state and then "correct" the ratings for these incidental variables. One of co-authors argued for this transformation, "In other words, all anyone has been able to do is to report the averages state-by-state, and the problem with doing that is you're not comparing apples with apples because the people who live in New York City are nothing like the individuals living in Montana." While there is nothing unethical about what the authors have done, the media reports on these rankings generally omit the fact that these ranking results differ from those that were reported in earlier analyses because of variations in the statistical procedures.

2. Differences between states in happiness are very small in absolute magnitude. Differences between the happiest and unhappiest states are on the order of .1 of a point on a 4 point scale. While focusing on the most unhappy states might make for good news copy, are any states really unhappy? It is easy to lose sight of the fact that even the most lowly Uticans report being very satisfied with life on average. Americans as a whole report remarkably high well being, with the mean well being about 3.4 on a 4 point scale (which is midway between satisfied and very satisfied with life).

3. The happiness rankings were actually not the point of the study. The authors' main analyses and main point was that individual subjective ratings of life satisfaction correspond with other more objective quality of life data, such as climate, traffic, taxes etc. This main point of the research has been largely ignored in all of the news coverage.

My takeaway: While Oswald and Wu's study deserves a careful look, the media presentation of it is grossly oversimplied. The intense coverage of this research, like that of previous state rankings, plays strongly on Americans' happiness insecurity. This happiness insecurity is especially pronounced this time of year. With Americans already asking, "am I in a happy state?," media attention to happiness rankings says less about the science of well being than the press of a cultural anxiety