Because most philosophies that frown on reproduction don't survive.

Monday, October 31, 2016

The Ultimatum Game Election

I attended a conference with my team from work last week, going to talks and workshops dealing with pricing techniques and theory. One of these workshops was on behavioral economics as it relates to pricing, and the ice breaker at the beginning of the session was based on the Ultimatum Game.

The Ultimatum Game works as follows: The facilitator breaks the participants into groups of two. He then explains to each group, "I am giving one of you $100. You can have this money on the condition that the other person in your groups agrees to your proposal on how to split the money. No negotiation is allowed. You make one proposal on how to split the money such as, 'I will give you $50' and your partner responds yes or no. If he agrees, you get the money and split it. If he refuses, none of you get any money."

I proposed giving $40 while keeping $60 and got a yes. One person proposed a $65/$35 split. The rest of the group went $50/$50, except for one group which had a refusal when a man offered the woman he was paired with a $70/$30 split and she turned him down. One guy who was left without a partner announced that he had conducted an imaginary transaction at $99/$1.

The imaginary transaction is actually the solution to the problem at a logical level. It's free money, and you get nothing if you don't reach an agreement, so if you're at the person who receives the proposal you should say 'yes' to anything. If you could count on everyone being logical this way, the person who was in charge of making the offer should offer a $99/$1 split every time in the confidence it would be accepted.

But people aren't simply rational profit maximizers, and in experiments people routinely refuse offers of $30 or less. This means that as the person making the proposal, you need to take into account the sense of fairness of the person to whom you are making the proposal.

This tendency of Person A to offer a more generous split for fear of offending expectations of fairness shows that while an individual Person B may get nothing due to refusing a low-ball offer, the tendency of people to refuse such offers actually means that in general people in the Person B role will do much better than if people didn't enforce this societal expectation. Someone who refuses a low offer gets nothing, but in reinforcing the expectation that low offers will be refused, he benefits others who are in the same position. Refusing may be 'irrational' in the sense of not getting a sure immediate gain, however it results in overall long term gains for the group.

Sitting where we are, a week from the election, it struck me that some of the intra-conservative arguments about whether to support Trump boil down to the two approaches to analyzing the Ultimatum Game. One side argues that if Trump is any short term gain at all to be had in electing Trump president over Clinton (judicial appointments, etc.) then we should vote for him. Doing otherwise is 'irrational'. The other side (to which I adhere) holds that in accepting a candidate as bad as Trump, conservatives would signal to their party that such a low offer will nonetheless by accepted. By rejecting him, we form the expectation that only better candidates will get our support.


Josiah Neeley said...

Consider the 2012 election as a case study. Romney ran a campaign that was perceived as being akin to a lowball offer to Hispanics. As a result, he did horribly with Hispanic voters and lost the election. The GOP got the message: in their election autopsy they identified the vital need to appeal more to Hispanic voters. And then Donald Trump won the nomination.

Unlike the Ultimatum Game, the GOP's "offer" isn't determined by a single individual or a unified group of people. It's the result of an emergent process involving a variety of contending forces within the Republican party. And when a group defects from a party, they make themselves less influential in future decision making.

Paul, Just This Guy, You Know? said...

I wrote essentially this same piece on my blog the year Judy Baar Topinka was defeated by Rod Blagoevich for Governor of Illinois. That year, I supported Randy Stufflebeam, the Constitution Party candidate. That worked out well.

What I later realized is, a candidate is not a product, and the party is not Wal-Mart. The party doesn't actually pick the candidate, as Trump's nomination should prove. The primary voters picked Trump. Nobody was less happy about the result than the party leadership. The primary voters tend to pick whom they like. They're not voting on whom they hope will appeal to you. The GOP had quite a crop of great candidates this time around. Unfortunately, they were all sufficiently self-absorbed that none of them could make a deal to team up and win as a team. So one candidate won, and the rest lost.

It is silly to sit around at the end of October expecting one party or the other to appeal to you by offering a better candidate. You may as well expect the Democrats to do that as the Republicans.

One of two people will be our next president. You can pick one to support, and/or one to oppose, or neither. Not to decide is to decide. If you're OK with either of them, then by all means, don't vote against either of them. The choice is yours, and you're not wrong to insist that a candidate earn your vote; no one is simply entitled to your vote automatically.

But the time to care about the quality of candidates is in the year before the primaries start, some two or two-and-a-half years from now, for the next cycle. For this cycle it was early 2015.

In 2020, it's likely we'll have an incumbent president running for re-election. Those are notoriously hard to beat. It's only happened twice in the last half-century. So the stakes aren't four years, but eight.

Just sayin'.

August said...

The problem with this analysis is that Trump comes after a long list of unconstitutional candidates. Trump is the punishment. The GOP is dead. Maybe he wins, pushes the party leadership out, and keeps the name, but it isn't the same party anymore.

It should have happened to the Democrats, but unfortunately, the left has always had this ability to be against the truth, while knowing enough of it to subvert the republic.

