Leave a comment, and send us questions @FiveThirtyEight.
UPDATE (March 16, 1:03 a.m.): NBC News is calling Clinton and Trump the “apparent” winners in Missouri.
UPDATE (March 16, 2:05 a.m.): We recorded a late-night podcast breaking down the night’s results. Take a listen to that here.
That does it for us, everyone. Thanks for sticking around. (If you want to relive it all chronologically, click here and scroll up through the nerdy goodness.) We’ll have a lot of analysis of tonight’s results for you to pick through along with your morning eggs and bacon. But as we shutter the live blog, there are two outstanding races, both in Missouri:
- The Democratic contest in the Show-Me State remains very close. With 99 percent of precincts reporting, Clinton leads by 0.2 percentage points. There are still some votes remaining to report in Jackson County (Kansas City), where Clinton is holding about a 10 percentage point advantage. It looks like she could eke out a win.
- The race in Missouri on the Republican side is equally tight, with Trump ahead by 0.3 percentage points. But Cruz and Trump are about equal in Jackson County and St. Louis County. Cruz needs a plot twist in the remaining vote to win the state.
Get some sleep.
Our colleagues at ABC News have also called Illinois for Clinton.
OK, it’s 12:30 a.m. on the East Coast, and Missouri’s barn is still burning. Trump is clinging to a lead of just 2,111 votes (0.3 percent), but I’m not sure that Cruz can overcome that with what’s out. Many of the remaining precincts are in St. Louis City, which shouldn’t produce many GOP votes. There are some other outstanding precincts in St. Louis County and Kansas City’s Jackson County, but neither batch looks likely to net Cruz a ton more votes than Trump.
Fox News and CNN just called Illinois for Clinton. She’s now guaranteed to go four for five on the night, and five for five is still in play depending on those last few votes in Missouri.
Consider this: Up until today, Trump had won 42 percent of all GOP delegates with just 34 percent of the vote. By my estimate tonight, it appears he’s on track to win about 67 percent of the delegates at stake, even though he’s averaged only a hair over 40 percent of today’s votes. That’s a huge disparity. The reason: Today’s transition from proportional allocation to winner-take-all.
From now on, even tiny Trump pluralities (like the one we’re seeing just barely in Missouri, barring some magical Cruz precincts) will net him enormous shares of delegates. Up until today, only 5 percent of all GOP delegates were awarded on a winner-take-all basis. Between tonight and the final primaries in June, 64 percent of GOP delegates will be allocated on a winner-take-all basis. That’s an enormous catalyst for Trump’s drive to 1,237.
The Missouri Democratic primary is something to behold. Sanders’s margin over Clinton in the state is down to less than a percentage point, and plenty of the remaining vote is in St. Louis, where Clinton is leading by 10 percentage points.
No one is totally sure why this works, but we have yet more evidence from tonight’s returns that Google searches for candidates in states that are voting are decent indicators of how those states will vote. They showed Kasich’s relative strength in Ohio, Cruz’s in Missouri and Rubio’s in Florida (relative to his performance in other states, not to Trump’s in Florida). This is still a very new area of study, and it remains to be seen how best to convert Google search numbers into vote shares, where it works and where it doesn’t, why it isn’t as useful on the Democratic side, and — the big question! — why searches are predictive of votes.
David and Carl, I think the #NeverTrump forces are also too disorganized at this late stage of the primary game. There was a lack of will to coalesce around Rubio back when he seemed like the golden ticket to Trump Neverland. Cruz has made many enemies among establishment Republicans, with Lindsey Graham going on a massive rant about what it would take for him to hold his nose and vote for the Texan. Kasich won Ohio tonight, but PBS reports that he would need to win 91 percent of the remaining delegates to win the nomination outright. I’ve been to many conventions, including the 2000 Democratic convention in Los Angeles, where police and demonstrators clashed violently, and a reporter friend of mine got accidentally clipped by an officer on horseback. If 2016 goes to a contested convention, I’m buying used riot gear.
So far, Trump has won 37.1 percent of the votes throughout Republican primaries and caucuses. That percentage is up tonight after Trump had strong results in Florida and other states. And it could climb further in subsequent states, especially with only three candidates remaining in the race. But the percentage is still on the low end by the standards of previous nominees. Since primaries became widespread in 1972, only George McGovern won his party’s nomination with a smaller share of the vote — just 25.3 percent, with McGovern winning by taking advantage of delegate rules that he had helped to write.
|NOMINEE||YEAR||PARTY||POPULAR VOTE SHARE|
Guess what? The last time a single state had two primaries where candidates were within 1 percent on the same night was probably … MISSOURI! In 2008, when Obama and John McCain won by less than two points.
We don’t know that tonight’s races will be that close, but they’ll certainly be tight.
CORRECTION: An earlier version of this post gave an incorrect winner for the 2008 Democratic primary in Missouri. It was Obama, not Clinton.
How unusual is it for a single state like Missouri to have two such close primaries in one night?
On a night when Hillary Clinton is likely to win four-of-five contests and extend her delegate lead over Bernie Sanders, pundits told Clinton to smile and criticized her for shouting.
Unfortunately, this reflects all too well a bunch of discouraging findings about women and the pursuit of political office. A study of newspaper coverage in 1998 found that readers were “more likely to read about female candidates’ personal qualities,” while news coverage was more likely to emphasize male candidates’ policy positions.
Recent scholarship has also focused on the supply of women candidates who enter politics in the first place. Richard Fox and Jennifer Lawless found that among a pool of qualified candidates, women were substantially less likely to consider or run for office and link this to how women are socialized. An even more arresting finding came from Kristin Kanthak and Jonathan Woon last year. In an experimental study, they found that women were specifically averse to competing in elections, even when they would volunteer to represent their groups in a random-selection process. When we see how women candidates are treated, it’s not hard to imagine why women might think twice about running for office.
David, if you’re right and Trump is on a clear path to 1,237 delegates and the nomination, the people cited in this Slate article will see their stock rise. They said a Trump nomination was realistic, even probable, when others were dismissing it. Although how definitively they stated the case for Trump; whether they got lucky or were prescient; and whether they’ll even prove to be right is hard to say.
Unless Cruz can pull a rabbit out of his hat, Trump’s going to win Missouri, based on the map. Most of Boone County (a Cruz stronghold) has reported, and Trump still leads by about 3,000 votes. We’ll see, though, as this election cycle has been crazy.
Here’s my gut takeaway from tonight’s primaries: It’s more difficult to see how Trump DOESN’T get to 1,237 delegates from here. He’s beating Cruz in red states, and he’s likely to beat Kasich in future blue states. Most of the delegates at stake from here on out will come from winner-take-all states. Whether #NeverTrump forces realize it or not, they are losing.
Right now, Trump is holding on to a 0.4 percentage point lead in Missouri. My guess is that the lead will probably hold, though it will be close. The question is whether Cruz can combine a big win on the remaining votes in Boone County (Columbia) with improving his showing in Jackson County (Kansas City) and St. Louis County.
The 66 delegates Kasich won in Ohio tonight are a big prize, but Kasich’s continued presence in the race has more ambiguous effects for Republicans hoping to stop Trump.
According to the estimates we developed last week, Cruz would be leading Trump in Missouri by about 4 percentage points in a two-way race. Cruz would also be leading by about 2 percentage points in North Carolina and trailing Trump by about 2 percentage points in Illinois.
Guess what? Trump could make up for all 66 delegates he lost in Ohio with huge delegate margins in Illinois and Missouri. In Missouri, Trump is clinging to a lead of just 2,400 votes, but IF things continue as they are, Trump will capture 47 of Missouri’s delegates to just five for Cruz. And in Illinois, where Trump is winning about 40 percent of the vote, he could win all but a handful of congressional districts, giving him perhaps 60 of the state’s 69 delegates. Wow.
In his speech, Cruz pledged never to compromise on Americans’ “religious liberty.” What exactly does that mean? Republicans have increasingly been using “religious liberty” to refer to opposition to abortion rights and the ability of small-business owners to refuse to provide services for same-sex weddings. In fact, the number of mentions of “gay marriage” and its variants during GOP debates has decreased since the 2008 primary season, while mentions of “religious liberty” have jumped from zero in the 2008 primary season to 29 so far in the current election season.
I’m not sure we’re going to see a tighter Republican primary all season than Missouri’s. All the vote counts I’m looking at right now have Trump and Cruz within 0.5 percentage points of each other, but Trump is winning.
Illinois and Missouri share a substantial border. Both are Midwestern states and have a mix of urban and rural areas. But politically, they are quite distinct. Missouri looks much more like the mix of a Western and Southern state that it is. In terms of presidential and Senate elections, Illinois looks more like the upper Midwest — or the other states with major cities that dominate their major elections. Illinois has voted for the Democratic candidate in every presidential election since 1992, and Missouri has gone for the Republican since 2000. But a substantial portion of Illinois is rural, Republican and pretty far south.
The election is now too close to call in Missouri, with Cruz having just taken the lead. Illinois has been called for Trump. This fits perfectly with the geographic pattern that has emerged over the past few weeks (and that I noted earlier this evening): Trump has done well in the Northeast and in the South — Illinois is kind of a weird mush of both, and if you look at this map, his areas of strength were around Chicago and the parts closer to Indiana and Kentucky. Missouri, on the other hand, may look more like the Western states that have gone so far, leaning instead toward Cruz. (Although one caveat here — the parts of Illinois that border Missouri also appear to have gone to Trump.) With the South and the interior West (plains and mountains, minus the Pacific Coast) making up the core of the Republican coalition, a possible geographic split is as significant as any other divisions we might observe.
