« The Mobi | Main | Russ Roberts asks about Beethoven's slow movements »
Assorted links
1. Tim Harford and Dan Ariely, mano a mano on Amazon
3. Money does buy happiness; the ongoing collapse of the Easterlin paradox
4. New video game detects your brain waves
Posted by Tyler Cowen on April 16, 2008 at 12:21 PM in Web/Tech | Permalink
Comments
The problem is that while large differences in wealth, especially as one climbs out of poverty, matter much for happiness ... there are surely declining returns.
This matters when individuals (or nations) are "wealthy" and must decide where to next apply their efforts.
Personally, I tend to knock off early and go mountain biking.
Posted by: odograph at Apr 16, 2008 1:42:22 PM
Maybe I didn't complete that thought. My "knock off early and mountain bike" might conflict with someone who says "let's sell off this park and build more homes." He might argue that total happiness is more improved by the increased wealth than by the recreational services.
... my point is that depends ... on how much wealth you have, or need.
Posted by: odograph at Apr 16, 2008 1:46:02 PM
The take-their-money-away-and-watch-them-scream test is suggestive.
Posted by: dearieme at Apr 16, 2008 2:03:57 PM
Still paradoxical: Venezuela's satisfaction is slightly higher than the U.S.; however, the U.S. GDP per capita is more than 4x that of Venezuela. You can do this type of comparison between many countries.
Posted by: at Apr 16, 2008 2:16:39 PM
How does globalization affect these kinds of cross-country comparisons of satisfaction? Might it be that the happiness reports of people in poor countries are impacted more by their relative income in a global economy than by their income relative to their fellow citizens? If so, then Easterlin's claims about the greater impact of relative than absolute wealth would not be undermined by showing that people in poor countries are less happy than people in rich countries. The time progression data would be more interesting (as Easterlin suggests).
Posted by: mike at Apr 16, 2008 2:32:17 PM
"The take-their-money-away-and-watch-them-scream test is suggestive."
But that's, presumably, an "all else being equal" test. Sure, all else being equal, more money is good. Real world questions, like money versus stable tuna populations, are harder.
Posted by: odograph at Apr 16, 2008 3:40:57 PM
Oh brother. This Stevenson-Wolfers paper is a big bunch of eyewash,
and for reasons that have been elucidated here earlier. "Fraglile"
time series"? Indeed. Last time I checked all the people crowing
about how much better the US is than Europe and Japan were pointing
to its higher rate of economic growth more recently, especially since
1990, since when Japan particularly has been stagnant. That they
are getting happier is evidence of income buying happiness? Nope.
The bottom line of this latest bunch of surveys is that the distribution
of happiness around the world has not changed much over the last several
decades. The countries that are happier now were happier back then,
including some poorer countries, notably in Central America, as well as
some richer, more egalitarian countries such as Denmark. There has been
little change over time in individual countries, but the outliers like
Central America aside, on average richer countries seem to be happier,
or say they are, than poorer ones.
An explanation besides that of Easterlin is that there is a globally
relative effect that shows up in such things as national power. If
you are poorer, you are also more likely to be dominated and ruled by
and humiliated by rich and powerful outsiders. Likely to make you
less happy. Duh. It is simply the relative income part of the Easterlin
paradox writ large and global.
No disproof here at all, indeed, confirmation if anything.
Posted by: Barkley Rosser at Apr 16, 2008 3:41:51 PM
Regarding the "take their money away and watch them scream," it is well
known that there is a huge asymmetry between gains and losses. People
very much dislike losing things, any things, including money, especially
when it is unexpected (most people do not mind their loss of pay when
they retire). This shows up in the much-studied asymmetry between the
willingness to pay for improvements (as in the environment) versus the
usually much higher (often an order of magnitude higher) demand for
compensation for losing as much. So, this proves mighty little.
Posted by: Barkley Rosser at Apr 16, 2008 3:53:47 PM
The explanation of the Easterlin paradox that I've seen emphasizes relative wealth. But the Stevenson-Wolfers paper seems to suggest that both absolute and relative wealth are important. If both absolute and relative wealth are arguments in the utility function shaped by Harry Markowitz, one gets an operational explanation of people who both pay unfair prices to assume risk and unfair prices to avoid risk:
http://ideas.repec.org/p/bsu/wpaper/199602.html
Posted by: indiana jim at Apr 16, 2008 4:48:04 PM
I have a cousin-once-removed named Seven, and yes she is a Utah Mormon. The website linked to had Seven in the boy column, however. Truthfully it does sound like a boy name.
