Market Failure Within Politics

Market failure, a technical term in economics, does not mean any situation where a market fails. It describes situations where individual rationality fails to produce group rationality, where each individual takes the action that he correctly believes is in his interest and most, in the limiting case all, are worse off as a result. One of my standard examples is an army losing a battle and being mostly slaughtered because each of ten thousand soldiers correctly believes that his decision to run has only a small effect on the chance that the line breaks, a large effect on the chance that he gets killed.

As that example suggests, the problem is not limited to markets narrowly defined; I argue in one chapter of the latest edition of my first book that market failure on the political market is the reason that having decisions made by government is usually a mistake.

Market failure also exists within the political system, supporters of a party or ideology taking actions that make less likely the outcomes they want. For example …

Why They Shouldn’t Pick on Palin

Back when Sarah Palin was running for the Republican nomination it struck me that her critics on the left were acting against their own self-interest. All of them preferred Obama to any of the Republican candidates. I expect most of them believed, as I did, that Palin would be one of the easiest for him to defeat. One would expect them to go as easy as possible on her until it became clear who the Republican nominee was going to be and only go back to making fun of her when it became clear that she had no chance of being the Republican candidate — or already was.

That was not what they were doing.

One possible explanation is that her critics on the left thought their attacks would make Republicans more willing to nominate her. An alternative I find more plausible is that what I was observing was the problem of market failure on the electoral marketplace. The left winger who refrains from attacking Palin, given a halfway plausible opportunity to do so, is giving up a private benefit, the fun and kudos of mocking the big bad witch. To refrain from doing so in order to increase her chance of being nominated would be to produce a public good, the public in question being all of the people who want to see her nominated and defeated. Public goods, goods whose producers cannot control who gets them and so cannot charge for them, are under-produced, a familiar example of market failure.

An alternative possibility is that her attackers were afraid she might get nominated and win and would be a worse president than McCain. I find that unlikely in the case of Palin, more plausible in the case of Trump in 2016, more plausible still in the case of Trump in 2024. I expect that there are a lot of Democrats this year who would prefer almost any Republican candidate, even one more likely to get elected.

How to Put a California Republican in  the Senate

California is a very blue state; the last statewide election won by a Republican was in 2006. California also has an odd primary system, a “jungle primary.” Candidates of all parties run against each other, with the two top scorers getting on the ballot. This sometimes results in two Democrats running against each other, with no Republican remaining in the race.

This year we had an interesting demonstration of the odd incentives produced by the system. Adam Schiff was the leading Democratic candidate, Steve Garvey, a retired sports star, the leading Republican. The next two Democratic candidates, Katie Porter and Barbara Lee, were both to Schiff’s left, with combined support, according to the polls, only a little below his. If one of them ended up sharing the ballot with Schiff there was a real chance that progressive Democrats would have enough votes to defeat him.

Schiff solved that problem by running ads attacking Garvey as a conservative Republican and Trump supporter, an argument calculated to reduce his support among Democrats, very few of whom were likely to vote for him anyway, but increase his support among Republicans, make sure they voted for Garvey instead of the other Republican in the race.¹ With all of the Republicans voting for Garvey, half the Democrats for Schiff and the rest split between Lee and Porter, Schiff would end up with no Democratic opponent on the ballot.

It worked — the senatorial race will be between Schiff and Garvey. Porter’s supporters were not amused:

Adam Green, a progressive activist and Porter ally, issued a statement attacking Schiff after the projection was announced.

“Adam Schiff put his own selfishness above democracy by lifting up Republican Steve Garvey, who will now turn out Trump voters in key House races that could determine control of Congress,” (NBC news, March 5)

Garvey got about 32% of the primary vote in a state where only 24% of the voters are registered as Republicans, evidence of his ability to attract independents, perhaps some Democrats, and sports fans. Combine that with progressive Democrats angry enough at Schiff to refuse to vote for him, perhaps write in Porter or Lee as a protest, and if 2024 turns out to be a Republican year it is possible, although not likely, that Adam Schiff’s clever tactic will end up costing his party what would have been a safe senate seat.