There’s a lot of angst out there, particularly among Republican voters, and it’s showing up in exit polls. Perhaps not surprisingly, Trump is doing well with the more discontented voters:
|Are falling behind financially||58%||46%||50%||49%||51%||51%|
|Are very worried about economy||—||45||46||45||42||44|
|Believe trade takes jobs||—||50||—||47||44||47|
|Are angry at federal government||59||53||49||53||51||54|
|Favor of an outsider||75||69||66||65||67||69|
|Favor deporting undocumented immigrants||63||58||53||53||57||58|
|Support banning Muslims||59||49||49||50||49||52|
We’ve mentioned 1,237 a half-dozen times already in this live blog. That’s the number of delegates a Republican candidate would need to clinch the nomination. Online search interest in the term is on the rise, according to Google Trends. As fans of crucial presidential election numbers, we’re happy to see the rise in awareness of 1,237 while remaining confident that the number of electoral votes (a number very important to us) will overtake it in public awareness come summer.
Cruz, speaking while the race in Missouri is too close to call, is making a conservative platform speech heavy on details. In the weeks ahead, one question is, with Rubio out of the race, will Cruz be able to break through not only with voters, but also with volume of press coverage, which arguably has been a huge factor in this election. For example, earlier tonight Trump complained about the barrage of negative advertising that came out before this primary … not that it seems to have had much effect. But it’s also worth noting that although Trump has spent $10 million on advertising, he’s received the equivalent of $1.9 billion in free media, according to The Upshot’s analysis of mediaQuant and SMG Delta data.
It’s not surprising to see Cruz going out of his way to praise Rubio; picking up those voters will be key to any chance Cruz has at the nomination. And Cruz has some good news on that front; exit polls in Michigan and other states have shown Rubio voters preferring Cruz by 4 to 1 over Trump. The complication is that those exit polls didn’t ask about a three-way race among Cruz, Trump and Kasich. Although Cruz is probably a better match for Rubio voters ideologically, Kasich’s voters have more in common with Rubio’s demographically.
Bottom line: This is a great night for both front-runners, Trump and Clinton, with Cruz and Sanders left clinging to hope for a win in Missouri. A Sanders victory in the Show-Me State would be more of a symbolic, morale-boosting victory, but not meaningful in terms of the math. A Cruz victory in Missouri would be much more meaningful: He’d grab 12 statewide delegates from Trump and maintain some momentum heading into the crucial primaries to come.
As a native Midwesterner, I’m a fan of the Show-Me State, but Missouri is a fairly annoying state to be coming down to the wire. It’s quite slow to count its vote. Furthermore, there can be huge differences in the vote among St. Louis, Kansas City, the suburbs, and the rural parts of the state, making it hard to make extrapolations. Partly as a result of this, the Associated Press prematurely called Missouri for Clinton in the 2008 Democratic primary; it eventually went for Obama after further votes from St. Louis were tallied after midnight.
The name of the game is congressional district wins in Missouri. If you win a congressional district, you win five delegates. Right now, Cruz and Trump are leading in three each. Two are not yet reporting.
Democrats have to defend just 10 seats, to 24 for Republicans. Strickland, Duckworth and Ross are the picks to challenge three vulnerable GOP incumbents.
Something to watch for Wednesday and beyond is for the Supreme Court vacancy left by Justice Antonin Scalia’s death last month to become more important as a campaign issue. Reuters reported tonight that President Obama is likely to announce either Judge Sri Srinivasan or Judge Merrick Garland as his pick for the court, as soon as Wednesday. Srinivasan and Garland serve together on the U.S. Court of Appeals for the District of Columbia Circuit, and both have some bipartisan appeal.
Sure, it matters. For one thing, there are 12 delegates for the statewide winner in Missouri. And it could also be important to how people interpret the results in the coming days. In contrast to some previous primary election nights, when Trump won a lot of headlines without necessarily doing all that well in the delegate count, this night has been sort of the reverse. He could have a pretty good delegate night even though the media coverage has been pretty equivocal. If people wake up in the morning and see that he’s gone four out of five instead of three out of five, that could make some difference at the margin.
Nate, does it matter at all if Cruz narrowly wins Missouri — which awards its districts at the congressional level — and Trump goes three for five on the night, or if Trump wins narrowly and he goes four for five?
After a disaster in Michigan, the polls look as though they’ll have a pretty good night on the Democratic side.
In Ohio, Clinton is projected to win by 15; she led by 11 in the polling average there.
In North Carolina, Clinton is projected to win by 18; she was up 22 in the polling average.
Illinois has not been called, but The Upshot projects Clinton to eventually win by 4 percentage points. She led by 7 in our polling average.
And Missouri, where Clinton led by less than 1 percentage point in our polling average, is projected to be just as close in the actual vote.
Where do Rubio’s delegates go?
Are they released and free to vote for whomever they choose? Nope, that would be too easy. What happens to a departed candidate’s delegates depends on the delegate’s state, and those details are fuzzy and complicated.
For example, Rubio’s seven now-released delegates from Kentucky must convene in a meeting with bound delegates in which they will vote in a secret ballot to reallocate Rubio’s delegates to another candidate. Some states release delegates to support whomever they choose at the convention (for example, New Hampshire and Tennessee). Other states continue to bind delegates to the withdrawn candidate (Iowa), or they reallocate the delegates among the remaining candidates as if the withdrawn delegate had never qualified (Alaska).
There’s really just no simple, pithy way to accurately summarize what happens to delegates who are released — even in general terms.
According to ABC, Rubio had 163 delegates at the time of his withdrawal (our delegate tracker reports a slightly different number) and 23 are still bound to Rubio. What happens to the rest of his delegates will prompt the first large-scale delegate fight of the election season.
Cruz has just closed a little bit on Trump in Missouri. He’s now down just 1.5 percentage points. Perhaps more importantly, he’s now barely leading in the Kansas City metropolitan area. He also seems to be closing in St. Louis County. There’s still a lot of vote out around Columbia, where Cruz is winning by 11 percentage points. The bottom line is that this race is very close.
In stark contrast to last week’s steak salesmanship, Trump’s speech tonight sounded like the Trump version of a mainstream political speech. He mentioned deals but contrasted them with “the politicians can’t make an agreement” — a Trumpish riff on Obama’s “changing the tone in Washington” argument. He stayed focused on trade in the beginning of the remarks, and then endorsements, plus turnout and enthusiasm, and he mentioned bipartisanship at least twice. His praise for campaign manager Corey Lewandowski was brief and pointed. Basically, I’ve concluded that Trump stays in the news cycle by vacillating between doing something really outlandish, like advertising steaks or talking about his anatomy, and then has a subdued debate and gives a rambly but fairly mainstream speech.
Trump also fits pretty well into another political science theory, political time. Put simply, political cycles go on for about 40 years. They tend to end with unsuccessful presidencies — think James Buchanan, Herbert Hoover, Jimmy Carter. One hallmark of these presidencies, especially in the 20th century, is that they tend to emphasize tinkering and getting things done rather than new big ideas (in contrast with those who start new cycles such as Lincoln, FDR, and most recently Reagan). Trump’s speech stressed his plans to make things work — health care, borders and, naturally, deals.
I’ve noticed that Trump often mentions corporate inversions in his list of business practices to which he will put an end, as he did in his speech tonight. Inversions allow companies to be acquired, in name only, by overseas corporations and thus dramatically lessen their tax burden.
“We cannot continue to allow our nation’s wealthiest corporations to avoid paying their fair share of taxes. These companies stash tens of billions of dollars in overseas tax havens while at the same time receiving billions in subsidies.”
But wait: that quote didn’t come from Trump’s speech. It’s from Bernie Sanders’ website. In fact, there is a subset of Sanders voters who would rather vote for Trump than Clinton — 7 percent according to a recent NBC News/Wall Street Journal poll.
The candidates’ strong positions often diverge, but this is one place where their two brands of populism meet.
Exit poll results are coming back and telling us a bit more about why Clinton won in Ohio and not in Michigan — her victory can be boiled down to her performance with white voters, older voters, and voters who are concerned about trade issues.
In Michigan, Sanders won white voters, but in Ohio, Clinton won them, 51 percent to 48 percent, and that seems to have made a difference. Her support with black voters in the state also remained strong — she won them 68 percent to 30 percent — but that was less an overwhelming win of the demographic than her record in other states.
It’s well-known that Sanders performs well with young people — there were fewer youths (ah, the youths) in Ohio. Sixty-two percent of Ohio’s electorate was 45 and older.
The last leg of the Clinton victory triangle is that she won on the all-important-in-the-Rust Belt trade question: Voters who said trade with other countries takes away U.S. jobs supported her, 53 percent to 46 percent. Sanders won that bloc in Michigan, 56 percent to 41 percent. It’s not quite clear what happened in the intervening week — our Andrew Flowers noted during the last Democratic debate that the Vermont senator seemed to be hammering the anti-NAFTA message less than usual, but it might also just be that Michigan was an aberration.
I’d be very careful with delegate projections in Illinois. Take a look at the 5th Congressional District, part of which is in Chicago. Trump delegates are winning two of the delegate slots in the Chicago part of the district, but one of Kasich’s delegates is actually winning the third slot. The loophole primary may have a few surprises yet.
