Sabato's Crystal Ball

For the Anti-Trump Forces, It May Be Now or Never

Meanwhile, Sanders gains on Clinton outside the South

Kyle Kondik and Geoffrey Skelley, Sabato's Crystal Ball March 14th, 2016

facebooktwittergoogle_plusredditpinterestmail

As we’ve suggested, the past few weeks have been defined by increasingly-loud talk of a contested convention and the possibility that the presidential contest will go beyond the first ballot, something that has not happened in either party since 1952. The highly unusual circumstances on the Republican side, where the polarizing Donald Trump has finished first in the majority of contests so far and has won more than a third of the delegates he needs for a first ballot nomination, make the outcome impossible to predict with precision at this point.

But wishful thinking has characterized some of the anti-Trump fervor. Yes, many Republicans detest Trump, and he does not appear to have anything close to solid majority support in the Republican Party. Yet he still has been getting about 35% or more of the vote in many places, and his level of support is geographically durable: He just added a convincing Midwestern victory, in Michigan, to his trophy case. And now, on Tuesday, there are two true winner-take-all contests in Florida and Ohio, in addition to potentially winner-take-most contests in Illinois and Missouri and a proportional delegate race in North Carolina.

While unlikely, it is not out of the question that Trump could sweep all five states. And if he does, we don’t see a practical way to stop him from winning the nomination, save for changing convention rules and guaranteeing a massive implosion in the GOP.

If Trump pulls off the quintuple play, he probably will have knocked Marco Rubio and John Kasich from the race, leaving Ted Cruz as the only one who could stop Trump. Cruz would find himself in this head’s up duel just as the nominating calendar moves west and north. The next major contest is winner-take-all Arizona on March 22. Doesn’t Arizona, the home of Trump’s champion, Maricopa County Sheriff Joe Arpaio, seem optimally suited to Trump’s anti-immigration message?

Future major contests include several Northeastern states, where Trump has posted some strong performances. Is a very conservative, religious candidate from Texas the best candidate to defeat Trump in these states? Anything could happen, but we have our doubts.

What we’re saying here is that March 15 has the potential to be D-Day for the Republican race. Unless someone steps up and defeats Trump.

John Kasich in Ohio seems like the best bet to beat Trump anywhere on Tuesday. Polls show Kasich running neck-and-neck with Trump, but as a popular governor we have to believe he will get a further home-state bonus that might not be reflected in the polls.

The Kasich campaign is a hard one to figure out. On the one hand, he has accomplished relatively little in this race. He has not won any states, and in 18 of 27 contests he has finished below 10% of the vote. On Super Tuesday, he ended up behind the now-former candidate Ben Carson in seven of 11 states. Yet Kasich, a solid experienced officeholder who has run a positive campaign, has endured in the race while bigger-name, ideologically similar candidates like Jeb Bush and Chris Christie have fallen by the wayside.

Kasich’s path in Ohio likely would mimic Mitt Romney’s in his narrow primary win over Rick Santorum four years ago. Santorum won most of the state’s counties but Romney captured the state’s three biggest, Cuyahoga (Cleveland), Franklin (Columbus), and Hamilton (Cincinnati). Kasich’s home base is in Central Ohio — he represented part of the region in the House for 18 years in the 1980s and 1990s — so he should be able to outperform Romney in that part of the state. Trump (and Cruz to a lesser extent) should do very well in Appalachian Southeast Ohio: Trump has generally dominated Appalachian counties in other states.

While many have pondered a Rubio-Cruz-Trump final troika in the GOP race, if Kasich wins Ohio but Rubio loses Florida, it’s possible that Kasich could replace Rubio in that grouping. But even with an Ohio victory, Kasich has little practical path to winning a majority of delegates during the primary season, yet Republican Party leaders and establishment delegates — the kind of people who might do some wheeling and dealing at a contested convention — would almost assuredly prefer Kasich to Cruz or Trump. That’s Kasich’s glimmer of hope.

Rubio’s star has dimmed considerably, and he appears less likely to win Florida than Kasich is to win Ohio. Part of his problem is a demographic one: Florida probably will have the oldest electorate in this primary season — it did in 2012 — and Trump has generally done better with older voters. A Rubio win in Florida would not be as big of an upset as Bernie Sanders’ recent Michigan win, but it would be close. Trump is clearly favored in Florida. And that is amazing, given Rubio’s past service as Speaker of the Florida House of Representatives and his tenure as one of the state’s U.S. senators over the past five-plus years.

One hidden factor that might help explain why Kasich probably has better odds in his home state than Rubio does in his is the lineage of the residents in their respective states. About three-quarters of Ohio’s residents were born in the state. Kasich is actually part of the quarter who was not — he’s from the Pittsburgh area — but he is a well-liked governor of a state whose residents might feel some pride in voting for one of their own (in that, the endorsement from Ohio State University football coach Urban Meyer on Thursday couldn’t have hurt — Buckeye football is a religion statewide, though particularly in Kasich’s Greater Columbus base).

Meanwhile, Florida’s electorate has less clear ties to the state: Only about a third of Florida residents were actually born there. While many Florida Republicans have certainly voted for Rubio before, they might be from the Northeast or Midwest or some other place and probably don’t care as much about a native son becoming the GOP nominee. Additionally, for much of this race there were two Floridians who had been elected to major offices statewide — Rubio and now ex-candidate Jeb Bush — which divided home state loyalties.

Again, this is a hard thing to quantify, but if Kasich outperforms Rubio, the composition of their home-state electorates might be a reason why.

