« Today's happiness research, part II | Main | Do women talk more? »
Testosterone economics
Remember the ultimatum game?
In this game, one player divides a pot of money between himself and another. The other then chooses whether to accept the offer. If he rejects it, neither player benefits. And despite the instincts of classical economics, a stingy offer (one that is less than about a quarter of the total) is, indeed, usually rejected.
Here is the latest result:
...the responders who rejected a low final offer had an average testosterone level more than 50% higher than the average of those who accepted. Five of the seven men with the highest testosterone levels in the study rejected a $5 ultimate offer but only one of the 19 others made the same decision.
In other words, irrationality isn't just a deviation due to imperfection, we are programmed to be spiteful. Here is information on how high testosterone levels are correlated with urges to compete and be dominant. Here are some other correlations; should we make a few leaps and infer that women in "sheer" clothing are more likely to be spiteful? Should you prefer to undertake joint projects with men of slight build and women in flat shoes? Should you deliberately seek out non-hot mates, in the realization that long-run cooperativeness is of more value than short-run hotness?
The pointer is from Daniel Akst.
Posted by Tyler Cowen on July 5, 2007 at 01:46 PM in Economics | Permalink
Comments
I know others have made this point ad infinitum, but the ultimatum game was more likely showing people
playing the strategy of not wanting to look weak. "If word got out that the second parties are pushovers..."
The experimenters of course, "assured" the subjects that the tests are, like, totally
independent. But the subjects likely have a broader definition of "repeated game".
Posted by: Person at Jul 5, 2007 2:18:16 PM
http://www.ns.umich.edu/htdocs/releases/story.php?id=3209
A study which found that high testosterone correlates with getting reinforced by frowny faces.
Posted by: Nancy Lebovitz at Jul 5, 2007 2:25:02 PM
Of course, we're programmed to be spiteful.
Posted by: josh at Jul 5, 2007 2:27:18 PM
I think the more interesting issue would involve ultimatum games with higher and higher stakes. Okay, so our testosterone types reject offers of $1 for them and $9 for the offerer. I can understand paying $1 to say "screw you."
So how much testosterone would you need to reject a split of $1000 for you and $999,000 for the offerer?
I think this can illustrate some of Caplan's valuable insights. After all, when your vote has a very tiny effect on the outcome, then the hormonal effects dominate the rational, which is why people turn into savages in the voting booth.
Of course, in the opposite ideological direction, if income confers diminishing marginal utility, the rich behave more irrationally when it comes to money. (More precisely, getting richer makes a person less rational.)
Posted by: Keith at Jul 5, 2007 3:03:46 PM
Testosterone is irrelevant. Gender is socially constructed. Biology does not influence human behavior.
Repeat 20 times before breakfast.
Ps. Aren’t these results really old? JEP article by Fehr and Gächter cite Bumhams result in 1998. The Economist is not exactly on the forefront on issues that are less than politically correct…
Posted by: Tino at Jul 5, 2007 3:30:20 PM
Keith: So how much testosterone would you need to reject a split of $1000 for you and $999,000 for the offerer?
I don't know ... but you'd need quite a bit to ask for the research grant to study this! :-P
Posted by: Person at Jul 5, 2007 4:11:38 PM
There's no irrationality here, people simply get value from being spiteful. Tyler needs to read his Mises again.
Posted by: Noah Yetter at Jul 5, 2007 4:19:15 PM
Well, I'd start with that, and then my counteroffer of a research grant using $10,000 amounts would seem downright reasonable. Heck, I'd be giving up 99% of my initial request!
Posted by: Keith at Jul 5, 2007 4:22:20 PM
With regard to the high total value ultimatum game split: it's an interesting question. Of course, you can't get someone to give you the grants.
Maybe they should try to get Blizzard (the computer game company) to sponsor the program. Blizzard could generate a few 1,000 gold amounts on World of Warcraft, and the test subjects could all be WoW players. The test would, essentially, be paid for via inflation of their virtual economy.
While 1,000 gold isn't worth $10,000, it's probably worth a hell of a lot more than $10, at least to a dedicated WoW player.
Posted by: Michael B Sullivan at Jul 5, 2007 4:35:06 PM
Muscle-bound macho men are not necessarily more handsome.
Posted by: Amelia Bedelia at Jul 5, 2007 4:40:21 PM
The next stage would be to examine how natural differences in testosterone levels and testosterone-receptor genes in different populations around the world affect cooperativeness.