Did the events of the past week, including the violence at Trump events in Chicago and other places, wind up helping Trump? Perhaps. On average among the five states voting today, Trump won 31 percent of late-deciding voters, according to exit polls. That’s a good-but-not-great number: more late-deciders went to Cruz in Illinois, Missouri and North Carolina, and to Kasich in Ohio. But it’s better than Trump normally does with late-deciding voters; his numbers have often been in the teens or twenties in previous states.
|SHARE OF VOTERS DECIDING IN LAST FEW DAYS|
|STATE||SHARE OF ELECTORATE||TRUMP||CRUZ||KASICH||RUBIO|
The only place left that could prevent Trump from batting four for five tonight is the Show-Me State, where Cruz is giving him a serious challenge. Trump is currently ahead 43 percent to 40 percent, with 20 percent of precincts reporting. Keep in mind, the statewide winner will take 12 delegates, while the winner of each district will claim five. The good news for Trump is that his lead across the state appears fairly geographically balanced, even though much of metro St. Louis and Kansas City have yet to report.
The major exception to Trump’s lead is Springfield and southwest Missouri, the state’s “Bible Belt,” where Cruz is beating him. (Springfield is the home of the Assemblies of God, which says it is the world’s largest Pentecostal denomination.) That means Cruz is likely to net at least five of Missouri’s 52 delegates. It’s still early in the vote count, so we’ll need to find out more about where things stand in the state’s major metro areas. But Trump is doing well in similar metro areas in Illinois. Thumb on the scale for Trump, but it’s far from over.
Kasich is being portrayed as the moderate Republican left in the race, with Cruz far to the right and Trump off the charts somewhere to the side. By our measure of candidate ideology based on fundraising, public issue statements and voting, Kasich is more moderate than the average Republican in the 2013-14 Congress. By the standards of a previous era in Republican politics, Kasich is more conservative: He stands to the right of the average GOP member of Congress in 1979-80. And on some issues, notably curbing access to abortion, Kasich is quite conservative.
North Carolina has been called for Trump by NBC News. His 4 percentage point lead has held as more of the vote has come in.
Sanders won Michigan last week partly because he received a larger share of the black vote than he had in the South. And there were some signs again tonight that African-Americans in Midwestern states are more favorably disposed toward Sanders than those in the South. Based on exit polls, Sanders won about 30 percent of the black vote in Illinois, Ohio and Missouri, compared with 17 percent in North Carolina and 20 percent in Florida.
|SHARE OF AFRICAN-AMERICAN DEMOCRATIC VOTE|
Are the stars aligning for Trump? It’s increasingly possible that Ohio could be Trump’s only loss of the evening. If that happens, you might not have been able to script the day more perfectly for Trump. Not only would he sweep huge delegate hauls from Florida, Illinois, and Missouri, but Kasich’s Ohio win keeps the anti-Trump vote split in the next few contests. Moreover, Kasich’s Ohio win might prevent #NeverTrump activists from embarking on a campaign to get an independent on the ballot. If Trump were to cross the 1,237 threshold on the final day of primaries (June 7), it would be too late for such an effort.
Clinton is holding on to a 6 percentage point lead in Illinois. Much of Cook County, where Clinton was expected to do well, has already reported. Still, much of the raw vote from Illinois generally comes from Cook. The bottom line is that the race is too close to call, with Clinton leading.
It’s still very early in the count, but another embattled local prosecutor, Timothy J. McGinty of Cuyahoga County, Ohio, holds a lead of just 3 percentage points over challenger Michael C. O’Malley. Protesters called for McGinty to be voted out of office after he decided not to charge the police officers involved in the death of 12-year-old Tamir Rice in Cleveland. O’Malley has criticized McGinty’s handling of the Rice case but has declined to say whether he thought the officers involved in the boy’s shooting broke the law.
BEREA, Ohio — Kasich took the stage here tonight at about 9:20 p.m., and before he could begin speaking someone who seemed to be wearing a Trump hat caused a small disturbance. The Ohio governor said something about peaceful protest and moved on, giving a nod to Rubio, who has suspended his campaign.
Flanked by his teenage daughters, whom he often invokes when talking about the ugly tone of the Republican race, Kasich went with his typical peace on earth, goodwill to men message.
“I want to remind people again tonight that I will not take the low road to the highest office in the United States,” he said to cheers. “Your purpose is to live your life a little bit bigger than yourself,” Kasich said, “to be a center of healing.”
He also seemed to shy away from his executive experience record somewhat, or at least tried to reframe it. Government, he said, wasn’t about the “bigwigs.”
“You hire us,” he said.
Kasich ended on a general-election note.
“We’ve got one more trip around Ohio this coming fall, when we will beat Hillary Clinton.”
ABC News and NBC News have called Illinois for Trump. Call it a case of the anti-Trump vote being divided. Trump leads with 40 percent, while Cruz is at 26 percent, Kasich is at 22 percent and Rubio is at 10 percent.
Cruz still has a chance to win — in decreasing order of likelihood — Missouri and North Carolina. From a delegate-math standpoint, a win doesn’t matter all that much. North Carolina’s delegates are allocated strictly proportionately. Missouri awards most its delegates by congressional district, so Cruz could pick up some delegates there even if loses the state. So does Illinois, which at least one network has called for Trump.
But in terms of the media coverage of the next few weeks of the campaign, notching a win somewhere could be important for Cruz. My guess is that Kasich will get a lot of attention for having won Ohio, despite it being his home state and his not really having a path to the nomination before the convention. Trump got big wins in Florida and Illinois, meanwhile, while Rubio will make headlines by having dropped out. That’s not a lot of news cycles left for Cruz, who will need to make up some ground to claim winner-take-all Arizona next week.
We’ve got a close race underway in Missouri on the Republican side. Trump is leading by 4.2 percentage points over Cruz with just 14 percent of votes recorded.
Here’s a twist: Trump could still remain on pace for the 1,237 delegates necessary to clinch the nomination after tonight’s primaries, even after losing Ohio and its 66 delegates to Kasich. That’s because he still has a good chance of winning an overwhelming share of delegates in Illinois and Missouri, where he currently leads and which allocate their delegates on a winner-take-all basis by congressional district.
According to our tracker, Trump owns 463 delegates but would need 719 by the end of the night to remain “on track” for 1,237 delegates. This morning, Trump captured the Northern Marianas’s nine delegates, which bumped him up to 472. He just captured 99 from Florida, which gets him to 571. He could win around 40 percent of North Carolina’s 72 delegates, which would get him to 600. But if he were to sweep Illinois and Missouri’s 26 districts, he would net another 121 delegates, putting him at 721, just above the 719 mark.
To knock Trump “off pace” for the nomination, Cruz or Kasich will have to start winning some districts in Illinois and Missouri before the night is over.
As I watch Kasich’s speech, I can’t help but think how it sounds very corny and yet can win him votes. I could easily imagine Kasich being competitive in suburban districts in Pennsylvania, Maryland and New York where more moderate Republicans live.
The establishment’s worst nightmare? Ted Cruz is now leading the Republican field in endorsement points among the three remaining candidates.
Man alive, Kasich got a huge majority out of Franklin County in Ohio, which includes Columbus. He leads there with 61 percent of the vote. Franklin, of course, is where Kasich’s campaign is based.
So Jody asks whether we are baked in with three candidates now — Cruz, Kasich and Trump. I think that’s likely, in part because this is a two-way race with three candidates. Or maybe two two-way races. Let me explain. It’s not a big leap to suggest that there’s a Trump movement and a stop-Trump contingent. Got it. But a couple of weeks ago that looked more like a “stop Trump and Cruz” movement. As I noted in our March 8 live blog, Cruz is very, very conservative. In addition to his apparent lack of friendships in the Senate, there are real electability concerns. At the same time, Cruz has a significant following among conservative leaders. So there may be at least two kinds of races going on there — a “stop Trump” race and a “stop Trump and Cruz/nominate an establishment candidate” one.
The important question that remains is whether there’s a “nominate Kasich” movement. One possible lesson from the last eight months or so is that campaigns might actually matter. If this is true, then a single, focused “establishment lane” campaign could make a difference going forward. But if we simply see a “stop Trump” campaign in which Kasich and Cruz get in each other’s way, then it will probably be a two-person race pretty quickly. But not for very long.
Here’s one of my favorite recent FiveThirtyEight graphics, from contributor Milo Beckman. As Kasich goes into his speech, keep your eye out for some of his favorite phrases: “balanced budget,” “in Ohio,” “end of the day.”
I just saw someone in left-leaning media tweet out something to the effect that tonight could be the night that closes the coffin on the Sanders campaign, ending the storybook narrative. I think while people have been saying for a while, over and over, that the math really isn’t likely to work for Sanders, the lost thread of the story tonight — i.e., Michigan doesn’t sew up Ohio — is likely to deflate those who were convinced that there was still a shot.
I will say that a lot of voters I talked to in Ohio today who were voting Democratic said they liked the Sanders idea, but that Clinton is the most experienced candidate, if not the most personally winning. They said experience was a powerful factor for them.
Clare, Clinton may get a sweep tonight, but there are a lot of Sanders-friendly states upcoming on the calendar. So if Clinton sweeps tonight’s states, do you think the media will treat Sanders’s chances as kaput? (Delegate-math-wise, they basically are, right?)
Not to rain on Kasich’s parade, but it’s worth remembering that after tonight, he’ll have won one of 29 states to have voted so far, and that one state is the one where he’s a two-term governor.
If there was a big shift of momentum toward Sanders after his win in Michigan, it wasn’t apparent in tonight’s exit polls. On average among the five states that voted tonight, just 17 percent of Democratic primary voters decided their vote in the last few days, according to exit polls, and those who did split fairly evenly between Clinton and Sanders.
|STATE||SHARE OF ELECTORATE||CLINTON||SANDERS|
Does this election give you an adrenaline rush, a feeling of shock and triumph, or just a need to nerd out? If you’re experiencing any of these feelings, we have a prescription for you: a weekly dose of FiveThirtyEight newsletters for your inbox and podcasts for your ears. Side effects include increased nerdiness and political savvy.