Here are a few quick observations on the three other major March 15 Republican contests:

— Illinois is likely to have a less conservative electorate than many of the states we’ve seen vote so far, akin to Michigan’s in many ways. Considering his performance among moderate and somewhat conservative voters thus far, Trump looks like the favorite. He’s also led the polling we’ve seen out of Illinois. Despite the Chicago meltdown at the Trump rally, the Land of Lincoln overall could be hospitable to him, and unlike Ohio there is no home-state favorite on the ballot. Cruz has been inching up here in polls: Is an upset brewing? Illinois is not winner-take-all, but the statewide winner typically takes most of the delegates.

— On March 1, Trump found success in Virginia, a state that is somewhat similar to North Carolina from a political perspective. Along with Florida, the Tar Heel State is the last in the South to hold a primary or caucus this cycle, and so far Trump has won every Southern contest east of Texas. So the real estate magnate almost certainly is the favorite in North Carolina, though perhaps Cruz could surprise. The downside for Trump is that the almost purely proportional rules in Tar Heel country will likely make it difficult to win a substantial delegate edge.

— Missouri’s GOP primary will probably be at least 50% white evangelical (as it was in 2008), and how those voters split could be pivotal. If they break more for Ted Cruz, it could present him with an opportunity to win one March 15 state. The good news for Trump is that Missouri is an open primary state, and thus far he’s won every election in such states outside of Cruz’s Texas. Geographically, the Show Me State is once again a Border State, except this time it’s between Trump and Cruz country: The Texas senator has claimed three states that border Missouri (Iowa, Kansas, and Oklahoma), but so has the New York billionaire (Arkansas, Kentucky, and Tennessee).

Is a contested convention ahead? It could easily happen, but we’re at the point where the non-Trump candidates have to earn it. They might not get a better chance than on Tuesday.

The Democrats: Clinton missed her chance at a knockout

If Hillary Clinton had won Michigan convincingly, as almost all polls erroneously suggested she would, and then followed up with a March 15 sweep of Florida, Illinois, Missouri, North Carolina, and Ohio, she would have had a compelling case to make that the Democratic contest was effectively over. And she would have been right, so long as the FBI moved on from its investigation into her use of private email while secretary of state.

Now she heads into “Titanic Tuesday” still holding a commanding position in the race, but the door is open for Sanders to continue fighting for the remainder of the primary season, turning Clinton’s once-expected coronation into an agonizing slog.

Michigan proved that Sanders could win a big, industrial, non-caucus state against heavy odds. It’s possible that Michigan was just a fluke: After the results came in, the Clinton campaign told some reporters that they might have taken the state for granted. And Sanders did run significantly more TV ads than Clinton did in Michigan.

However, the Michigan upset could suggest bigger problems for Clinton. Crystal Ball Senior Columnist Alan Abramowitz looked at the prior results on the Democratic side and developed a model that explains the results. To paraphrase — and we encourage you to check out the whole piece, which we published late last week online — Clinton has cleaned up in the South thanks in large part to the region’s heavy African-American population. Sanders has mostly competed impressively with her outside of the South, and he has done particularly well in states that are less diverse.

Abramowitz’s model shows that Sanders has a decent chance to win Missouri and Ohio, and an outside shot at an upset in Illinois.

Table 1: Predicted vote shares and winners of March 15 primaries according to Abramowitz model

Polling over the last several days has largely conformed to Abramowitz’s model. Clinton retains big leads in Florida and North Carolina, while Illinois, Missouri, and Ohio have all gotten close.

Clinton’s once-big lead in Ohio polls has nearly disappeared, and the state is similar enough to Michigan to raise questions. However, Michigan has a stronger liberal Democratic tradition than Ohio, and Ohio’s Democratic electorate is more dispersed throughout the state than in Michigan. Last Tuesday, more than half of the Democratic vote was cast in Wayne County (Detroit) and its neighboring counties. By contrast, in 2008, Cuyahoga County (Cleveland) and its neighbors cast only about 30% of the total Democratic vote. Clinton did well in Greater Detroit, but not as well as she needed to do because Sanders swept much of the rest of the state. Will the more geographically and culturally diverse Ohio be harder for Sanders to organize? We’ll soon see.

Ultimately, a Sanders win in Ohio would be a surprise despite Abramowitz’s model and the tightening polls, but not a shocking surprise like Sanders’ win in Michigan was.

Just as on the GOP side, we don’t have much information about the Democratic contest in Missouri. The state has something of a Southern character despite being a Border State, and Sanders isn’t a great ideological fit for its voters. But that didn’t stop Sanders from winning Oklahoma, for instance, which is a very conservative state, more so than Missouri. The Show Me State is probably Sanders’ best shot at a victory: In fact, we think he has a better than 50-50 chance to win it.

Clinton should run up big margins in Florida and North Carolina — anything less than 60% would probably be a disappointment, given that Clinton’s worst Southern performance was 64% in Virginia — and while Clinton remains slightly favored in Illinois, late polls showed Sanders closing (and one even showed him narrowly leading). Sanders’ attacks on unpopular Chicago Mayor Rahm Emanuel and the clash between Sanders and Trump supporters on Friday night might be giving the Vermont senator some late momentum. Illinois, like Michigan, is dominated by a single metropolitan area (Chicago). If Sanders can hold down Clinton’s performance in Chicagoland, downstate could deliver him the statewide victory.

We’ll get a bit more into the calendar after Super Tuesday, but the schedule decompresses after March 15 and features many small states with caucuses that should be very favorable for Sanders. Clinton should continue to expand her delegate lead on Tuesday, but no matter what happens, Sanders will likely fight on. Will that make Clinton a better general election candidate — or stoke the fires of dissent on the left?