It's Time for Rubio to Get Out

I love Marco Rubio.  I haven't talked about him too much here on the blog, because I haven't had much to say about him that hasn't been said better by others elsewhere.  But I think he's a great American, an inspiring leader, and an all-around decent guy. I will always be grateful to him for giving the best debate answer in history -- a politically risky defense of human life in the heart of hostile New Hampshire, even as his so-called "Catholic" opponents, Christie and Bush, attacked him for daring to stand up for the most innocent, most defenseless people in our society.  I wrote a stirring pro-Rubio speech for the Minnesota caucuses last week, and was very proud to live in the first (and, so far, only) state to deliver Rubio a victory.*  Somehow, he has been tarred with the brush of "Establishment" over a single mistake he made over an immigration bill several years ago; I know that this is not the case, and that Rubio is one of our finest conservative rebels. (I mean, c'mon, guys, the Conservative Review, no friend of Rubio's, still admits that he's the 7th-most-conservative guy in the Senate!)

Sen. Rubio performed perhaps his greatest service ten days ago when he finally, finally turned the guns of the Republican Party on Donald Trump, leveling our frontrunning fraudster with mortar and grapeshot, from Trump's progressive record to Trump's moral repugnance to Trump's victimization of middle-class Americans to puerile insults -- the same sort Trump has used to take the lead, and thus a necessary evil.  Sen. Rubio's have borne fruit, both in the better-than-expected results of Super Tuesday and the much-better-than-expected results of Super Saturday.

But Sen. Rubio's efforts haven't benefited Team Marco. Instead, they've apparently redounded to the good of Sen. Ted Cruz, who has pulled decisively into second place (with an assist from Dr. Ben Carson), while Gov. John Kasich (a man who seems determined to become Trump's VP) continues to deny Rubio oxygen at every turn.

The conventional wisdom (today, at least) is that Rubio should drop out if he can't win Florida on March 15th.  I think conventional wisdom is wrong.

Rubio should drop out now.

Many people are saying that Rubio needs to stay in simply because we need the 99 delegates from Florida to stop Trump from winning the nomination.

I don't deny that getting those 99 delegates would be wonderful.  If Trump loses them, it makes his march to "clinching" the nomination (1,237 delegates; he currently has 391) much harder.  However, I think this argument both overestimates the chances that Rubio will actually succeed and underestimates the costs of Rubio staying in through Florida (even if he does succeed).

First, let's talk about Rubio's chances of winning Florida: they're bad.

Rubio trails in every recent poll.  A poll released today by Monmouth, conducted after last Thursday's debate (which appears to have done serious damage to Donald Trump), shows Rubio still losing in Florida by 8 points.  This is somewhat better than pre-debate, when Rubio was losing by an average of 16 points.  But it is also a long way from a win.  Rubio supporters may point to an internal poll released by the anti-Trump Our Principles PAC, showing Rubio down by only 5 points.  But this is even worse: Nate Silver notes that campaigns only release their best internal polls, which means internal polls released to the public biased by, on average, six points.  So the best way to read that internal poll is that it shows Rubio losing by 11 points -- again, even after the debate.  Perhaps most damning of all, the Monmouth survey shows Rubio still in a too-close-to-call battle with Trump even if the other candidates drop out.  Given that the entire Republican strategy must be focused on consolidating the field ASAP, the fact that Rubio does not clearly win his own home state in a fully-consolidated field makes him an implausible candidate to serve as the anti-Trump standard bearer.

Now, I say that Rubio's chances in Florida are bad.  They are not impossible. FiveThirtyEight uses both a polls-only model and a "polls-plus" model (which counts endorsements and so forth).  The polls-only model has been, by far, the more accurate instrument this year, and it gives Rubio only a 14% chance of winning Florida.  But the polls-plus model has occasionally been right, and it currently gives Rubio a 40% chance (the Monmouth poll was actually a big boost: he was at 31% this morning).  My own gut, given what I know about Rubio and Florida, gives him about a 20%-25% chance of winning -- smack in between those two models.  A 1-in-4 or 1-in-5 chance is bad, but not insanely bad.  I've bet on horses at those odds, and, occasionally (about one in every five races), I've won those bets!  So nobody would be shocked if Rubio won Florida, the way I was shocked when Cruz won Kansas by 25 points.

So why not leave Rubio in the race?  He's taking his shot at Florida.  If he loses, we're in the same place we are today if he drops out: Trump gets 99 delegates.  But if he wins, Trump is denied those 99 delegates!  It might be a bad bet, but why not let Rubio make it?

This brings us to Reason #2: Rubio's continued presence in the race has inherent costs outside Florida.

