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.

This has been trumpeted in lots of media and lots of blogs, including Marginal
Revolution recently, as supposedly disproving the Easterlin Hypothesis, that
rising income does not necessary lead to rising happiness, when it is the income
of everybody in a society. People get happier when their incomes rise relative
to that of others within their own society. To quote Robert Frank: “The happy
man 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 studies
of time-series data, starting with a study of Japan in the 1970s, where income had
soared at double digit rates for a couple of decades. No increase in overall
happiness, although the rich tended to be happier than the poor. This has been
replicated widely, including in the US, where reported happiness actually maxed
out in 1956. In short, the current set of relations, with higher income countries
being happier than poorer ones, probably held back a half century ago, with not
much change in those rankings, a kind of global version of the relative income
effect 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 there
may be other factors at work in these findings as well, such as pride in national
power, which is correlated strongly with income, and so forth.

I cannot resist adding a really unpleasant point in my view. I have been given to
understand by high level individuals that the Gallup organization is not making this
data set generally available. It seems to be available to Kahnemann and a handful of
others, but not to other reputable and respected researchers who have requested access
to it. Given the huge publicity that has accompanied this data set, much of misguided
as I have just pointed out, I find this peculiar and unscientific. Indeed, I consider
it a travesty and downright reprehensible. I find it unfortunate that the very respected
Professor Kahnemann appears to be partaking of this regrettable effort to suppress
scientific reseearch.

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.

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.