Posted by: Steve Sailer at Jul 5, 2007 4:53:30 PM
I think that spite is certainly the explanation - when you do this one-off test where the counterparty is a computer, people do not reject the split.
Increasing stakes do not change the results greatly; there is a late 90s paper from an Australian researcher who did the Game in Indonesia using a month's salary, and got similar results as with relatively small stakes.
Posted by: cure at Jul 5, 2007 5:01:15 PM
Aristotle defines anger in terms of someone's failure to accord you the status you deserve.
I guess that would turn out to be testosterone-dependent as well.
Posted by: Anderson at Jul 5, 2007 6:31:58 PM
I would like to see an iteration where an attractive woman is the one dividing the money and a man is choosing to accept/decline. I predict very different results.
Posted by: fustercluck at Jul 5, 2007 9:38:45 PM
Standard advice:
"It's just as easy to fall in love with a rich [girl/boy] as it is with a poor [girl/boy]."
Posted by: chug at Jul 5, 2007 11:19:59 PM
"Ps. Aren’t these results really old? "
Yes, I believe the experiment was run in 1996 and there have been numerous working papers in circulation. The fact that NO editor at an economics journal has had the courage to publish these findings is not something for which one can hold Dr Burnham accountable. Credit to Proceedings B for finally making sure these ideas are disseminated to a wider audience.
Burnham also has an important paper on the proper interpretation of laboratory evidence in the obscure journal Analyse & Kritik. I urge all aspiring experimentalists to read it.
Posted by: Imperator at Jul 6, 2007 12:04:28 AM
I'm not sure spite is the relevant emotion, but rather an insistence of being treated with respect and not being seen as a pushover -- which could rationally be worth more than the monetary value of the low offer. Yes, participants are told these are anonymous transactions, but people still use their standard 'programs' for striking deals in these situations, I believe.
Posted by: Slocum at Jul 6, 2007 8:22:42 AM
There was an old game which had this as a component. It was a third-world "simulation". The president would draw seven bills off of the top of the stack (foreign aid), and propose a budget (without anyone being able to see what he drew). People could vote for or against, with atternative resolution (coup attempt) available. The goal of course is to accumulate as much cash in Switzerland as possible.
FAR more subtle effects.
Posted by: Nathan Zook at Jul 6, 2007 8:46:26 AM
Nathan_Zook: The game you're talking about, in case anyone is wondering, is Junta. I played that with my friends once. Pretty fun.
Posted by: Person at Jul 6, 2007 9:38:35 AM
"Yes, participants are told these are anonymous transactions, but people still use their standard 'programs' for striking deals in these situations, I believe."
Absolutely, and this, in my opinion, goes to the heart of behavioral sciences. The vast, vast majority of people do the same thing they always have even if the underlying game dictates a different optimal strategy. The vast majority of people do not abtract in any way, nor is their behavior dictated in any way by abstractions. They do what they've been raised to do, what they've been told, and what has worked for them in the past.
This means that people's behavior will appear to be very optimizing during stable periods, where the templates and norms that have developed over time still work. But during periods of rapid change and novelty, many people will fail miserably. The small group of people who actually do "game things out" correctly will do very well, partly at the expense of the non-abstracters, and their behavior will become the basis of future social norms.
Posted by: Keith at Jul 6, 2007 10:43:27 AM
I wonder if you'd have different results in Hong Kong vs Berkeley? In other words, how much of this is due to Marxist-think brainwashing?
Of course if those groups were significantly different it doesn't necessarily mean "brainwashing", it could be at least partially selection bias, with those inclined to think like spiteful Neanderthals choosing to locate in Berkeley, and those with more sense choosing to leave ASAP.
Posted by: happyjuggler0 at Jul 6, 2007 10:44:34 AM
hi everyone, just happened to come across this discussion completely randomly and was wondering:
has anyone read/ heard of the book 'Passions Within Reason: The Strategic Role of the Emotions' by Robert H Frank? (It's a great book, btw)
Because this goes to the heart of Slocum's comment
"I'm not sure spite is the relevant emotion, but rather an insistence of being treated with respect and not being seen as a pushover -- which could rationally be worth more than the monetary value of the low offer."
Frank's thesis is pretty much exactly this -- that seemingly 'irrational' emotional reactions actually have deeper rational purposes. And it is emotions which determine that the people's 'standard programs' kick in, or rather, emotions effectively are humans' 'standard programs'.
Anyway, hope that's of interest. Cheers!
Posted by: Jonathan Mills at Jul 10, 2007 9:12:29 AM
Posted by: 鑽石 at Apr 2, 2008 9:01:24 PM
