Tuesday, June 1, 2010

Happiness May Come With Age, Study Says

New York Times article "Happiness May Come With Age, Study Says" (May 31st, 2010) speaks on research involving, as its main variable of interest, surveys concerning the states of mind of individuals of different ages.

A large Gallup poll has found that by almost any measure, people get happier as they get older, and researchers are not sure why.

This may be considered a more subjective and methodological correction, but Corrections considers such research to generally be invalid due to heterogeneity in responses by class. For example, wealthier people may simply respond to questions differently than poor people. Not because they are "happier," but due to some difference in type--rich people are likely different from poor people. Similarly, older people may respond to questions about happiness different from younger people.

There may be a saving grace for happiness research, exogeneous variation. For example, we might survey 1,000 individuals and discern their subjective happiness responses. We may then observe them a few months later--perhaps one has won the lottery or inherited some large sum of money. In this case, we can difference out his previous responses to his new responses, and attributed the difference to the large exogenous shock in wealth.

However, we can never get any such variation in age. We cannot take a young person and suddenly put them in an old person's body. The research, as Corrections sees it, is largely bound by the untenable structural assumption that old people's responses mean the same thing as young people's responses.

1 comment:

  1. Saying that "by almost any measure" old people are happier...lends a wholly unjustified air of uncontroversiality to this claim. _All_ of our happiness yardsticks suffer from the same fatal flaw, and so the fact that almost any measure is consistent with the claim...demonstrates very little. When almost all physicists agree on something, it should give us real confidence in its correctness. Gallup has not earned this species of confidence here, although the NYT would have us believe it.

    By the way, wouldn't it be great if, instead of throwing suggestive statistics or selected expert opinions at us, the media regularly polled experts for their probability of something being correct? Here, for example, a selection of economists and psychologists would immediately reveal that there was far less certainty to the claim that old people are happier.