Posted by: incrediblub at Apr 16, 2008 7:01:47 PM
As a reasonable approximation, happiness questions are measuring *personality*.
Because personality is stable over time, then so is happiness.
Because personality questionnaires are self-rated, they vary between cultures in unpredictable and (at present) non-explainable ways.
Until we get some objective personality measurements (probably by discovering genetic correlates of personality), this type of happiness research is a misleading waste of time - from a scientific perspective.
However, this type of happiness research (ie. research showing a lack of correlation between 'happiness' (actually personality) and GDP) has a hidden agenda of anti-modernization.
For example, happiness reseach is used to ask 'why bother increasing GDP if it doesn't make you 'happier''.
Or happiness research has the hidden agenda of advocating leftish command economy tactics to replace markets - suggesting that happiness surveys leading to bureacratic solutions should replace consumer choice as a outcome measure.
This kind of 'happiness' research is popular in the UK public services where the state-run general practice (family physician) service is being reorganized around patient survey data; and where a government national student satisfaction survey is being built-into the state-run university system.
In other words, in this happiness-research world; the idea is that people should say they want in surveys, and using this data bureaucrats will decide what people should get; rather than people being allowed to do what they want from the choice offered by markets.
Happiness research as a basis for public policy is mostly a new name for a very old idea. It is almost precisely the kind of system used in the later Soviet Union - when consumer survey data was used to decide what people 'really' wanted - the idea being that this was a superior to market expressed preferences.
So mainstream 'happiness research' (when considered as merely one example of consumer satisfaction survey data) is scientifically misguided/ trivial; but politically of core importance to those who advocate centralized command economy modes of organization.
Posted by: BGC at Apr 17, 2008 1:55:34 AM
at's comment is a perfect example of how, focusing on an outlier, you can completely miss what a data set is telling you.
Easterlin is right to point out that the positive correlation within nations could represent a relative income effect, and that the positive correlation among nations could represent an underlying shared cause (perhaps culture, or, as Barkley suggests above, national power). But to claim, in the face of that NYT graphic, that the correlation doesn't exist or is refuted by the case of Venezuela or some other outlier, is ridiculous.
Posted by: David Wright at Apr 17, 2008 4:47:26 AM
"But to claim, in the face of that NYT graphic, that the correlation doesn't exist or is refuted by the case of Venezuela or some other outlier, is ridiculous."
I'll buy that completely David; but I no of no evidence that refutes the hypothesis that BOTH absolute and relative wealth contribute to happiness.
Posted by: indiana jim at Apr 17, 2008 10:06:23 AM
Sorry about the typo:
Should read: I know of no evidence that refutes the hypothesis that BOTH absolute and relative wealth contribute to happiness.
Posted by: indiana jim at Apr 17, 2008 10:07:42 AM
indiana jim,
Well, the original work of Easterlin and large chunks of the subsequent literature
are hardly supportive of the absolute wealth argument. Stevenson and Wolfers have
to strain at things to get some countries to maybe fit, and they are odd ducks.
Let us focus on the US. What year was recorded happiness at its peak? 1956.
Were we absolutely wealthier then than now? I doubt it. Of course that was a year
of reasonable peace and prosperity after many that were neither or only one. It was
also the year that preceded the largest number of births of any in US history, and
Kahnemann and some others have shown that "intimate relations" are the activity on
average giving at least women in Columbus the greatest happiness, with commuting at
the bottom. So, given all those births in 1957, we know what was going on in 1956...
Posted by: Barkley Rosser at Apr 17, 2008 5:45:01 PM
In the current Journal of Economic Literature, Clark, Frijters, and Shields use similar modern data on happiness and utility to support the Easterlin paradox.
The authors propose that there is a spike in happiness during initial increases in national income (especially in developing nations like Venezula) but the happiness eventually flattens as the nations become more developed and the nations around them become more developed.
Posted by: Patrick at Apr 17, 2008 5:57:50 PM
8th hightest female name in 2004: Brooklyn.
Really.
Really?
Posted by: Adam at Apr 17, 2008 10:28:26 PM
8th hightest female name in 2004: Brooklyn.
Really.
Really?
Posted by: Adam at Apr 17, 2008 10:29:03 PM
Barkley,
"we know what was going on in 1956..."
Good point, this poses an interesting challenge to the assumption of ceteris paribus.
Posted by: indiana jim at Apr 18, 2008 7:30:14 AM
Barkley,
Following up on my last post, maybe the proper way to conduct a "happiness survey" is to include a question or series of questions about the activities of the respondent over the days and/or hours prior to the filing of the survey.