The election turnout will be analyzed and Trump will be imitated by politicians who want to win in particular areas. The only was this phenomenon will fade rapidly is if he gets in and fails miserably.

S. Armaticus said...

Your approach would work if the Democrats thought the same way!

Too bad they don't.

Which brings an old W.C.Fields quote to mind...

Bacon Wrapped said...

If I am understanding the Ultimate Game correctly, it is not a repeated game, so by refusing a low offer there is little gained by Person B. Person B only gains if there is a reasonable assumption that there will be some other opportunity to negotiate or repeat the initial offer process. So signaling that low offers will be refused is still irrational, right?

Darwin said...

Josiah and Paul,

It's certainly clear that the nomination process is more complex than an just making an offer, but the combination of the nomination and the election do provide a information conveying process to which the party (in all its many parts) responds. For instance, the Trump candidacy pretty arguably could have been destroyed early on, had party leaders not allowed themselves to think that it would either wither away or be guidable. The worse this goes, the more they'll learn that's not the case and be willing to spent the political capital next time.

There is also a way in which losing in clear ways seems to guide a party. For instance, the Dems ran pretty far left candidates in '68 and '72, in part due to their own party realignment and breakup. With Carter and then Clinton 1 and 2 (arguably even with Obama to a fair extent) they've pretty much returned to running center left candidates.

I don't know if left leaning moderates were explicitly thinking of themselves as rejecting an offer to guide future offers when they defected to Nixon, but that does seem to have been the result.

Darwin said...


Since I don't agree with your first sentence, I'm not sure where to go from there.

S. Armaticus,

I'm not sure how the Dems failure to police their own candidates would affect things one way or the other. It might mean that the Dem elected in the year in which part of the GOP rejected the GOP nominee would not be to the taste of all Dems either, but if we assume that's the case with Hillary, that should make it that much easier to beat her in 2020 with a halfway decent GOP candidate.

Bacon Wrapped,

It depends what we mean by irrational. The Ultimatum Game is not iterative, so if Person B rejects he gets nothing. However, the fact that most people playing Person A know that fairness is expected (and unfairness may result in rejection) means that offers are higher than they might otherwise be. The lowest real offer in our workshop was 70/30 (and it was rejected). This means that the fear of being rejected kept even the lowest offer 30x higher than it could have been. So while the individual Person B does not benefit, the cultural expectation created by the willingness to reject benefits most Person B players a lot.

And that's basically, what I'd propose a rejection of Trump would achieve for conservatives. Yes, we'd get a lousy president for four years, but we're going to be a pretty bad president no matter who we support. If we make it clear that borderline insane celebrity candidates with incoherent policy agendas lose big, we should see less of those in the future (and see party leaders be more willing to act to keep them from winning.)

Paul, Just This Guy, You Know? said...

The other problem with your argument is, there really won't be a Republican Party by the time the next election cycle comes around.

August said...

I have a constitutional metric. And it got a lot stricter once I got mad enough to apply it.

You have to have a metric. The metric must not be a deep personal hatred for a guy, especially when a large portion of America realizes that most people who hate this guy, hate us.

Darwin said...


I'm unclear how one can have a constitutional 'metric', since I'm not sure how one converts constitutional adherence into an objectively measured numerical value.

That said, one of my key reasons for distrusting Trump is his apparent ignorance of the structure of government laid out in the constitution, and complete rejection of some aspects of it. For example, he pretty clearly does not accept the first amendment and believes we should have libel and press laws similar to those of Great Britain instead.

Joseph Moore said...

If elections were simply iterations over time of the same game, then this comparison would hold water. Instead, elections inform what is possible for the next election. I am struck (as apparently few are) by the 100+ history of the Chicago Machine and its predecessors: how is it that NO viable alternative candidates arise in Chicago? What are considered radical reformers are, based on what happens, only slightly less in bed with the Machine, if that.

Now that Machine holds the White House. They don't love Hillary, but, like Mafia dons, can work something out.

If this is true at all, voting however it takes to get the Machine out of the White House NOW may be our only chance to do so for a long, long time. (As an aside: it seems most of the numerous Machine politicians who have been jailed over the years were nabbed by the FBI - which is now front and center. The Justice Dept, more firmly under White House control, should have nothing to say in this situation, but is forcing its way in. As is said of demonic possession: it only shows when it's not complete. Once complete, it becomes invisible.)

This is why I'm still undecided. The thought of voting Trump makes me ill; the thought of Hillary (or, by any number of crazy theories flying around, someone more to the Machine's liking) as president is even worse. And if she wins, the chances of a candidate better than Trump actually getting elected go down.

It is a sad situation.

Josiah Neeley said...

The Democratic nomination in 1972 was an outlier. It was the first time the Dems had selected their nominee via primaries, and McGovern had helped write the delegate selection rules, so he was able to get the nomination with 25% of the vote.

When a party nominates someone extreme, it's not surprising that subsequent nominees are less extreme. That's just reversion to the mean.