A few days ago, Robin posted on the Edge’s annual question, which this year is about the changing of minds. One of the participants (a social scientist who undoubtedly knows lots) is Daniel Kahneman. It’s impossible to overstate Kahneman’s eminence. He’s unquestionably one of a handful of top researchers ever, and arguably the most important yet alive, on the subjects that make up the theme of this very blog. In addition to being one of the inventors of the "heuristics and biases" research program, as well as prospect theory, he also won the 2002 "Nobel Prize" in economics.
The finding you refer to is within a society, based on the relative income effect.In Japan in the late 1940s, there were a lot of people on the verge of famine,especially right after 1945. Still, not much difference in happiness with 1974.
I seem to recall reading that absolute income does increase happiness but only to a point; once you're out of poverty and no longer have to worry about where your next meal is coming from (figuratively speaking), getting richer doesn't help happiness very much. In "poor countries" there are a lot more people who do worry about whether they're going to survive the next few months, which seems as though it would have a negative influence on happiness, treadmill or no treadmill.
I have much respect for Kahnemann, but for some reason he is pushinga seriously distorted line that is based on one of the most obvious andflagrant of errors in statistics, confounding cross-section with time-seriesresults. He has been supporting the arguments of recent distributors of theGallup survey results around the world that find a positive relation betweenincome and reported happiness (and satisfaction, they are strongly correlated),as supposedly showing that income does cause happiness after all.
This has been trumpeted in lots of media and lots of blogs, including MarginalRevolution recently, as supposedly disproving the Easterlin Hypothesis, thatrising income does not necessary lead to rising happiness, when it is the incomeof everybody in a society. People get happier when their incomes rise relativeto that of others within their own society. To quote Robert Frank: "The happyman is he whose wife's sister's husband makes less money than he does."
So, what is the deal here? Easterlin's findings are based on extensive studiesof time-series data, starting with a study of Japan in the 1970s, where income hadsoared at double digit rates for a couple of decades. No increase in overallhappiness, although the rich tended to be happier than the poor. This has beenreplicated widely, including in the US, where reported happiness actually maxedout in 1956. In short, the current set of relations, with higher income countriesbeing happier than poorer ones, probably held back a half century ago, with notmuch change in those rankings, a kind of global version of the relative incomeeffect that Easterlin found within societies, the rich are happier than the poor,but only because they are above the poor. I have suggested elsewhere that theremay be other factors at work in these findings as well, such as pride in nationalpower, which is correlated strongly with income, and so forth.
I cannot resist adding a really unpleasant point in my view. I have been given tounderstand by high level individuals that the Gallup organization is not making thisdata set generally available. It seems to be available to Kahnemann and a handful ofothers, but not to other reputable and respected researchers who have requested accessto it. Given the huge publicity that has accompanied this data set, much of misguidedas I have just pointed out, I find this peculiar and unscientific. Indeed, I considerit a travesty and downright reprehensible. I find it unfortunate that the very respectedProfessor Kahnemann appears to be partaking of this regrettable effort to suppressscientific reseearch.
Back in 1990, I had the honor of a taking a class from Amos Tversky, Kahneman's frequent research partner who had the misfortune to pass away before the 2002 Nobel was awarded.
His observation, which I assume is shared by Khaneman, is that cognitive biases are similar to perceptual biases. Just because you know about them, doesn't mean you can easily correct for them. Psychologists know that distance perception is affected by all sorts of factors such as atmospheric clarity, presence of reference objects, etc. That doesn't mean a psychologist is any better judge of distance than anyone else. It's just the way everyone's brain works.
You need a ruler to do better. The scientific method is like that ruler for cognitive bias (though imperfect). This single idea has always stuck with me and is probably the most significant factor in my desire to read this blog.
Oh, sweet irony.
Doug S.,
The finding you refer to is within a society, based on the relative income effect.In Japan in the late 1940s, there were a lot of people on the verge of famine,especially right after 1945. Still, not much difference in happiness with 1974.
I seem to recall reading that absolute income does increase happiness but only to a point; once you're out of poverty and no longer have to worry about where your next meal is coming from (figuratively speaking), getting richer doesn't help happiness very much. In "poor countries" there are a lot more people who do worry about whether they're going to survive the next few months, which seems as though it would have a negative influence on happiness, treadmill or no treadmill.
I have much respect for Kahnemann, but for some reason he is pushinga seriously distorted line that is based on one of the most obvious andflagrant of errors in statistics, confounding cross-section with time-seriesresults. He has been supporting the arguments of recent distributors of theGallup survey results around the world that find a positive relation betweenincome and reported happiness (and satisfaction, they are strongly correlated),as supposedly showing that income does cause happiness after all.
This has been trumpeted in lots of media and lots of blogs, including MarginalRevolution recently, as supposedly disproving the Easterlin Hypothesis, thatrising income does not necessary lead to rising happiness, when it is the incomeof everybody in a society. People get happier when their incomes rise relativeto that of others within their own society. To quote Robert Frank: "The happyman is he whose wife's sister's husband makes less money than he does."
So, what is the deal here? Easterlin's findings are based on extensive studiesof time-series data, starting with a study of Japan in the 1970s, where income hadsoared at double digit rates for a couple of decades. No increase in overallhappiness, although the rich tended to be happier than the poor. This has beenreplicated widely, including in the US, where reported happiness actually maxedout in 1956. In short, the current set of relations, with higher income countriesbeing happier than poorer ones, probably held back a half century ago, with notmuch change in those rankings, a kind of global version of the relative incomeeffect that Easterlin found within societies, the rich are happier than the poor,but only because they are above the poor. I have suggested elsewhere that theremay be other factors at work in these findings as well, such as pride in nationalpower, which is correlated strongly with income, and so forth.
I cannot resist adding a really unpleasant point in my view. I have been given tounderstand by high level individuals that the Gallup organization is not making thisdata set generally available. It seems to be available to Kahnemann and a handful ofothers, but not to other reputable and respected researchers who have requested accessto it. Given the huge publicity that has accompanied this data set, much of misguidedas I have just pointed out, I find this peculiar and unscientific. Indeed, I considerit a travesty and downright reprehensible. I find it unfortunate that the very respectedProfessor Kahnemann appears to be partaking of this regrettable effort to suppressscientific reseearch.
The phrase "renounces result" in the title of this post is inaccurate and misleading.
Back in 1990, I had the honor of a taking a class from Amos Tversky, Kahneman's frequent research partner who had the misfortune to pass away before the 2002 Nobel was awarded.
His observation, which I assume is shared by Khaneman, is that cognitive biases are similar to perceptual biases. Just because you know about them, doesn't mean you can easily correct for them. Psychologists know that distance perception is affected by all sorts of factors such as atmospheric clarity, presence of reference objects, etc. That doesn't mean a psychologist is any better judge of distance than anyone else. It's just the way everyone's brain works.
You need a ruler to do better. The scientific method is like that ruler for cognitive bias (though imperfect). This single idea has always stuck with me and is probably the most significant factor in my desire to read this blog.
Um... Isn't misinterpreting data what social science is all about?
If this can happen to Daniel Kahneman, it can happen to anyone.
Your emphasis is backwards: if this can happen to anyone, it can happen to Daniel Kahneman.