Incumbent Anita Alvarez is losing big to Kim Foxx in early counting of votes for the Cook County state’s attorney Democratic primary. Alvarez is blamed by many in Chicago for her role in the city’s cover-up of video showing a police officer shooting and killing 17-year-old Laquan McDonald, and her office repeatedly has declined to file charges against police officers who fatally shot people.
Dave’s right that Trump is leading in most of the districts in Cook County — except one! I’m currently on the Chicago Board of Election Commissioners’ website, and in the 7th Congressional District (only part of which is in Chicago), Kasich delegates are holding a lead over the Trump delegates with 51 percent of precincts reporting. Trump is leading in the rest of the districts in the city, which is in Cook County.
All right, this is going to get nerdy, but it’s absolutely critical. There are very few GOP voters in Cook County, Illinois (Chicago), but the best news of the night for Trump so far might be that he’s winning it handily. That’s because there are three delegates at stake in each of Illinois’s 18 districts, regardless of how blue or red they are. In fact, Trump could win three delegates from a Chicago district with as few as 5,000 votes but then collect zero delegates from a Cruz-won downstate district with as many as 50,000 votes.
In early returns, Trump is taking 44 percent in Cook County to 23 percent for Kasich and 21 percent for Cruz. If that holds, he could win a bigger delegate margin out of the city of Chicago than he would win with a 15 percent margin of victory in the entire state of North Carolina. It’s a fascinating quirk of the Republican National Committee’s delegate allocation rules, and it wasn’t simply an accident or sign of provocation that Trump planned a Chicago rally last week.
It’s been a rough election for governors-turned-presidential-candidates. Most of them never connected with voters and dropped out. Kasich’s Ohio win might end up being the only state victory for a current or former state executive.
Sanders could still pull out wins in Illinois or Missouri, but with North Carolina, Ohio and Florida all having been called for Clinton, I’m reminded about what I said in our chat earlier today. So far, the primaries have often defied momentum. After a shocking loss in Michigan last week, Clinton is poised to have one of the best nights of her campaign.
ABC News and other networks have called Ohio for Kasich. It’s not a huge surprise — Ohio polls broke toward Kasich at the end, and he was a fairly heavy favorite in our “polls-plus” forecast. Still, that’s 66 delegates that Kasich has taken off the board from Trump and Cruz, increasing the probability that nobody will achieve a majority before the convention.
Kasich needs a contested convention to win, so what does he do next? One answer is to stay out of Cruz’s way. That might mean focusing on northeastern states such as New York, Connecticut and Pennsylvania, where Cruz has little chance to win but Kasich has strong favorability ratings.
Wisconsin, which votes April 5, could be a trickier call because the state’s delegates are awarded winner-take-all (some by congressional district and some based on the statewide vote). It’s easy enough to imagine both Kasich and Cruz finding pockets of support in Wisconsin, but Trump winning the state with 40 percent of the vote while Kasich and Cruz have 30 percent each.
In the projected delegate spreadsheet I keep for myself, I had a lot of Illinois’s delegates going into Trump’s column. Why? I thought he would do very well in the congressional districts around Chicago. And he’s leading by 20 percentage points in Cook County, which includes Chicago.
Trump and Cruz are now one candidate closer to their desired one-on-one matchup. It’s tempting to look at national polls to see how Trump and Cruz compare in voter support: Trump leads Cruz by 38 percent to 22 percent, according to our national polling average, while Cruz beats Trump head to head in some polls of their hypothetical matchup.
But after each state or territory holds its primary or caucus, its voters can no longer help either candidate. And Cruz is weaker in places that haven’t voted yet than in places that have.
Trump led Cruz by 17 points in places with votes on or before March 15, according to data provided by the online-polling company SurveyMonkey, based on its interviews of 8,624 Republican registered voters from Feb. 29 to March 6. But Trump’s lead expanded to 24 points in places that vote later.
In a hypothetical head-to-head against Cruz, Trump led by 1 point in places that had voted by today, but by 8 points everywhere else. As our delegate tracker indicates, Cruz needed a lead over Trump by now to be on track for a majority of delegates, because the voting gets tougher for him from here.
Networks just called Ohio for Kasich, who was leading Trump 43 percent to 34 percent with nearly 300,000 votes reported (there were about 1.2 million votes cast in the 2012 GOP primary). Those were almost entirely early votes as opposed to day-of ballots.
As expected, Trump did very well in eastern Ohio, which fits squarely into Trump’s Appalachian sweet spot and bodes well for his prospects in Pennsylvania and West Virginia. But Kasich not only activated Mitt Romney’s 2012 suburban coalition (Romney won that year’s primary while taking just 19 of the state’s 88 counties), he did vastly better than Romney in northern and western Ohio, which are rural and conservative in character but may not share Appalachia’s sense of alienation.
It’s really early, with three states yet to be called on the Republican side, but so far there are bits of good news for both Trump and anti-Trump voters. On the one hand, it looks as though Trump will get closer to 40 percent of the vote across the board tonight — and about 45 percent in Florida — rather than the 35 percent or so he’s been averaging previously. That inches him meaningfully closer to a majority. On the other hand, there seem to be fairly clear signs of tactical voting among supporters of the other candidates, with Cruz as Trump’s main competition in Illinois, Missouri and North Carolina; Kasich in Ohio; and Rubio (less successfully) in Florida.
Rubio’s exit helps Cruz and Kasich way more than Trump nationally. Among the 3,372 Republican and Republican-leaning independents interviewed from March 11 to 15 by Morning Consult, the online media and polling company, there were 412 Rubio supporters. Among them, 47 percent had Cruz as their second choice, 27 percent had Kasich, 13 percent said Trump, and the rest liked someone else or didn’t know.
Want to know why Clinton won Ohio? Look no further than white voters. According to the exit polls, she won white voters 51 percent to 48 percent. She lost them in Michigan by 14 percentage points.
In the beginning of the 2016 election season, Jonathan Bernstein identified the Walker-Bush-Rubio trio as the three likely GOP nominees. I think this probably represented mainstream opinion among political scientists in and out of the blogging community. We disagreed somewhat about which of those candidates would be the strongest. And to some extent I think a lot of us thought that Trump’s popularity over the summer was due to the party’s failure to coordinate among these three conventional candidates. Cruz seemed like a sideshow, and until he appeared in the debates, a lot of us forgot Kasich was running.
A lot of what we think we know about nominations hinges on whether this argument about coordination failure was right. And it now kind of looks like it wasn’t. Walker got winnowed out during the debate stage. Bush couldn’t turn his candidacy into a significant presence in the early contests, and Rubio’s disappointing run ends with this big loss in his home state. It seems like our misfire may have been not about those candidates getting in each other’s way, but a fundamental misunderstanding about what Republican primary voters want and about the power they wield in the process.
If Trump wins in Missouri, he might be able to thank Rubio. Thanks in part to the absentee vote being reported so far, Rubio is pulling 18 percent statewide. I expect that to plummet as today’s votes are counted, but man, Cruz wants those votes that Rubio currently has.
If Clinton’s current margin in Florida holds, she’ll get somewhere around 140 pledged delegates, while Sanders will get about 75. That would be well ahead of Clinton’s target of 116 delegates, while Sanders would fall under his target of 98 delegates.
Trump’s lead over Cruz in North Carolina is shrinking as votes come in, down to just 40 percent to Cruz’s 34 percent. But in a big red flag for Trump, Raleigh’s Wake County is reporting that Cruz is ahead 38 percent to 30 percent there with almost half of today precincts reporting. I wouldn’t turn my eye away from the Tar Heel state just yet.
If you’re looking for hints that Kasich may pick up some delegates in Illinois, he is leading the early vote in Lake County. That’s a big well-to-do suburban county in the state. We’ll see if that holds.
If you’re looking for Missouri results, I suggest you follow along at the Missouri secretary of state’s website. Later on, the site will also give us the crucial breakdown by congressional district, which is how most of the delegates will be allocated.
Just as the pre-election polls were off in Michigan’s Democratic primary, it’s possible today’s exit polls could be just as far off the mark in Ohio. The exits showed Clinton leading Sanders 53 percent to 47 percent in Ohio, but right now Clinton appears on track for a much bigger victory, dooming Sanders’s comeback hopes. With most early votes reported, Clinton leads Sanders 68 percent to 31 percent. Either Sanders has a huge vote surge in his back pocket, or the exit polls here are way off.
“We will still be rich, and we will still be powerful, but we will not be special,” Rubio added after conceding, citing the legacy of Pilgrims, enslaved people and immigrants as key to an American common culture. The question now is where our common culture will head after the divisions of this election. Will the fault lines that have riven the electorate be healed by any one leader from any one party? On top of that, Rubio’s concession speech raises the question we’ve mused about occasionally here at FiveThirtyEight: whether there could even be a split in the Republican Party. People from many other nations are used to multiparty coalition governments. It’s clear we have more political wings of the electorate in America today than we have parties, but our system is hard-wired — for now — to remain firmly two-party.
Our colleagues at ABC News have also called North Carolina for Clinton.
BEREA, Ohio — The polls just closed in Ohio and it’s looking like it’s going to be a squeaker. As Dave and Ella pointed out earlier, Kasich’s strongholds lie in areas filled with white suburban voters along the I-71 corridor, who tend to be well educated and well off. That was certainly the case when I went to the heavily Republican city of Westlake on the west side of Cleveland. When I stopped by the city’s rec center about an hour before the polls closed, the parking lot was so full that people were double parking — the precinct director told me that it had been busy like that all day. Voters there, like Ron Karpuszka, 49, seemed to be mostly casting their ballots for Kasich.