There are twelve states and territories that vote between today (March 7th) and Florida (March 15th), including five other states that vote on March 15th itself.  If Rubio is still in the race, he divides the anti-Trump field in all those states, making it easier for Trump to snap up precious delegates.

This is very obvious in the winner-take-all states which vote on the 15th: In Ohio, for example, the winner in the state will get all 66 of Ohio's delegates, no divisions, no compromise.  Gov. John Kasich is currently polling just behind Donald Trump in Ohio (36% to 33%), while Marco Rubio absorbs 5% of the vote.  If Rubio drops out, more of his votes in Ohio will go to Gov. Kasich than to Mr. Trump (there is precious little polling on second-choice candidates to support this, but I trust it is an unexceptionable claim).  In a close race, this could be Kasich's margin of victory: the difference between Trump getting 66 delegates all to himself or Trump losing all 66 delegates to someone else.

The same pattern holds true in the other winner-take-most states voting on March 15th, namely Missouri (52 delegates) and Illinois (69 delegates), where Rubio can't realistically win (especially now that he's turtled up in Florida), but where he is currently serving as a likely spoiler to the other non-Trump candidates.

And it doesn't stop there.  Even in the proportional states voting between now and the 15th, Rubio's presence strengthens Trump's.  For example, Michigan has a candidate viability cutoff of 15%; if you get 14% of the vote, you get zero delegates.  Guess where Rubio's polling last put him?  Yep -- 13%.  If he misses that cutoff, then every single vote cast for him is wasted, and the overall allocation is calculated as if none of those votes were ever cast.  Why this is so damaging is a bit hard to conceptualize at first, so instead consider these models:

First, we'll assume Rubio stays in, but just misses the viability threshold, exactly as 538 currently projects. Here's how delegates are divvied up if that projection is correct: Candidate Vote % Delegates Trump 39% 28 Cruz 23% 16 Kasich 22% 15 Rubio 13% 0 Now assume Rubio drops out before Michigan, and his voters split up, half to Cruz, half to Kasich: Candidate Vote % Delegates Trump 39% 24 Cruz 30% 18 Kasich 29% 17 Notice what happens here: it's not just that Rubio's dropping out allows Cruz and Kasich to pick up more votes (and thus more delegates). The mere fact that more votes were cast for candidates who made the viability threshold changes the allocation calculation, and Rubio's dropping out causes Trump to lose four delegates, even though Trump got exactly the same number of votes!  That's an eight-delegate swing in favor of #StopTrump!  If Rubio stays in, on the other hand, those delegates are lost... or, at the least, put at serious risk, since Rubio is more likely than not to miss his viability thresholds.

Now, repeat this exercise in every other state that votes before the 15th, all of which have similar viability thresholds.  (North Carolina, thank goodness, with its 78 delegates, does not.)  Rubio's main contribution in those races is likely to be either (a) missing the threshold himself, wasting all his votes and handing delegates to Trump, or (b) causing another candidate to miss the threshold, with the same effect.  It's hard to say exactly what that adds up to, but my back-of-the-envelope math puts it around 10-15 delegates.  Not an irretrievable loss -- certainly not as dangerous as losing Ohio or Missouri or Illinois -- but a substantial number nevertheless.  If someone offered me 15 certain delegates or a 1-in-5 shot at Florida's 99 delegates, I'd bank the 15.  And that's without adding in the literally hundreds of delegates at stake in the winner-take-all/winner-take-most states.

So Rubio staying in has costs.  Substantial costs.  Even if he wins Florida, with its 99 delegates, there is a large chance that his staying in the race until then will cost the #StopTrump campaign nearly that many delegates in other states (if we lose Ohio because of a Rubio spoiler, it would add up to more than Florida).  And the odds that he wins Florida are low, as we've discussed: 1-in-5, maybe 2-in-5 at best.  He still has to exit the race eventually anyway, because a divided field will lose to Trump in the end if we do not rally around a single candidate by mid-April, and Rubio simply doesn't have a plausible path to overtaking Cruz by the convention at this point -- not without simultaneously allowing Trump to clinch the nomination.

There are lots of possible outcomes to Rubio's continued presence in the race, but most of them range from bad to worse.  The most likely outcome, by a wide margin, is the worst of the batch. There is a possible good outcome in there (Rubio could conceivably win Florida while spoiling no other races), but it requires him to walk a very, very thin tightrope, and must be considered very unlikely. I don't think it's worth staking so much on that outcome. The likely costs of Rubio's continued presence in the race clearly outweigh the likely benefits.

Marco Rubio is a great American whom I supported in last week's primary and would be proud to support again in November.

For the good of his country, Marco Rubio should abandon his presidential bid immediately.

*Sorry, Puerto Rico.  You're great, but you're not a state.