A more likely result in the same direction is suggested by the Adam Green quote above. A Republican in the senate race is a reason for Republicans to bother to turn out to vote; even if that isn’t enough to win Garvey’s race it might make the difference in tight House races, help the Republicans keep their control of the House.

On the other hand, a close Schiff-Porter race would have been a reason for both candidates to spend the money of Democratic donors on their campaigns. Porter’s donors are now free to spend their money on other races and, if Schiff’s race against Garvey looks like a sure thing, he can funnel money from his donors to other Democratic candidates.

All things considered, it is hard to be certain whether spending Schiff money on Garvey’s campaign will end up helping or hurting Schiff’s party but it is pretty clear that that was not his central consideration in doing it.

Trump and the Republican Party

In the political cases I have been describing the actors, Democratic politicians and their supporters, had a common interest, the electoral success of their party, an interest arguably harmed by their self-interested actions. How willing they were to take such actions depended in part on how large the negative effect on their group interest was relative to the private benefit. Each individual Palin critic could reasonably believe that any negative effect of his writing via a reduction in the chance that Palin would be nominated, hence an increase the chance that a stronger Republican candidate would be nominated and elected, was tiny. Adam Schiff could be pretty sure that even if Steve Garvey could match his votes in the primary Schiff could beat him in the actual election.

That was not the case for Trump’s political actions over the past four years. In 2021 there were runoffs for both Georgia senate seats. Trump’s attacks on the Georgia election cost the Republican candidates votes, probably cost their party control of the Senate. The 2022 elections for the House were supposed to be a red wave. The party barely kept their majority, due in part to Trump’s insistence on nominating candidates loyal to him.

A second factor in how willing someone is to serve his individual interest at a cost to the group he is part of is how much the group interest matters to him. Trump’s actions were calculated, probably correctly, to maintain his control over the Republican party and the 2024 nomination. Pushing Republicans who did not support him out of the party by running candidates they would refuse to support was a loss for the party but a gain for Trump. Fewer Republicans in the House and Senate meant that more bills supported by Democrats and opposed by Republicans would pass, but that does not matter to someone with no ideological commitment to Republican policies. Trump, so far as one can tell, has no ideological commitments at all.

A Dubious Example

If you imagine all voters aligned on a one dimensional scale, any Republican candidate noticeably to the right of the median should be defeated by a Democratic candidate a little less to the left of the median. It follows that a party that wants to win should nominate a centrist candidate, a Republican on the left of his party or a Democrat on the right of hers — the median voter theorem.

When I started thinking about the public good problem in politics, one of the first examples that occurred to me was the failure of party primaries to produce that result.

I was wrong, for two reasons. Market failure is about situations where individual rationality fails to produce group rationality. If Republicans, even right wing Republicans, prefer a centrist candidate who can get elected to a right wing candidate who can’t, that is whom it is in their individual interest to vote for. If instead they vote in the primary for the candidate whose views are closest to theirs, the obvious explanation for the failure of the median voter theorem to predict which candidate gets nominated, they are making a mistake, being irrational.

Alternatively, perhaps they are not making a mistake. Political outcomes are not as simple as the theorem assumes. How many votes a candidate gets depends partly on his political position, partly on how charismatic he is, partly on how much money his supporters are willing to spend on his campaign, partly on the mood of the electorate, partly on random chance, even on the weather on election day. If the choice is between nominating a centrist candidate who will do almost the same things as the centrist candidate of the other party or a candidate closer to my views, it my be worth accepting a somewhat lower chance of winning in order to get the candidate I want. The Republican who votes in his party primary for the candidate he likes or a Democrat who does the same may not be making a mistake.

1. Porter attempted to counter the tactic by running ads attacking Early, the other Republican, as the real Trump supporter.

[In writing this post I confused Palin’s 2008 campaign for Vice-President with her suggestion in 2011 that she might run for president in 2012.]