“He’s the least of all evils,” Karpuszka said. “I’m not happy with any of the choices.”
In my ongoing quest to see if blue-collar Democrats would be tempted by the fruits of another party in the guise of Trump, I headed to Our Lady of Angels in the West Park neighborhood of Cleveland, where many of the city’s police officers and firefighters make their homes. Trump has been hitting the law-and-order notes pretty hard this election.
Eileen, a 58-year-old resident of West Park who didn’t want to give me her last name, was a Trump voter that I met coming out of the parish gym. She had found the election to be an incredibly “exciting” one.
“I paid more attention to this than any other year before,” she said, noting that that she had watched all of the debates. Eileen, who voted for Obama in the last election, said that it was Trump’s “enthusiasm and realizing this country needs such a change,” that was the ultimate deciding factor.
Dave Malloy, a 27-year-old engineer, voted for Bernie Sanders, but said that his choice had been between the Vermont senator and Trump. He was looking for a candidate, he said, who “doesn’t seem like everyone who’s been in power since I’ve been alive.” I asked him what he thought about the violence at Trump rallies of late. “I think that’s part of the game,” Malloy said. “Could have just been some drunk jackass.”
Rubio, who lost badly to Trump in Florida tonight, just announced that he’s suspending his campaign. I published a long pre-obituary on Rubio last week, which contemplated the perils of being a high-ceiling, low-floor candidate, so I’ll refer you to that for now and perhaps add some thoughts in the coming days.
One thing I forgot to mention in that article, however, is that having a relatively even spread in your support across different Republican geographic and demographic groups can be a disadvantage for building momentum, since it becomes hard to win states if your support isn’t more concentrated. Up through Super Tuesday, Rubio had won roughly as many votes as Cruz. But he won only one state, Minnesota, while Cruz won several. This was hardly Rubio’s only problem, but it’s still something for future candidates to consider.
In his concession speech suspending his campaign — saying it was “not in God’s plan” — Rubio hit hard on both the idea of a conservative movement with values and a “new political establishment,” but he also began by outlining the issues with jobs and the economy. He said that his campaign was “on the right side” but “not the winning side” of this election. Real household income is lower than it was 10 years ago, and no matter who the presidential nominees are, the economy consistently ranks among the top issues.
Rubio’s speech, which is edging ever closer to a farewell, is an interesting contribution to the growing “Republicans respond to and repudiate Trump” genre, but he draws a clear distinction between the party establishment and the conservative movement, blaming the former for the party’s problems. I wonder if he’s been too conciliatory for the angry base and too abrasive for elites.
Because Illinois awards most of its Republican delegates by congressional district, the geographic splits from the exit poll are worth looking at. Cruz and Trump are tied at 33 in the Cook County (Chicago) suburbs, with Kasich at 24 percent. Trump is leading with 42 percent in the exurban collar counties. Cruz and Trump are essentially tied in central and southern Illinois, while Cruz leads Trump in Northwest Illinois.
The exit poll didn’t report results from the city of Chicago because of a small sample size, but there are plenty of delegates at stake there too.
The North Carolina Republican margin is down to 5 percentage points. Cruz is expanding his margin in Wake County as election day votes are counted. The fact that Cruz is down only 5 percentage points with many of the votes counted being early votes could be a good sign for him.
It could be a long night of vote counting in Missouri and Illinois. Based on the gender split in preliminary exit polls, Trump narrowly leads Cruz 40-36 in Illinois, with about 20 percent of the vote for Kasich. In Missouri, Cruz has a very narrow 42-40 lead over Trump from the exit polls.
The Democratic side is just as close, although with good early news for Sanders. He leads Clinton by 4 points in the Missouri exit poll, and by 3 points in Illinois.
We have to watch the election day vote in Ohio, but Clinton is literally winning every county but one in the state. In the early vote Clinton leads by 37 percentage points, and more than 150,000 votes have been counted.
NBC News has declared North Carolina for Clinton. She leads by 19 percentage points with 27 percent of the vote counted.
Nate’s point on delegates — regarding whether there is any reason for Rubio to continue — is an important one. The other one is that the numbers I’m seeing suggest even if Rubio and Cruz had been a single candidate (assuming all the same votes), that candidate would not have beaten Trump in Florida. Is it worth it for Rubio to drop out just to help Ted Cruz win second place? Like so many aspects of this race so far, that’s a question without a good answer, and that’s part of why the race looks like it does right now.
Just watching this Florida margin on the Republican side. Trump is up by 18 percentage points with most of the votes already counted. Even the most ardent Rubio haters have to feel just a little bad for him.
To echo Harry, I’m bullish on Kasich’s chances of holding onto his home state based on what we’re seeing so far. Trump is faring well in the blue-collar, coal-heavy southeastern counties of Ohio, but Kasich is holding his own in white-collar suburban counties. More importantly, the governor is over-performing Mitt Romney’s 2012 vote share in Ohio’s northwestern counties, which are more rural but less susceptible to the Appalachian anger that has fueled Trump’s margins in coal country.
So, those 1,700-plus anti-Trump ads that ran in Florida clearly did … not very much, given his win. Earlier today I posted a piece about anti-Trump TV ads for Ides of March primaries, and Monday on the podcast I spoke with Nate, Harry, Clare and Jody about race and the violence at Trump rallies. Well, let’s mash up political ads, race and history. What you get is this 1964 ad by the Lyndon B. Johnson campaign vs. Barry Goldwater.
During this four-minute (!) ad, a man (presumably an actor) recounts being a lifelong Republican who sees no choice but to vote Democratic in the face of extremism. Key line from the ad: “When the head of the Ku Klux Klan, when all these weird groups come out for the candidate of my party, either they’re not Republicans, or I’m not.” And this no-holds-barred ad from the same campaign very directly takes on the issue of the KKK endorsing Goldwater, echoing the recent fight over whether Trump waffled on repudiating the Klan and David Duke.
Not really. Rubio would only help Trump by continuing. One wrinkle, though, is that the rules vary from state to state about what happens to a candidate’s delegates when he or she drops out of the race. So there might be some legal nuance in exactly how Rubio stops campaigning.
Is there any reason for Rubio to continue?
So far, there is zero evidence that Sanders has carried any momentum from his Michigan victory into tonight’s contests. Clinton’s easy victories in Florida and (though it has not yet been called) North Carolina, as well as her strong showing so far in Ohio, suggest that she will carry at least three of the five states voting tonight. In terms of delegate math, the size of Clinton’s victories in Florida and North Carolina may be the most devastating news for Sanders fans.
We just spent 20 minutes in the newsroom trying to figure out why the AP called the North Carolina primary for Trump when only 1 percent of precincts had reported. This very human tweet answered all our questions:
There’s a lot that’s asymmetrical about the way this race is shaping up for the two parties, even though both certainly have an establishment/insurgent dynamic going on. But for the Democrats, we’re still seeing a lot of the kind of analysis that’s dominated the last 40 years (see my earlier comment about Muskie and the rough period of adjustment to the informal post-reform system). That race is being seen in terms of state victories, upsets, and expectations, not solely the count of delegates. In other words, many analysts are interpreting Democratic election results to develop a narrative about what kind of candidate Sanders is and what his candidacy means for Clinton (including, but not limited to, the possibility that he might win the nomination).
For Republicans, it’s raw delegate math. The symbolism of Rubio’s loss in Florida has been on the radar for weeks. The delegate allocation rules for each state matter. The main anti-Trump strategy embraced by party elites seems to be a contested convention.
One possible implication of this is that we’ll see a reversal of how the parties work. The recent scholarship that describes asymmetry argues that Republicans are ideological while Democrats are a patchwork coalition of interests. As Republicans scramble to mobilize an anti-Trump coalition and Democrats argue about the ideological direction of their party, these descriptions may switch.
I’d remembered Florida as being a Clinton-fecta state: one won by Bill Clinton in the 1992 primaries, the 1992 and 1996 general election, and Hillary Clinton in both 2008 and 2016. But my memory is wrong: Bill Clinton lost Florida to George H.W. Bush in 1992, despite an overwhelming Electoral College victory nationally.
ABC News has called Florida for Trump and Clinton.
For Trump, the victory must taste very sweet. He went into Florida and absolutely crushed Rubio. In doing so, he took all 99 delegates. Rubio, on the other hand, may leave the race after this embarrassing performance.
Clinton has a very large advantage in the actual vote being counted, so this shouldn’t be a surprise. The state is quite diverse and and its population on the older side, which is perfect for her. The fact that 214 delegates were up for grabs means that Clinton will extend her delegate advantage when all the votes from all the states are counted tonight.
The early results look good for Kasich in Ohio. He’s winning the areas that Romney won in 2008, and he seems to be doing better in the western part of the state than Romney did.
Missouri polls close momentarily; here’s how the state voted in 2012:
With the NCAA Tournament starting this week (technically, it’s underway right now) and the rollout of our March Madness predictions, I thought the time was right to compare the states voting today to first-round matchups in the tournament. For each party and state, I looked for the matchup where the favorite’s win probability was most similar to that of the leading candidate in our polls-only forecast.
Polls in Illinois also close at 8 p.m. Here are the GOP results from four years ago:
Remember, the North Carolina Republican primary is proportional with no threshold. Winning with 41 percent doesn’t net a candidate any more delegates than losing with 41 percent.
