Justin Wolfers kindly responded to my recent post. His thoughts, reprinted with his permission.
Interesting, fun, provocative, and well written. Your math looks to be right to me.
Some thoughts:
1. I’m not sure that controlling for confounds necessarily would reduce the causal effect of income. (My prior is similar to yours, but I haven’t thought through the issue enough).
2. I actually think 0.35 is pretty big. Said another way, it’s big enough that it can explain why people in Burundi are at 3.5/10 on a happiness scale, and Americans are at 8/10. My interpretation is that big gaps in happiness are easily explained by big gaps in income. So why do we interpret things differently?
a. I think raising happiness by a standard deviation is huge. Basically I see incredibly miserable people and incredibly happy people all around me. Moreover, if you think there’s measurement error in happiness, then the standard deviation in measured happiness is even bigger (and you are talking about raising someone’s measured happiness by one measured standard deviation).
b. The other way of saying this is that it doesn’t matter that the effect of income on happiness is “small”: if there exist massive disparatives in income, then a small coefficient can have a big effect. And I think there exist massive disparaties in income (and these largely explain the massive disparaties in happiness).
3. In my careful moments, I see my data as a shocking refutation of whether money has no effect on happiness. In my less-guarded moments, I see it as a refutation of whether money has little effect on happiness.
4. Your final thought experiment (“an extra $820,585” is clever). But it is just as interesting of a thought experiment in the opposite direction: If I raised your income from $3,000 (roughly the average income in the US at the turn of the century) to $50k, I would increase your happiness by one standard deviation. How many people would move from being “depressed” to OK? (If, for argument’s sake, depressed = bottom 5% in 1900, then depressed = % of people with z-score less than 1.6. Shift that distribution 1 standard deviation to the right, and the proportion who are depressed = % of people with z-score+1 less than 1.6, which is 0.5%. So we would nearly eliminate depression.)
But, as always, really interesting stuff, and great fodder for the blog.
One more thought: The claim made by some that money matters less than we think and hence we should focus less on it, always struck me as important, and plausible, and one that my evidence is silent on.
J.
READER COMMENTS
Tom West
Feb 28 2014 at 7:59am
Whether money matters a “lot” is simply a matter of context. Obviously money changes are pretty meaningful for development from an agrarian economy to an industrial one.
But for developed world policy debates on economic measures that cause social unhappiness in return for economic gains, the happiness curve is pretty darn flat in return for the unhappiness that change and the lack of security causes.
I’d say for most of the people here who would like to use the results as ammunition in a policy debate, the net is money doesn’t matter all that much to happiness compared to the other effects of policy.
AS
Feb 28 2014 at 8:38am
Correlation does not imply causation. Also can’t assume the relationship is perfectly log-linear at all points. Also can’t assume what works in aggregate will work for individuals, who each their own unique “happiness-income” curve, which may be non-linear, non-continuous, and non-differential. Individual results may vary.
Eric Falkenstein
Feb 28 2014 at 9:12am
I don’t think he even wants to understand the alternative hypotheses, because he’s been writing about it conspicuously for a while and continues to offer up caricatures in the manner of a politician. There’s no large group in the ‘money does not matter’ camp that he presents as his main refutation. The alternatives are first, the Easterlin position is that 1) after a modest amount of welfare has been achieved (say $50k in modern USD) money does not matter or 2) that after a modest amount (say $50k) only relative income matters.
It’s important because the utility function is the premise in so many models, and if it doesn’t have the property u'(w) is continually positive, u”(w) is negative, where u() is the function of happiness/utility and w is the individual’s wealth/consumption, then a lot of models are irrelevant. Note that if you also want to hit the stylized fact that interest rates are pretty stable over the past 150 years, you need a very special functional form that isn’t even being considered (the CRRA form w^(1-a)/(1-a))
David R. Henderson
Feb 28 2014 at 10:00am
@Eric Falkenstein,
You’ve got me curious. What alternative hypothesis or hypotheses would you suggest?
Eric Falkenstein
Feb 28 2014 at 10:25am
The latter, relative income after a certain level. I wrote a book, The Missing Risk Premium, on that argument (available in paperback and Kindle!).
Justin Wolfers
Feb 28 2014 at 10:29am
Eric: Go back and read Easterlin, and you’ll quickly learn that you’ve mis-characterized his position. And it’s not hard to find many others who take a similar position, either.
ggc
Feb 28 2014 at 2:43pm
[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]
Eric Falkenstein
Feb 28 2014 at 8:22pm
Justin: In what way have I mischaracterized Easterlin’s position? My minority position is the latter, the Easterlin position, to my mind, is that after a certain level, income does not matter. What do you think Easterlin’s position is?
Comments are closed.