A silly leftover from yesterday to cleanse the palate. Dude, it’s happening.
Well … no, it probably isn’t.
American Research Group found Gov. John Kasich of Ohio trailing Trump in a two-way race, 42% to 48%, among likely Republican primary voters, with 9% undecided. Trump leads another possible rival, Sen. Jeff Flake of Arizona, 49% to 33%, with 18% undecided.
The results come ahead of Kasich’s third visit to the nation’s first primary state since ending his 2016 bid for president. He will speak Tuesday evening at New England College in Henniker. Flake, who is retiring from the Senate this year, made a trip to the state last month…
The more anti-Trump challengers in New Hampshire, the easier the path for Trump. In a hypothetical primary where Flake and Kasich both run, Trumps leads the three-way field with 51%, followed by Kasich at 34%, and Flake at 4%, with 11% undecided.
Look again at that last number. In a three-way race with Trump running as a populist and Kasich running as a centrist, leaving Flake the entirety of New Hampshire conservatives and libertarians as his natural base, Flake manages … four percent. New Hampshire is “quirky,” I realize, but that feels like a useful snapshot of the state of the conservative movement circa 2018.
Anyway. Time for Trump to panic, right? BuzzFeed notes in its story that the pollster in this case, ARG, did extremely well in calling the New Hampshire primary two years ago. Their final poll had the top two as Trump 33, Kasich 17. Actual results: Trump 35, Kasich 16. They know what they’re doing! If they say it’s Trump 48, Kasich 42 right now, that’s what it must be.
Perhaps. But here’s something from NYT data guru Nate Cohn written in January 2016, about three weeks before New Hampshire’s primary:
The American Research Group has an unusually lengthy record of high-profile misfires, including for almost all of the 2008 Democratic primaries and in the 2012 general election.
Its surveys often show a pronounced “house effect,” in which the results are consistently biased toward the same candidate (like Hillary Clinton in 2007 and 2008). So far this year, it has consistently shown Mr. Kasich doing better than he has been in other polls — which raises the possibility that the most recent A.R.G. poll might be an outlier on top of a tendency to lean toward Mr. Kasich more generally.
“House effect,” you say? Well, as noted, there was no house effect in ARG’s New Hampshire primary survey. They nailed Kasich’s share of the vote almost perfectly. In other pre-primary polls later that spring, however, ARG did indeed evince a strongly pro-Kasich house effect. Their final poll before the South Carolina primary had him at 14 percent; he finished with just 7.6. In Wisconsin they saw him taking 23 percent of the vote; he ended up with 14.1. Days before the Michigan primary they predicted a momentous upset, with Kasich nudging past Trump for a narrow victory, 33/31. In reality he finished third with 24.3 percent, half a point behind Ted Cruz for second. That’s a “house effect” for you.
Oh, almost forgot: ARG also polled New Hampshire last August, trying to gauge Kasich’s strength against now-President Trump in a 2020 primary. They saw a 52/40 lead — for Kasich. The new poll showing Trump ahead 48/42 represents a net 18-point swing *in Trump’s favor* among Republicans in the past seven months. That should be the headline, that even GOPers in states that are lukewarm to Trump have gotten more comfortable with him over the past half-year.
I think the two polls are fair evidence, though, that Kasich could poll enough of the vote in certain states in 2020 to embarrass Trump in a primary even if he didn’t win. And he probably wouldn’t win any (except maybe Ohio): Some Republican voters who are on the fence would end up tilting towards Trump if only to minimize the humiliation to a sitting president who they know would inevitably win the nomination and end up leading them into battle in the fall. What if Trump won 70/30, though? 60/40? It’s not nuts to think Kasich might pull a third of the party as a protest vote, especially if the economy slows down in the next two years. But that also assumes he’d choose to run in the primary as a Republican instead of skipping it and running in the general election as an independent. That seems like a smarter play to me. And if he does, he could end up hurting Trump badly by pulling just enough votes from the right in Rust Belt states to turn Trump’s razor-thin margins of victory in places like Michigan last time into margins of defeat.
Semi-serious exit question, then: If Kasich looks like he’s going to make life harder for Trump in 2020, should the White House consider offering him the VP slot? Conservatives would hate seeing Mike Pence replaced , I know, but so what? Conservatives are a marginal sliver of the modern Republican Party. Republicans are Trumpists now and will go along with whatever Trump wants to do. Meanwhile, replacing a conservative like Pence with Kasich would give Trump more centrist cred (and more Rust Belt juice) in the general election, particularly if the Democratic nominee is hard-left. Trump would have purchased the quiescence of a challenger and probably improved his chances of reelection in one fell swoop. And for all his palpable disdain for Trump, Kasich might be tempted to say yes since it would make him an instant frontrunner in 2024, when he’ll be 72 — old, but probably not too old for the presidency nowadays. How about it?