Sure enough, the very first returns from Ohio are from Mahoning County (near where Trump held a rally last night), and Trump leads Kasich there 47 percent to 34 percent. That’s not yet a danger sign for Kasich, though. He’s counting on white-collar areas of the state to carry him to victory, and the Mahoning Valley should suit Trump quite well.
With Rubio’s chances in Florida looking … nonexistent, here’s an image you should expect to see cited mockingly many times in the coming days:
Presuming that Florida is called for Trump at 8 p.m., most of the focus will be on Rubio’s demise. But let’s be clear: This is also a really impressive performance for Trump. He’s currently getting 46 percent of the vote in a populous, diverse state. Even if more Cruz and Kasich supporters had gone to Rubio, that would be a tough total to beat.
It’s looking like a mixed night for Ayn Rand fans from the state of Florida who are looking to keep their title hopes alive. Rubio, who fell hard for Rand’s books early in his political career, is likely to get walloped in his home state. But Florida Gulf Coast University, where every major in the Koch-funded-economics department gets a copy of “Atlas Shrugged,” is routing Fairleigh Dickinson in the men’s March Madness play-in game.
On top of a romp in Florida, Trump looks like the early favorite to carry North Carolina, but Cruz is mounting a strong showing, keeping in close among early voters in Greensboro and Charlotte and winning the early vote in some western counties, including Asheville and Hendersonville. That’s the first hard evidence that tonight may not be an across-the-board Trump blowout.
Here’s how Florida’s Democrats voted in 2008:
In Ohio, it’s possible that Rubio’s suggestion that his supporters vote for Kasich could make a difference. Rubio has just 2 percent of the vote there, according to early exit polls, while Kasich has a narrow lead over Trump.
We’re about 20 minutes from the last polls closing in Florida. Here’s how the 2012 GOP primary in the Sunshine State shook out:
Extrapolating from the gender splits in the exit polls, they have Clinton ahead of Sanders 53-46 in Ohio and 54-42 in North Carolina. Sanders won’t get much credit for it, but that would be a pretty good result for him in North Carolina, given how much closer it would be than neighboring states Virginia and South Carolina. Still, let’s wait for some actual votes to come in.
You’re going to hear stuff like “too close” and “too early to call.” What’s the difference? Too early merely suggests that there isn’t enough information to make a call. Too close means we have enough information, and the race is too close to make a projection.
My biggest pet peeve this week has been when pundits pontificate that Ohio is the only GOP state to watch tonight. These pundits claim that a Trump win in Florida is a foregone conclusion and that Illinois, Missouri, and North Carolina are “proportional” states and thus less important. In fact, Illinois and Missouri aren’t proportional — they’re effectively winner-take-all by congressional district. And they’re hugely important: in fact, they could tell us more about the odds of a contested convention than Ohio could.
If Kasich wins Ohio, it’s a close call as to whether that’s good or bad news for Trump’s odds of getting to 1,237 delegates. On one hand, it would deny Trump 66 delegates, which is 5 percent of what’s needed to clinch the nomination. But it could also deny Cruz a one-on-one matchup against Trump, heightening Trump’s chances of winning future winner-take-all states. In other words, Ohio’s outcome tonight might not shed a lot of light on the odds of chaos in Cleveland.
However, Illinois and Missouri will tell us a lot. If Trump sweeps both states by large margins and wins in all 26 of their congressional districts, he’ll win 121 delegates — more than he’ll win in Florida. However, if Cruz can win at least 10 of their congressional districts, it will be a positive indicator for his competitiveness against Trump in future primaries and significantly heighten the odds of a contested convention.
Last week an Ohio judge ruled that 17-year-olds who will turn 18 by Election Day on Nov. 8 were allowed to vote in today’s primary, overturning a move by Ohio Secretary of State Jon Husted in December to bar 17-year-olds from voting. How much will this help Sanders? It might get him an extra half a percentage point of votes relative to Clinton.
Here’s how we estimated this:
Sanders leads Clinton among 18- to 29-year-olds by about 45 percentage points — national polling from Reuters and an Ohio poll from YouGov show roughly that margin. We don’t have much polling on 17-year-olds, so the simplest thing is to assume Sanders’s lead among them is the same as the next-youngest group of voters.
There were about 150,000 Ohio residents age 17 in 2014. Let’s assume that the age composition of the state is the same and that about half of them can vote today — since Election Day is about half a year away. That leaves about 75,000 new young voters in the state. There were nearly 9 million people of voting age in the state in 2014. So less than 1 percent of voting-age Ohioans are age 17 — 0.8 percent is closer. Since 18- to 29-year-olds are more likely to vote Democratic than Republican — about 60 percent of those who had decided in the YouGov Ohio poll said they’d vote in today’s Democratic primary — 17-year-olds are likely to make up about 1 percent of the Democratic electorate today. If they go to Sanders by a margin of about 45 points, that would translate into a gain in his margin in the state of 0.45 percentage points.
This is all a rough estimate. Many of the newly franchised may not make it to the polls; on the other hand, if Sanders does better the younger a voter is, he might win them by an even bigger margin and drive their enthusiasm. They might even swing the state — if voters 18 and older are evenly split.
CORRECTION (March 16, 7 p.m.): A previous version of this post misstated when 17-year-olds who will be 18 by Election Day in November were allowed to vote in Ohio. They have had that right since 1981, not for the first time this year.
As you contemplate Rubio’s probable shellacking in Florida, it might be worth taking a look at my article from last week on Rubio’s demise. It argues that while Rubio had a high ceiling — something that made us and other reporters optimistic about his chances at various points in the campaign — he also never built up much of a base. The perfect example of that would be if he ends his campaign losing by double digits in his home state.
In order for Sanders to keep the delegate count close after tonight, he needs to do well in Florida. There is absolutely no sign that he’s going to do so from the early votes being reported in the state. He is trailing in big Democratic counties: Broward by 51 percentage points, Miami-Dade by 61 percentage points, Orange by 37 percentage points and Palm Beach by 50 percentage points. And with so many Floridians casting early ballots this year, Clinton looks like she’ll win the vast majority of the 214 delegates up for grabs in the state tonight.
This year, I’ve been keeping track of the exit polls as they were initially released to the public by the networks right when states closed, and comparing them against our pre-election forecasts. The exit polls have had some misses — for instance, they wrongly projected Trump to win Iowa and Bernie Sanders to win Massachusetts. Overall, however, the exit polls provide useful information. Once you have exit polls in hand, it turns out that the best mix to predict a candidate’s eventual vote share has been roughly 75 percent exit polls and 25 percent the pre-election polling average.
Nonetheless, sometimes both exit polls and pre-election polls are way off. The margin of error on the blended exit poll/pre-election poll estimate in predicting a candidate’s vote share is roughly +/- 6 percentage points. So exit poll results can be nice, but returns are much nicer.
As we watch another round of primary returns come in, and read more about the potential implosion of the Republican Party, it makes sense to keep asking questions about what exactly we’re seeing. In the podcast last night, the tea party came up as a possible forerunner to the Trump phenomenon. Tea party scholar Rachel Blum suggests in her dissertation that they might be connected, but it’s too early to tell.
The data presented Monday by Dan Hopkins and Diana Mutz presents a bit of a challenge to this. They find that compared with Rubio and Cruz supporters, Trump supporters are more supportive of government, are slightly less anti-Obamacare and don’t stand out as especially conservative. This profile contrasts somewhat with the one constructed of tea party “true believers” by Christopher Parker and Matt Barreto — they find tea partiers, not surprisingly, to be concerned about the size and spending of government, and much more opposed to Obama overall.
Where they agree is in their skepticism of immigrants. There’s also some geographic overlap: Trump and tea party congressional candidates have done well in the South. However, Trump has also done well in northeastern Republican primaries, where there are few Republican legislators, much less tea party types. And, as this map shows, tea party House candidates were slightly less dominant in certain parts of the West (especially compared with the South), where Trump has also been weaker in the primaries and caucuses thus far.
The Ides of March primary happens to fall on the 83rd birthday of the Notorious RBG, otherwise known as Supreme Court Justice Ruth Bader Ginsburg. This year’s election will probably help determine who gets to appoint at least one of Ginsburg’s colleagues on the bench, and possibly several more.
In the last GOP debate, Cruz said that the next president “is going to appoint one, two, three, four Supreme Court Justices.” That high figure seems like a rather grim prognosis for the lifespans of the justices, though it could suggest voluntary retirements à la Sandra Day O’Connor. But there’s no question that the empty seat left by the death of Justice Antonin Scalia, and the ages of Justices Anthony Kennedy and Stephen Breyer (both are in their late 70s) has raised the question of how the next president will shape the court.
How big a factor is the Court’s future for voters? A Public Policy Polling survey in four states including North Carolina found that voters disagree with the plans of Senate Republicans to reject any of Obama’s nominees to the court. According to the survey:
Supermajorities of voters in all four of these states — 69/25 in Arizona, 66/24 in Missouri, 66/25 in North Carolina, and 66/26 in Iowa — say that the Senate should at least wait and see who’s put forward before deciding whether to confirm or deny that person. Even Republican voters — 56/35 in Arizona, 54/38 in North Carolina, 52/37 in Missouri, and 50/39 in Iowa — think their senators are taking far too extreme of a position by saying they won’t approve President Obama’s choice without even knowing who that choice is.
May the Notorious RBG and all her fellow justices live long and prosper. Meanwhile, we’ll see how the tussle over the next Supreme Court nominee shapes the race, particularly after the party nominees are clear.
As expected, Trump appears to be headed for a sizable Florida victory based on early votes in several counties. Trump is posting large leads all over the state, though Rubio was expected to do best among early voters. The only good news for Rubio? It doesn’t look like he’s in much danger of coming in third behind Cruz.
