Saturday, March 23, 2019

Alan Krueger And Happiness

It took awhile for me to remember after his apparent suicide that the late Alan Krueger was the coauthor of what I consider to be the best paper published on happiness economics, "Develpments in the Measurement of Subjective Well-Being," Journal of Economic Perspectives, 2006, 20(1), 2-24, .  (I apologize if that link is no good.)

This paper was coauthored with is Princeton colleague, Nobel-Prize winning psychologist Daniel Kahneman.  What stood out was that this was as well done an empirical study of this difficult topic as I have ever seen, and above all in terms of his professional accomplishments, Krueger's abillity to carry out very well done empirical studies stands above all else, and this paper with Kahneman certainly fits into that category.

Happiness research, or more properly research on "subjective well-being" as their paper titel put it, has been a controversial mess for some time.  The journals I have edited have been major outlets for this research, so I have seen lots of papers on it.  Not only that, but I have seen the referee reports on those papers, and this area of research has somehow attracted a lot of unhappy people taking out their unhappiness on papers that they referee.  This is not true  of all people in this field, and  I shall especially single out Richard  Easterlin, the fathr  of the field, who retired from USC last year at age 92 and deserves a Nobel Prize.  But many others in the field are not as happy as he has always been.

There are three main sources of data for these studies: cross-sectional cross-national, time-series within a naton, and panel studies of individuals over time (hence, also time-series).  I have listed them in increasing order of scientific credibility. Even so, these least credible cross-section studies get lots of media attention. I mean, wow! the UN just released its latest World Happiness Report, and, wow! Finland is Number One, despite having a way-above average suicide rate. But, hey, the suicides are not showing up in the survey.  And wow! the US has continued to decline, now at #19 or thereabouts, which is probably meaningful given that the US is the only high income nation in the world with a declining life  expectancy.

The middle of these, time-seires inside a nation , does not have the cross-cultural problems that the previous method has, but it has the problem that people die over time so the sample is different people at different time periods. In any case, it  was these data sets on Japan and the US that led Dick Easterlin to enunciate his "Easterlin Paradox," (no, he did not name it that), that while a nation with rising real income per capita always showed that at any point in time higher income people were happier than poorer people, over time overall happiness levels did not rise.  Indeed, for the US they have gradually declined, with 1956 being the year of maximum recorded happiness in the US, even though real incomes per capita have risen quite substantially since then.

Which brings us to the final method, the one I take the most serously, panel studies of individuals.  Yes, this is what Kahneman and Krueger did with a set of 500 women in Columbus, Ohio over aa period of time, monitoring almost minute by minute, very closely, the most detailed and credible study of this sort I have seen.  They found that these women preferred being with their friends more than with their husbands or lovers, and those more than their children.  The only people they were less happy being around than being alone with were their bosses, oh, big surprise.  In terms of more specific situations, they most liked "intimate relations," so those spouses and lovers got their due, with their least favorite social situation being stuck in traffic jams, which suggests to me that having people commute with friends is a good idea.

So, there we have it. Alan Krueger was a serious researcher in happiness economics.  It has also come out that in reent years he has also studies effects of the opiod crisis and also mental health issues more broadly. I have not read these more recent papers, but I am confident that he applied the high standards of professional empirical research to these topics that he did to his exceptionally outstanding paper with Daniel Kahneman on happiness.  Pretty obviously now in hindsight that he was concerned with these topics suggets that indeed this interest may have been driven by a deeper personal concern.  Pretty clearly, while he may have coauthored the best paper ever written on happiness, in the end he was not himself all that happy.

RIP, Alan.

Barkley Rosser


marcel proust said...

I have listed them in decreasing, order of scientific credibility.

Based on the rest of the post, I believe that decreasing s/b increasing. said...

Good point, marcel. Will fix that.

Fika Fikalestari said...

Magic is something you make