I rely on FiveThirtyEight to tell me the direction of the election, and a few weeks ago, Utah suddenly became very interesting. Evan McMullin, a previously unheard of candidate for President (who entered the race only a few months before Election Day), appeared in Utah’s forecast numbers with a surprisingly large probability of winning Utah’s electoral votes. Benjamin Morris then wrote about how Evan McMullin could plausibly not only win Utah’s electoral votes, but the Presidency as well.
Evan McMullin appeared in Utah polls not long after the hot mic video of Donald Trump bragging about groping and sexually assaulting women, and many Utah politicians denounced him and pulled their endorsement (though Chaffetz, being the unprincipled, ambitious flake that he is, has since said that he will vote for Donald Trump). Utah Republicans and conservatives have always had a hard time supporting Trump, and Mormons, unlike the evangelical Christian voting block, have been less willing to sell out on their morals and support a man antithetical to everything Christian (if not just plain decency). After the 2005 video of Trump, Evan McMullin surged in Utah polls.
I’m generally skeptical of American third parties. I have eagerly bashed Jill Stein, and I argue on Facebook with someone I will refer to as “Frank Green” (aren’t I punny?) about third parties on a regular basis (these arguments helped inspire me to write my recent post on Hillary Clinton. And I believe that the usual criticism of voting third-party still holds: first-past-the-post voting systems, particularly the American system, produce a two-party state, for reasons I won’t discuss here but will simply assume true.1 That said, there are caveats unique to American federalism that are often not discussed yet matter here, along with those unique to how the electoral college and the rules for choosing a President work. As Benjamin Morris clearly explained, McMullin has a delightfully insurgent path to the presidency that, while very unlikely (Nate Silver said at one point that, in simulated election outcomes, McMullin won the presidency about one out of 5000 times), is not impossible and actually makes Evan McMullin more likely to win the presidency than any other third-party candidate, despite others’ larger national support.
Three things must happen for McMullin to win: 1) he must win Utah; 2) neither Trump nor Clinton must have enough electoral votes to win the presidency; and 3) McMullin must earn the support of the House of Representatives as they decide who becomes President. Step 1 is difficult, but as McMullin fever spreads throughout Utah, McMullin may pull it off. Step 2 is extremely difficult; Clinton is easily the favorite to win. And no one knows how likely Step 3 is to be completed; my guess is it would be easier to achieve than step 2 but harder than step 1 (which is hard enough as it is). So “long shot” does not begin to describe McMullin’s chances, but it could happen, and I find his path a brilliant use of limited resources for a potentially huge gain.
Before I go on, I should probably say something about who McMullin is an what he stands for. Evan McMullin is a Utah native, a Mormon, a former CIA operative, has worked in investment banking for Goldman Sachs, has advised the House Committee on Foreign Affairs and was chief policy director for the House Republican Conference. His platform is unapologetically conservative; it reads like the platform of the nominee Republicans should currently be running. While there are a couple positions here and there I may agree with, I disagree with nearly all of what he calls for; I would prefer a Clinton presidency any day. He is supported by a super PAC, Stand Up America, founded by Republican operatives with experience supporting third-party candidates. Make no mistake; I am not considering supporting McMullin because I like his ideas.
So why the appeal in Utah? As said before, Utah Republicans (Mormons in particular) dislike Donald Trump, who finished third in Utah’s Republican primary. Many pundits explain McMullin’s appeal by citing the fact that McMullin is a Mormon. I will neither agree nor disagree with this; the Mormon community seems to be annoyingly fixated with electing a Mormon President (Ben Carson was a Seventh-Day Adventist, the religion I was raised in, and I was about as unimpressed with him as I was with Donald Trump; but to be fair, I suppose that few Mormons celebrate the fact that currently the most powerful Mormon in Washington, and thus the world, is Harry Reid), but I think there are reasons they support him that go beyond his religion. McMullin offers Republicans and Mormons a way to vote for their principles without voting for Donald Trump, and at the same time, not feel like they are wasting their vote (since recent polls actually put McMullin in the lead to win Utah, and as of this writing, FiveThirtyEight’s polls-only model gives McMullin a 17% chance to win Utah, with Nate Silver admitting that the models will struggle to accurately place McMullin’s chances and and Benjamin Morris admitting the models may be underestimating McMullin’s chances). While we have not seen much of Evan McMullin’s personality, it’s probably better than Donald Trump’s (like getting pricked by a needle is probably less painful than getting shot in the foot; most things are), and McMullin has at least had experience as a CIA agent and in policy, unlike Donald Trump, who throughout the campaign (especially in the debates) has demonstrated how naïve he is about nearly everything policy related while seeking the most difficult policy-related job in the world. These are all good reasons to support Evan McMullin over Donald Trump that go beyond his religion and a mere protest vote (though a protest vote is a good reason to vote against Donald Trump).