Earlier, Andrei posted a very interesting table from the early exit polls about whether Republican voters in each state would be satisfied with a Clinton vs. Trump matchup and whether they would seriously consider voting for a third party in such a matchup. Most of the states look the same in that table, except for one: Florida.
|SATISFIED W/ TRUMP VS. CLINTON||SERIOUSLY CONSIDER THIRD PARTY|
More Florida Republicans would be satisfied with that matchup and fewer would seriously consider a third-party candidate than in any other state. That would suggest that the pre-election polls showing Trump strong in Rubio’s home state may have been accurate.
If there’s one thing we’ve learned this primary season, it’s that blue states aren’t coming to the GOP establishment’s rescue. But red states are absolutely rescuing Hillary Clinton. So far, Bernie Sanders has won 52 percent of pledged delegates (327 to 301) in states won by President Obama in 2012, according to our handy delegate interactive. However, Clinton has clobbered Sanders in Mitt Romney-won states, taking 68 percent of their delegates (466 to 221).
Of course, most of this red/blue divide has to do with the fact that Clinton has racked up huge margins among African-Americans and Latinos, who have dominated early Southern primaries such as Georgia, Texas and Virginia. But if polls are accurate (the cliché caveat after Michigan), tonight could be Clinton’s best opportunity ahead of her home state of New York to finally turn the tide in blue states. Even if Illinois and Ohio are close, she seems likely to win a healthy delegate margin from Florida.
Unfortunately for Sanders, there are no winner-take-all states to make up a big deficit like there are on the GOP side, and he’s still well behind his delegate targets. But on the plus side for Sanders, tonight marks a turning point in the Democratic race. Up until now, 55 percent of Democratic pledged delegates awarded have come from red states. Between tonight and the end of the primaries, 76 percent will be awarded in blue states, where Sanders has performed better.
There’s the possibility for some historical symmetry here — the first post-reform primaries in 1972 were shaped by Maine Sen. Ed Muskie underperforming in New Hampshire. I suspect that we’re on the verge of another round of reform, and some of the story behind that will be Rubio’s decline in Florida.
If Kasich wins his home state of Ohio tonight, it will mark the first time in the modern primary era (since 1972) that four or more Republican candidates have won states in the same year:
|YEAR||CANDIDATES WINNING STATES|
|1976||Ford* (27), Reagan (23)|
|1980||Reagan (44), Bush (6)|
|1988||Bush (41), Dole (5), Robertson (4)|
|1996||Dole (46), Buchanan (4), Forbes (2)|
|2000||Bush (43), McCain (7)|
|2008||McCain (31), Romney (11), Huckabee (8)|
|2012||Romney (37), Santorum (11), Gingrich (2)|
|2016||Trump (15), Cruz (8), Rubio (1)|
Note that I’m going by popular vote rather than delegates, so I don’t count Ron Paul as a winner in 2012 although he won the plurality of delegates in a few states. Democratic nominations have often been much wilder. Eight Democrats won states in 1976, for example. But Republican nominations have usually been fairly orderly — not so this year.
Although the Republican primary calendar will slow down after tonight, there’s another big prize at stake a week from today: Arizona, which votes March 22 and has 58 winner-take-all delegates. Trump is an 84 percent favorite in Arizona according to our “polls-plus” forecast, and Cruz has a 13 percent shot. Note, however, that the forecast is based on two recent polls, both of which showed a fairly large undecided vote. The polls could also be shaken up if one or two candidates drop out tonight.
But Trump’s focus on immigration should presumably play well in Arizona, and the fact that the state’s electorate is older could also give him an advantage. On the other hand, Cruz has performed fairly well in the West so far, and Arizona holds a closed primary, which has sometimes hurt Trump in the past.
Polls close in Ohio in about 45 minutes; here’s how Democrats in Ohio voted in 2008.
CORRECTION (March 15, 7:21 p.m.): Due to an editing error, a previous version of this post gave the incorrect poll closing time in Ohio. Polls close at 7:30 p.m., not 7 p.m.
Greetings from Brooklyn … Ohio. I am sitting in my third coffee shop of the day on the hunt for free Wi-Fi (we’ve moved from coffee to green juice in order to avoid the deep jitters) and have been checking in with the voters of the northeast corner of this state, who seem to be of interest to us political types this year more than ever; Ohio is the center of a pitched battle between its governor, John Kasich, and Donald Trump, and even the Democratic race seems a bit closer than we might have anticipated only a couple of weeks ago.
I’m in Ohio’s 11th Congressional District, where the estimated Republican support for Trump, according to a handy little visualization our friends at the Upshot whipped up a couple of months back, is at about 30 percent. This is Cuyahoga County, which is a name you will recognize if you’re a dedicated reader of this site because, well, it’s home to a big honking chunk of this state’s voters, and it’s oft-mentioned on election nights. As the guys at the Cuyahoga County Board of Elections told me today, it’s not their Super Bowl just yet (that’ll come in November), but it’s certainly playoffs time.
But when I talked with voters filing in and out of St. Leo the Great Parish Community Center, there wasn’t one Trump voter — people seemed to be leaning toward Kasich. Old Brooklyn, the Cleveland neighborhood where St. Leo is located, is largely white and working class — the median annual income is about $38,000 — which has been thought to be Trump’s demographic. As you get farther from the city center, voting patterns lean more Republican, as this precinct vote breakdown map of Cuyahoga County from 2012 shows; Parma, which you can see on the map, is near Old Brooklyn and right at the edge of where blue votes turn to red ones.
Kathy Yatsko, 56, of Old Brooklyn, told me she was voting for Kasich because she liked his record in Ohio. She wasn’t so sure about his prospects, though, and that didn’t make her happy. “If Cruz or Trump makes it, I won’t be voting in the general,” she said. “I think they’re both nuts.”
Linda Horabik, 61, also of Old Brooklyn, told me she was switching her affiliation from Democrat to Republican — “I don’t like anyone,” she said. But “if it’s Hillary and Trump in the general, I’ll vote for Hillary — I can’t in my right mind vote for that man.”
Among Democratic voters heading into St. Leo, there seemed to be an even split (anecdotal, of course, dear empirically minded readers) between Clinton and Sanders.
Kim Streicher, 38, who had come to vote with her two small daughters in tow, said the much-talked-about Sanders persona had nothing to do with her vote for him. “I’d like to get our troops out of Afghanistan, and I’m not sold on Obamacare,” she said.
And Bob Dvorak, 73, a retired janitor, was as straightforward as could be when explaining his support for Clinton.
“Because nobody else I like,” he said, before getting in his car and driving away.
While Trump’s opponents have jockeyed to be the “anti-Trump,” his support has remained relatively steady, allowing him to rack up a healthy delegate lead in a divided field. But Trump has done somewhat worse in more heads-up races:
|STATE/TERRITORY||COMBINED VOTE SHARE OF TOP TWO CANDIDATES||WINNER|
Aside from Florida (where Rubio and Cruz remain close), today’s races seem to feature pretty clear two-way contests between Trump and Kasich in Ohio and between Trump and Cruz in Illinois, North Carolina and Missouri. Although Trump leads most of the polling, he has typically received about the same percentage on election day as the polls already had supporting him. But what about his opponents? Here’s how all of the remaining candidates have performed versus their polling: Each line (or floating point) is the range of polls for one state that came within 31 days of the state voting. Each point is a poll, and the trend is weighted by how close to the election each poll was (largest bubbles represent the final polls taken before voting).
(Not making the cut because of insufficient polling: Alaska; Hawaii; Maine; Minnesota; Puerto Rico; Washington, D.C.; and Wyoming.)
In the RealClearPolitics averages Trump leads Cruz 41-29, 36-30, and 36-29 in North Carolina, Illinois and Missouri, respectively. This is dangerous territory for Trump! In states where Cruz has been the most competitive, he has outperformed his polls significantly, while being in the high 30s to low 40s for Trump has tended toward poorer results.
Of course, any effect we can see here is based on a very small sample, but we can say that if Cruz wins some states today — potentially even all three — it would be consistent with the dynamic we’ve seen so far.
Illinois Republicans are holding a version of a “loophole primary” today: While some of the state’s delegates will be awarded based on the statewide vote for president, most delegates will be elected directly at the congressional-district level.
What that means for voters is that they’ll be able to cast two votes that will affect the selection of presidential delegates (and in the votes cast directly for delegates in each congressional district, voters can select up to three choices). Below is a sample ballot from McLean County:
If Kasich wants to hang on in his home state against Trump tonight, he’ll need to do well among the same group of urban, well-educated Republicans that just barely propelled Mitt Romney to an Ohio primary victory in 2012. Romney beat out Rick Santorum 38 percent to 37 percent that year by winning just 19 of Ohio’s 88 counties.
All of those counties were located in just three clusters linked together by Interstate 71, also known as the “Three C’s:” the Cincinnati area (extending north to Dayton), Columbus (and its bedroom Delaware County to the north), and Northeast Ohio (mainly Cleveland, but extending east to the Mahoning Valley, south to Akron and Canton, and west to Lorain and Sandusky).
It’s possible Kasich could be headed for a bigger win tonight than Romney achieved, but it’s also possible that his support is so concentrated in well-educated suburbs that his path could be even narrower than Romney’s. If there’s one place where Trump’s campaign might feel they can make inroads into the Romney map, it’s probably the economically distressed Mahoning Valley (Youngstown and Warren), where “Trump Force One” touched down for a last-minute rally last night.