Posted by: indiana jim at Apr 18, 2008 10:26:13 AM
i.j.,
Could be. Needless to say, the whole happiness thing is now a major cottage
industry. One can find some correlations with rising income or wealth and
rising happiness, with the most dramatic examples being in some recent years
in the transition economies. Of course, these increases have followed massive
declines in happiness that occurred as they fell into major economic declines
earlier. But then, one can say that those correlated declines show that absolute
income/wealth do matter.
I am not going to say that absolute wealth is zero. What does seem to be true
is that the relative income/wealth effect is far stronger than any absolute such
effect, if the latter are even greater than zero at all in a resilient way.
Also, of course, the only reliable studies, to the extent that there are any
such things in this fuzzy biz, are the panel studies of specific people. The
cross-section stuff is mostly unreliable for the reason that Easterlin pointed
out, all those cultural differences. In some places it is not considered proper
to admit that you might be unhappy (high suicide rate Nordics who go "ja ja ja"
when asked if they are happy) while in some others one is viewed as an idiotic
fool if one is not complaining all the time about everything (Russians and some
other notoriously well known groups).
Posted by: Barkley Rosser at Apr 18, 2008 10:51:57 AM
Barkley,
I hypothesize, as in a JEBO publication you know about, that if one's absolute wealth rises (falls) significantly, he/she will tend to alter associations to being among wealthier (less wealthy) peers. ("Significantly" here would mean a large deviation from the current peer group's average wealth.) As we say in the paper, while the changing of friendships associated with climbing the career ladder is decried by some moralists it is consistent with the hypothesis of reciprocal altruism.
And the changing of such associations may mean that a change in absolute wealth ends up (after associations are changed) altering relative wealth by very little (that is if the relevant comparison group IS one's "peer" associates).
But to test this kind of relative wealth hypothesis, one would have to gather information not only about individual survey respondents, but about the wealth levels of their peers.
Maybe you or others can tell me if this type of investigation has been considered.
Posted by: indiana jim at Apr 18, 2008 1:20:13 PM
indiana jim,
Hmmm. Now I am confused. Are you referring to a paper you have published in JEBO or have
(or had) under consideration there? I am not in my JEBO office, and indeed I see a lot of
these papers. Not sure I remember seeing one by you along the lines you are saying, although
that may just show that I am getting senile if I have(heck, I just turned 60, by gummy). If
I am forgetting something I should be remembering, I apologize.
Thinking about it offhand, I do not remember seeing any papers that have brought reciprocal
altruism into this discussion, although it could certainly be relevant. You are certainly right
that distinguishing properly between absolute and relative income effects very much has to do
with whom one identifies as one's peer group, which can get tricky. This can cut a variety of
ways, with a hedonic treadmill effect kicking in if one upgrades one's peer group, thus finding
oneself not much happier after that big raise and promotion once one becomes another schlunk in
the executive washroom after the initial buzz and thrill.
Easterlin's stuff on the life cycle fits in here. People make a whole lot more in middle age
than when they are young or old (although they are often wealthier when older), but they are not
that much happier, having adjusted their expectations and reference groups accordingly as they
move through the cycle.
BTW, my favorite story on the relative effect is one I heard from Bob Frank, although I know it long
predates him. He said that "The man is happy who makes more money than his wife's sister's husband."
Posted by: Barkley Rosser at Apr 18, 2008 5:36:54 PM
Barkley:
Coelho and I had a paper there in 1998: "Social Context and the Utility of Wealth: Addressing the Markowitz Challenge".
JEBO had us cut it down drastically, but the working paper posted at RePec was a lengthy discussion of reciprocal altruism and status seeking:
http://ideas.repec.org/p/bsu/wpaper/199602.html
I'm only 51 and I'm forgetting lots of things that my wife brings to my attention.
Posted by: indiana jim at Apr 18, 2008 10:08:14 PM
Jim,
I just checked. I was not editor then, and while I saw your paper I do not remember it well. Will check on it
later when I get to the office on Monday. I am aware of your more recent paper, which is not quite so directly
on these issues, although arguably peripherally so somewhat. But, you must realize that JEBO is a central clearing
house for this cottage industry, and I reject more papers on this than ever even get looked at by the ed board.
After all, Easterlin is one of the most senior members of our board and has recently been elevated to being an
Honorary Editor. However, that does not mean that we only publish papers that support his "paradox," and have
published papers that can be viewed as supporting an independent role for asbolute income or wealth, especially
some of these papers on transition, a hot topic in the cottage industry, although they remain open to a variety
of interpretations.
Posted by: Barkley Rosser at Apr 19, 2008 2:22:57 AM
