Tuesday, 9 December 2008

Freakonomics: Coolness inequality declines

Over the last forty years, the American General Coolness Survey has asked subjects the following question:

"Taken all together, how do you think you shape up these days? Would you say that you are: very cool, pretty cool, or not too cool?"

In a new paper published today I demonstrate some dramatic results: coolness inequality has diminished hugely in the four decades covered by the survey. Simultaneously, overall coolness has increased substantially.

In 1968, 5% of people reported themselves to be "very cool". A further 10% were "pretty cool" and 60% "not too cool" (a substantial proportion of respondents did not answer this question).

By the 1970s, coolness had increased but so had inequality. 40% were very cool, 28% pretty cool and 30% not too cool. In the late 80s and early 90s, however, overall coolness fell and society became more equal - with a majority reporting themselves "not too cool" again.

By the time of the 2008 survey, coolness had definitely got its mojo back and equality has returned. The results today are: 31% very cool, 50% pretty cool and 14% not too cool. Notably the number of non-respondents has diminished substantially to only 5%.

Comparing two different measures of inequality - the standard deviation of the distribution of the survey results and the number of people reporting "pretty cool" - inequality is at a historic low. The question is, why would this be? Especially since other measures traditionally correlated with coolness - such as skin colour - are at historically high levels of inequality in an increasingly diverse American society.

We hypothesise that...no, it's no use, I can't go on.

These results are, of course, nonsense. There is no coolness survey. If there were, its results would be a referendum on the evolving usage of the word "cool" and not on any meaningful measure of how cool people really are.

Surely the Freakonomics guys - Justin Wolfers and Betsey Stevenson, who wrote a paper on this in August - can see that? Their paper is on inequality of happiness and not coolness. But really. How can this possibly be a meaningful result?

Outside of major exogenous events (let's say the Second World War, and even then, only the beginning and end of it) people's reported happiness is much more a result of their own attitude to life and the word 'happy' than anything to do with social conditions.

If there were some objective, repeatable way to measure happiness that might give useful results. But using self-reported happiness to measure actual happiness is no more valid than using self-reported sexual activity to measure the birth rate.

Admittedly there may be some significance in the results relating to race or sexual orientation, but the authors conclude that the largest component of the effect is independent of this.

Yesterday Justin wrote a gently mocking article about someone else's results on happiness. Fair enough. I am sure he can take a bit of it too.

No comments: