The ultimate experiment

Author: Carlos Alberto Gómez Grajales

Queen's famous ballad 'We are the champions' was voted as the world's favourite song, according to a poll sponsored by Sony in 2005. It has become an anthem for sporting victories and if you believe some researchers, it is the catchiest song in the history of popular music.

Yet I'm not here to discuss popular rock songs, though there's plenty to say about that. It is just an excuse to discuss the experimental equivalent of ‘We are the champions’. It is a research methodology as popular and catchy as Freddie Mercury's anthem and just like Queen's work, you've probably heard about it but it just never gathered much of your attention. It is the ‘Ultimatum Game’.

The ultimatum game was first presented to the research community in 1982 in an article entitled 'An Experimental Analysis of Ultimatum Bargaining'. The authors devised a scheme to study their subjects natural reactions and strategies for bargaining. The experiment has then been replicated hundreds of times, with interesting results that challenged the perception of economists.

The ‘Ultimatum Game’ is simple in its definition: it is a game where two people are bargaining on how to split a fixed amount of money, by following certain rules. The first player proposes how to divide the sum between the two players, and the second player can either accept or reject this proposal. If the second player rejects, neither player receives any money. If the second player accepts, the money is split according to the proposal. The game is only played once and no counter offers are allowed, hence the ‘ultimatum’ name.

When the experiment was devised, economic theory predicted that the second player would happily take any amount more than zero. Game theory was popular among economists at the time and they modelled that any normal, rational human being would gladly prefer to earn one dollar, than to earn zero dollars (remember that rejecting the offer meant that no one gets any reward). Interestingly, experimental evidence quickly refuted this argument, the most common offer during the first runs of the experiment clearly favoured an even split.

According to some of the researchers, the reaction among the economists after the first results of the ultimatum game arrived was usually ‘clearly these people are stupid’1. Why would anyone give away half the loot if the second player would clearly take any amount of free money? Actually, further studies confirmed that this response was far more common than economists had ever imagined. Not only is an even split the most common solution found in most cultures, it became noticeable that offers of less than 20% of the amount of money were often rejected: people were refusing free money.

Only in an isolated culture in the jungle did economic game theory triumph. The Machiguenga people of eastern Peru, with an almost non-existent economy, were one of the few ‘rational’ subjects in the world where economic theory held and low offers were gladly accepted. After the first awkward results of the ultimatum game, economists thought that very unequal allocations were rejected only because the amount of money was too low. So in theory, the absolute amount of cash involved was not enough to produce strategically optimal behaviour. You are rejecting free money because it is just a dollar, your pride wouldn't be so hurt if you were offered something around ten or fifteen bucks.

However, economists failed once again. The experiment was repeated on several occasions and this time the amount offered was substantial. Studies found that the higher the stakes, the closer the offers approach an even split. Even in a $100 game played in Indonesia2, $30 offers were turned down. To put things in perspective, $30 can be the equivalent of up to two week's wages in Indonesia.

Due to these controversial results, just like Queen’s famous ballad, the ultimatum game has been sung and repeated around the world countless times. The experiment, in many variations or 'covers', has been replicated using various different individuals and cultural backgrounds3: men and women, Asian and Israelians, children, autistic persons and individuals with exceptionally high IQ's among others. Almost every time the experiment is replicated, some cool statistical technique is used, ranging from non parametric tests to some interesting recent developments that involved Logistic Regression models aimed to predict the behaviour of the participants.

Ran Spiegler, author of ‘Bounded Rationality and Industrial Organization’, once described the ultimatum game as ‘by far the most intensely studied experimental game’. Rock stars tend to die young, but don't worry, the ultimatum game has been studied for more than three decades by now and is still going stron


  • 1. Poundstone, William. Priceless: The Myth of Fair Value (and How to Take Advantage of It) Hill and Wang; First Edition edition (January 4, 2011)
  • 2. Cameron, L.A. (1999). “Raising the Stakes in the Ultimatum Game: Experimental Evidence from Indonesia.” Economic Inquiry. 37, 47–59.
  • 3. Oosterbeek, Hessel et al. Cultural Differences in Ultimatum Game Experiments: Evidence from a Meta-Analysis. Experimental Economics, 7:171–188 (2004)

Bookmark and Share

Comment on this article

Submit your comment
  1. Image of unique ID


Carlos Alberto Gómez Grajales

What I find particularly interesting about these experiments is how flawed the theoretical basis of Economics theory was and how terribly it predicted the outcomes. I mean, back in those days was a huge shock that implied that all theory had to be re-written. It is a great example of how experimentation and statistics may be used to test even some of the most fundamental theories. We should keep doing experiments like this in all fields!

By the way, here's some info about the Machiguenga tribe in Peru, the only "rational" men in the world:

Thanks for reading!

reply to this comment

Alex Wolfe

I think something would be better than nothing and am surprised how often neither party gets any money

reply to this comment

Gerhard Adam

First, they have a strange definition of "rational".  In the second place, the confusion occurs, because of the false belief that people are fundamentally selfish.  So, the assumption is that people will accept whatever is given to them, because they view only their own gain.

However, as this game illustrates, people are not selfish, they are cooperative, so when someone displays selfish or greedy behavior [i.e. a non-equitable split] they are more apt to react in punishing the individual than considering personal gain. 

An interesting variation would be to see what happens if the first person offers more than 50% and then allow an opportunity for the second individual to return some of that.  It would be informative to see how many people [even given a chance to hold a larger amount] would strive for a more equitable distribution.

reply to this comment


As soon as I see a study is using this artificial game-theory crap, I stop reading.  You can be 100% certain the result is transparently wrong, attributing reverse causation and ignoring innate differences.

Maybe once a year a published study will use valid methods (natural non-computerized experiments with real people who are NOT psych undergrads).  I'll read and respect those extremely rare studies.  They may not be correct, but they're never instantly and obviously wrong.

reply to this comment

John Callaway

Simple reason.  Everyone knows there is no such thing as a free lunch.  When offered a free lunch, participants declined because they could not imagine this was "just an experiment" and instead bid 0 to avoid a "gimmick". 

reply to this comment


Quite clearly, this is due to some inherent sense of "fairness," where people feel that it is preferable to punish someone for attempting a "low-ball" offer than it is to receive a less-than-fair amount. This has been demonstrated before, and humans are not the only species this behavior has been observed in.

I would like to know more about the culture in which low offers were always accepted; is it a communal-type culture? There it would make sense to always take whatever offer is provided, as it would be for the benefit of the entire tribe regardless of who makes the most in the game.

reply to this comment


Taking some time to think about this, if I was one having to accept the offer, I would decline if the split was too small.  And if I was the one making the offer, I would offer a 50/50 split to ensure that the offer wouldn't be refused.

reply to this comment

Eric Fanwick

I find it funny that the only "rational" people in the world seem to have a name that resembles the Yiddish word for crazy.

reply to this comment

Skip to Main Site Navigation / Login

Site Search Form

Site Search