Has the GOP Race Permanently Shifted to Trump?
For most of the year, the GOP primary could best be explained as a sort of collective action problem. Donald Trump seemed to have 35 percent to 40 percent of the Republican electorate locked down, but had difficulty expanding his coalition. The expectation was that he would perform well so long as it was a multi-candidate field. But as the field narrowed, the anti-Trump forces could be expected to mount a late comeback.
The upshot of this, however, was that the non-Trump candidates were incentivized to continue splitting that vote: The (reasonable) expectation was that the one who found himself (apologies to Carly Fiorina) as the last man standing would have a very good chance of becoming the nominee. So the field winnowed much more slowly than people expected.
The evidence for this was pretty strong. We can start with national polling. On January 1, when there were still 15 candidates (!) in the race, Trump was at 35.6 percent of the vote. On February 1, he was at 35.8 percent. On March 1, with the field narrowed down to five candidates, he was at 35.6 percent of the vote. And on April 1, with the field down to three, he was at 42.1 percent.
Contrast this with Ted Cruz, whose vote share increased from 18.6 percent to 31.7 percent during this time period, and John Kasich, whose vote share went from less than 2 percent to 19.3 percent. In other words, Trump received about 20 percent of the vote from the candidates who dropped out during this time.
This was also apparent from state level polling and exit polls. As I observed after Super Tuesday, a majority of the GOP electorate in Virginia, Arkansas and Texas found Trump to be an unacceptable nominee, yet Trump won in the former two states because those voters were split.
We can also see this in the head-to-head polling that has been conducted. For example, Trump won South Carolina by 10 points, but head-to-head polling with Marco Rubio suggested a one-point race. Trump won Mississippi by 11 points, but exits suggested a six-point race against Cruz (the state awarded delegates proportionally). Trump won Michigan by 11 points, yet he trailed Cruz by two in a head-to-head matchup.
This has been consequential. We can see it with Jeb Bush’s decision to stay in the race well after it was apparent there was almost no appetite for his candidacy, which froze donations and endorsements for much of the fall. We see it with Chris Christie’s Hail Mary attempt to take down Rubio at the New Hampshire debate (as I noted at the time, he might have done the GOP a favor). We see it with Rubio’s decision to stay in the race until Florida, probably costing Cruz victory in Missouri and North Carolina.
So when the race finally winnowed down to three, people thought that the anti-Trump coalition was about to have its due. And for a while, it looked like it might happen, with Cruz taking all of the delegates in Utah, most of the delegates in Wisconsin, and sweeping the delegate selection process in Wyoming, Colorado and North Dakota.
You’ll notice that most of the foregoing is written in the past tense. It seems now as though the race has shifted. After nearly sweeping New York and winning over 60 percent of the vote on April 19, Trump once again nearly swept all of the delegates last night. He now has about 950 of the pledged delegates, with another portion of the unpledged delegates announced in his camp.
We’ve also seen an uptick in his polling nationally, along with polls showing him in the lead in Indiana and California. The Cruz/Kasich vote share has declined a bit. All of this adds up to a reasonably compelling story that the Republican electorate is either (1) coming to terms with Trump, (2) deciding it doesn’t like the non-Trump alternatives, or (3) just growing tired of the whole business.
In truth, this isn’t necessarily an “either/or” proposition; I think there is some truth to all of these factors. But I also think a bit of perspective is needed. Trump was supposed to romp in these states. My projection has almost always included an absolute Trump sweep of the Northeastern delegates.
Now to be clear, I didn’t expect him to win 60 percent of the vote. Before chalking this up to a massive shift in the race, though, we have to consider a few things. First, as Nate Silver has noted, it isn’t unusual for winners to run away with uncompetitive races, as opposing voters just decide that it isn’t worth it to cast their ballots. Second, and more important, Trump’s best state prior to New York was Massachusetts, where he won 49 percent of the vote. If you give Trump Ben Carson’s votes, and add in 20 percent of the Rubio vote, you end up with 56 percent of the vote in the Bay State. That is still indicative of a shift, but not as dramatic of a shift as it appears at first blush.
What’s interesting is that we’re now moving from very favorable Trump states to less favorable ones. If you look at this map of election results, the clear tendency is that his vote shares drop off as you move westward. You can chalk this up to the prevalence of caucuses west of the Mississippi, or to the Mormon vote, or demographics, but it is real.
Will this hold? I don’t know, and I don’t think anyone honestly knows. I think Trump will do better in these states than he would have if they had followed Wisconsin directly, but his baseline in these states was also pretty low to begin with. We also don’t know whether this new Cruz-Kasich “alliance” in Indiana will backfire, or whether it will work.
Regardless, I do think we can say that the race has shifted in the past two weeks. Whether it has shifted enough to give Trump a clear path to the nomination is still an open question.