# The ultimatum game: An experimental economics poll

I’m reading a book right now that has an interesting experimental game, and I’d like to try it on all y’all since the outcomes from your average people are never as the economists doing the experiment expect. Please read the premise before voting.

Two people are given \$1000, which they are entitled to keep as long as they can agree on how it should be divided up. The catch is in the decision procedure they must follow - one person is given the entire sum of money then allowed to make a single, take-it-or-leave-it offer to the second person. The second person can either accept the proposed division, in which case the money is divided up between them under the agreed terms, or reject the offer, in which case the money goes back to the experimenters and nobody gets anything.

So, my question to you - you’re the first person; what offer do you make to the second person? What is your rationale?

As the second person, what would you accept as an offer?

I’ll talk about how economics people usually expect this to go later.

I am an economics person, so perhaps my view doesn’t count.

This is game is only vaguely interesting (to put it charitably) because most people don’t act like their lives are endless series of one-shot games. If you model social interactions like repeated games (since most of the time in life, they are), the experimental results and the theory align somewhat better.

The pinhead equilibrium is \$.01. Why give a whole buck when all it takes is a penny?

I’d give \$100, because I think that’d be enough to overcome the pleasure from denying someone their \$999.99.

I go with 50/50 for two reasons. First, and most important, I’ve run into enough bull-headed, stubborn people that I think it likely that the other person might be the type to turn down \$400 even if it’s free money if they think that I’m dicking them out of their fair share. An even split ensures that this will not happen. Second, it would feel unfair to take more for myself. But mostly the first reason.

The obvious answer is to take the lion’s share. The question is how much can you offer a person before they weigh the loss of X potential dollars as less important than their wounded pride over not getting a “fair” deal? Do we know anything about the subject participants? If it’s anything like many of these types of studies they are probably mostly broke college students. So I would be comfortable taking at least \$800 because I think most people who need the money would rather have 200 bucks in their pocket than some ideal of fairness that they imagined in their heads.

Yeah, apparently this experiment comes up in first year Economics, and isn’t tremendously complex; I am curious how this set of people respond to the idea, though.

Keep in mind that among Dopers there are far more hopeless idealists than the populations at large, as well as it being a wealthier than average cross-section. They wouldn’t play the game the same way as a sample that’s representative of the general population. I fully expect a lot more to propose the even split.

That’s why I’d like to see how Dopers approach this question; I’m guessing somewhere between what economists would expect and what the general population does.

Minor notes I feel compelled to make:
Of course, the answers one gives to a poll may not align with the answers one would give if one were actually in the situation of playing the game with real money on the line.

Also, you seem to want to know what people would do both as the first person and the second person, but your poll, I presume, is only for learning about the first.

Not answering since my daughter is working on a PhD in this area, and we jointly teach a class using these topics as applied to engineering.

I wonder if an anchoring experiment would work better.

Yes, this. For me, your second reason would weigh heavier though. (Not to imply that I’m this overly ethical/moral person, but in this particular case I think that would be important to me.)

I voted 501-999 but have two objections.

(1) In the question, but not OP, it says he’s your “buddy.” I think I’d go 500-500 with a close friend, but perhaps 700-300 with a distant acquaintance.

As a matter of “fairness” it seems fair that the advantaged player gets to exercise his advantage! The more handsome buddy would get the dating preference, etc. That’s not “unfair”; it’s just common sense. Exactly where one draws the line is unclear. Almost anyone would reject a 970-30 offer out of annoyance, but it would be hard to find fault with 550-450. That brings me to my second objection:

(2) Isn’t 501-999 an overly broad range?

Not really. The poll choices could just as easily have read, “take less for yourself,” “equal split,” and “take more for yourself.”

I voted 50/50 because the question said the other person was my buddy. I would share with my buddy. With a stranger, I likely would offer less, probably 60/40. I don’t screw people, but the decision was left to me. If the decision was left to the other person and I was offered \$300 for nothing, I would accept it.

50/50

I can’t imagine any set of circumstances where I’d want to be greedy and offer less to anyone: especially a “buddy.”

I’d go 50/50. There’s no reason not to. If I give 500 the other person knows I’m not getting more, and they’re not going to get any more. If I offer less I take a chance that they say no way and no one gets any money. I’d rather have the 500 for sure then any other amount.

50/50. It’s not like either one of us did anything to earn it, and if I kept more than that I’d feel like I was cheating the other guy somehow.

Because the second person can make sure that no one gets anything, the 50/50 split is the only sensible offer that maximizes the return for both parties.

Assuming a stranger, I’d offer a 60-40 split. Unless the sum of money were even larger, in which case I might offer 50-50 to be safe.

Depending on how badly i wanted the money, I’d probably accept up to a 65-35 split.

Then again, i know the literature in this area.

That’s my reasoning.