Those are reasons for Republicans to support Evan McMullin. So what about Democrats? Like I said, there’s little, if anything, in Evan McMullin’s platform to make Democrats like him. I do believe that McMullin is preferable to Donald Trump for the reasons I listed above, and this is a reason to hope he gets the presidency instead of Trump. However, that’s a far from satisfying reason to vote for McMullin, who is definitely an ideological opponent, all while Clinton appears to be in a strong position to win the presidency.
This is the point where “Frank Green” and other third-party voters and supporters, who have been beaten over the head for decades with Duverger’s Law and the “third parties can’t win election; they’re spoilers at best” arguments, may feel some vindication. I’m not recanting those arguments. They are still very much true and still hold… at the state level, not the national level. Basically, all elections in the United States are held at the state level or lower; there is no such thing as a national election (or, at least, not one you participate in). This is obvious for offices such as Senator or Representative, but it’s true for the presidency as well. We often fail to remember we don’t vote for the President, but instead vote for who our state’s electors will support for President, meaning that the winner-take-all voting dynamics apply at the state level. The difference is subtle but crucial, and is the reason why we got George W. Bush instead of Al Gore as President in 2000 (thanks, Florida).
Duverger’s Law says that third parties don’t have a chance because as the election carries on, voters abandon candidates who are trailing behind the leading two candidates, in order to minimize the maximum lost utility2 and elect an individual that does not differ too much from a voter’s own ideological position. If the polls or FiveThirtyEight are to be trusted, Hillary Clinton is on track to finish third in Utah, behind both Evan McMullin and Donald Trump, either of whom have enough votes to beat Clinton single handedly. This pushes McMullin from the roll of spoiler to being a serious candidate for receiving Utah’s electoral votes. It seems that distaste for Donald Trump has not been enough to push Utah voters to support Clinton. Clinton now finds herself in the position of a third-party candidate in Utah, and according to Duverger’s Law, the optimal strategy for Utah voters (including Democrats) is to abandon Clinton and support one of the leading two candidates. As mentioned above, McMullin is probably the preferred between the two.
ManyClinton supporters may not see this as sufficient reason for voting for Evan McMullin, so I will also mention that Evan McMullin winning Utah is better for Clinton than Donald Trump winning Utah. The best outcome would be Clinton winning Utah, but that’s unlikely to happen if we trust current polling data. So the next best outcome would be McMullin winning.
McMullin’s path to victory gives Clinton a better chance of winning the presidency than Trump. If Utah’s electoral votes go to Evan McMullin, then Donald Trump lost six electoral votes he probably would have otherwise earned. If the election is sent to the House of Representatives, then either there was a tie in the electoral college, or Donald Trump would have otherwise won. While sending the election to the House probably does not bode well for Clinton, it’s much better than an automatic victory for Trump, and support for McMullin in the House may lead to a split vote that gives Clinton the election. (And if McMullin actually won, as I have said before, that’s probably better than a Trump victory even from a Democrat’s perspective.)
Now, this argument is good reason for Utah Democrats to vomit a little in their mouth; it seems the best thing to do for Clinton is vote for an ideological opponent who, in many circumstances, we would oppose vehemently. You now know the sick feeling many third-party supporters probably feel when told to vote for one of the two major parties, and see why we badly need to dump our first-past-the-post voting system and the electoral college in favor of a better voting system. I would eagerly have that conversation, but no one is going to change how voting is done less than two weeks before the election.
I’m not going to give Evan McMullin a ringing endorsement or actually instruct Utah Democrats to vote for McMullin; I’m the one who wrote this article and I’m not sure if I actually want to turn this theory into the most disgusting vote I’ve made. But I am considering voting like this, and I hope other Democrats do at least consider it as well.
As for Utah Republicans… please don’t vote for Trump. You know he’s terrible. Keep some shred of decency and vote for someone else. Perhaps Evan McMullin.
- Duverger’s Law is the cited reason for the dirth of third parties in fist-past-the-post voting systems. I may write a future blog post discussing these reasons in greater depth, but I feel it’s a separate discussion. For now, CGP Grey has a good video explaining the phenomenon. ↩
- Notice that I said “minimize the maximum lost utility” instead of “maximize gained utility”. This is known as the minimax principle, a key idea in decision theory, behavior theory, game theory, and so on. There is a big difference between “minimizing maximum lost utility” and “maximizing gained utility”. The former is preferred over the latter because the former is a much more stable means of making decisions. You can justify playing the lottery if you are trying to “maximize gained utility”; you’re trying to get the best result, chances and state of reality be damned! Minimax optimization, though, leads to an achievable and not-too-bad result. Thus, minimax is seen as being the more rational of the two. Voting for a third party in a first-past-the-post system with a stable two-party system can be justified by the “maximize gained utility” view, but not the minimax perspective; it’s not a minimax-optimal voting strategy. This explains why voting for third parties is seen as being suboptimal at beast, irrational at worst. ↩