There’s going to be a lot of attention paid to Florida and Ohio on the Republican side tonight. They are winner take all, and if Trump wins both he will likely make this a two-candidate race with Cruz. Still, the name of the game at the end of the evening is delegate math. Fellow live-blogger Dave Wasserman has pointed out that Illinois and Missouri are really important in determining where the delegate race is heading after tonight.
Unlike what you may have heard, Missouri and Illinois are not proportional states. They are basically winner take all, some by congressional district and some by the statewide vote. If one candidate wins one of the states by a wide margin, he is likely to come close to taking all the delegates. That’s what happened in South Carolina, when Trump won all the delegates with just 33 percent of the vote. If Trump ends up winning both states by more than 5 percentage points, despite losing Ohio, he’ll have had a strong night. Indeed, more delegates are at stake in Illinois (69) than in Ohio (66), and Illinois and Missouri , together (121), have more than Florida (99).
Given that Trump is likely to win Florida, Cruz has to do well in Missouri and Illinois to keep his delegate deficit to Trump within 200. That’s a much more manageable race than a 300-delegate deficit would be if Cruz has hopes of winning a plurality of delegates. Further, if Trump does poorly in Illinois, he may end up doing poorly in Wisconsin next month (which is also a winner-take-all state on the congressional district and statewide level). If Cruz can win in Missouri, it shows that he may be able to consolidate the anti-Trump vote when the race shifts out West. There are a lot of delegates in those western states starting with Arizona (58) and Utah (40), which vote next week.
One reason to treat early exit polls as preliminary is because … well, they change. They’ll be weighted as more waves of the exit poll and actual results come in. You really don’t have to look further than North Carolina, where MSNBC’s Steve Kornacki reports that just 29 percent of (early) Democratic exit poll respondents were black. I’d be really surprised if that ended up being the case.
Black voters made up 34 percent of Democratic primary voters in the 2008 Democratic primary exit poll. That means there would have to be a 5 percentage point drop in the percentage black voters make up of the electorate. I find that unlikely because of the nine Southern primaries to vote so far, only one (Virginia) saw such a drop. Six of the nine (Alabama, Arkansas, Georgia, Oklahoma, Mississippi and South Carolina) saw a rise. In fact, on average in all nine states, the black percentage of Democratic voters rose by 5.2 percentage points.
Also, the early voting data doesn’t support it either. According to the New York Times’s Nate Cohn, 38 percent of early voters in North Carolina were black. That means that there would need to be a significant white vote on election day to make the exit poll right. If anything, the past bias for exit polls is to overestimate the proportion minorities make up of the electorate versus the voter files that states put together.
Wyoming’s Republican county conventions, held Saturday, gave nine delegates to Cruz, one to Rubio and one to Trump (one uncommitted delegate was also elected). Wyoming’s delegate-selection process is quite different from most other states. The first step was precinct caucuses, in late February and early March, where delegates to the county conventions were chosen. (The Wyoming Tribune Eagle has a good piece describing a precinct caucus in Cheyenne.) Then, 12 of Wyoming’s 23 counties chose one national delegate each at their county conventions (that’s what happened Saturday). The other 11 counties were relegated to selecting alternates to the national convention. (The exception is Laramie County, which chooses both a national delegate and an alternate.) Every four years, the counties switch places; the counties that chose alternates this year chose delegates in 2012.
There’s one more step: the state Republican convention, April 14 to 16, where 14 more of Wyoming’s national convention delegation will be chosen. That leaves three additional delegates to get to the state’s total of 29. Those three — state party officials — are free to support whichever presidential candidate they choose.
Bernie Sanders and his supporters are hoping for repeats today of last week’s historic Michigan polling miss. Sanders trailed in the polls by an average of 21 points but beat Clinton by 1.5 percentage points, and he could need to outdo his polling numbers by margins like that — today and beyond — to close the delegate gap.
One notable difference between today’s races and Michigan’s: The pollster that missed Michigan by the most, Mitchell Research and Communications of East Lansing — which found Clinton up 37 points two days before the race and 27 points the day before — didn’t release polls about any of today’s races. The firm is experimenting with new methods, but the tests are not for public consumption. “We are not going to sit back and wring our hands, we are going to try to figure out how to do a better job!” CEO Steve Mitchell said in an email.
The problem in Michigan, Mitchell said, was chiefly that the firm underestimated turnout among younger voters. And one big reason for that is that the firm didn’t dial cellphones — in part based on voter turnout estimates from Mark Grebner, founder of Practical Political Consulting, who provided Mitchell’s lists of voters to call. Grebner had found that people under 50 made up no more than 27 percent of voters in August Democratic primaries — for local races — in 2010, 2012 and 2014. And Mitchell figured from those sorts of numbers that it was safe to miss people who have only cellphones. “I’m afraid I just got kind of complacent.”
Grebner told me that was the wrong decision. “You had to have some strategy to deal with cellphone people,” Grebner said. “I assumed Steve had some strategy for dealing with them.” He cited the streetlight effect — based on the story of a drunk searching for a lost item under a street lamppost, where it’s easiest to look, not where the item is likeliest to be — and said, “Landlines are the lamppost.” Grebner sometimes does landline-only polls for local clients. “We do them a crappy little poll, and our crappy little polls are mostly right in most cases.”
But not in this case. “This one is a disaster from a polling perspective,” said Grebner, who previously has helped uncover suspected political polling fraud. “It’s bad for everyone’s reputation; it’s bad for my reputation.” He added, “I’m sorry if I’m partly responsible for Steve’s disaster. Everyone else was off, too but …” I said, “He was off more.”
Politico’s Shane Goldmacher reports that influential conservatives are gathering to discuss the possibility of running a conservative on a third-party ticket, should Trump be the Republican nominee.
So far, there’s more smoke than fire — a person quoted in the Politico article described the planning process as being in the “embryonic” stages. But I’m less skeptical about these reports than I usually am about third-party bids. The reason, as I outlined in a piece last month, is that Republican conservatives really do have a lot of reasons to oppose Trump and perhaps even have incentives to see him lose the general election. As I wrote:
If you’re one of these ideological conservatives, it may even be in your best interest for Trump to lose in November. If Trump loses, especially by a wide margin, his brand of politics will probably be discredited, or his nomination might look like a strange, one-off “black swan” that you’ll be better equipped to prevent the next time around. You’ll have an opportunity to get your party back in 2020, and your nominee might stand a pretty decent chance against Clinton, who could be elected despite being quite unpopular because Trump is even less popular and who would be aiming for the Democratic Party’s fourth straight term in office.
But if Trump wins in November, you might as well relocate the Republican National Committee’s headquarters to Trump Tower. The realignment of the Republican Party will be underway, and you’ll have been left out of it.
In other words, I don’t necessarily see this as a bluff. For certain types of movement conservatives and certain types of “establishment” Republicans — not all of them or necessarily most of them, but some of them — Trump becoming president is probably as bad a long-term outcome as Clinton becoming president. Thus, there wouldn’t be the usual fear about throwing the election to the Democrats, even though (for instance) a Mitt Romney-Ben Sasse ticket would make that more likely. Meanwhile, the conservative ticket could serve as a sort of Republican Party-in-exile, with vulnerable Republican senators up for re-election in swing states declaring their allegiance to that candidate rather than having to support Clinton or Trump.
At the same time, having credible incentives to formulate a third-party ticket and getting around to doing so are two very different things. If Trump has a really good night, we’ll probably see a lot more talk about this in the coming days. If he loses a few states, Republicans may focus more on still trying to defeat him in the nomination process.
Although I sit at a microphone for the taping of each week’s elections podcast, I feel much closer to a listener. It’s my chance to check in with some pretty sharp and fun folks and get some clarity about the confusing state of the presidential race. Here are a few key moments from Monday’s taping that are still rattling around in my brain:
Farai Chideya on violence at Trump rallies and whether it’s like anything we’ve seen before:
When I look at this violence, I see it as part of a continuum of the struggle of different groups of Americans — not just black people and not just immigrants, but in the context of the Trump rally, those two stand out — for full political participation.
Nate Silver on how Trump encourages violence and how everyone else encourages him:
There’s just this gigantic cloud of dust with everything Trump does, and sometimes that can lead to a lack of clarity, which is: This is wrong. It should be denounced. It should be denounced by people in the media. It should be denounced by other Republicans. I think it’s pathetic that Marco Rubio and Reince Preibus and Ted Cruz and John Kasich aren’t clearer about that.
Some of Harry Enten’s benchmarks to look for as the Democratic results come in. Remember, it’s not just about wins and losses, but delegate math:
For Bernie Sanders to have a realistic shot at the nomination, he should really hold Clinton to 20-point margins or less in North Carolina and Florida.
Take a listen to the full podcast below, and subscribe on iTunes so that you don’t miss any future episodes.
Listen to the latest episode of the FiveThirtyEight elections podcast.
This is preliminary exit poll data, and it can/will change, but you can see the #NeverTrump contingent isn’t limited to Republican elites in D.C. — a solid chunk of voters would consider voting for a third party candidate if Trump wins the Republican nomination.
|SATISFIED W/ TRUMP VS. CLINTON||SERIOUSLY CONSIDER THIRD PARTY|
The five states voting today don’t break down into neat economic categories the way the Super Tuesday states did. But given how much the election has focused on the economy, it’s still helpful to think about where today’s states fit into the bigger picture, and how they compare to states that have voted so far.
Like Nevada (where Trump won handily and Clinton narrowly), Florida rode the housing rollercoaster, booming during the mid-2000s and suffering when the bubble burst. Unlike Nevada, Florida has since seen a strong recovery; its unemployment rate is down to 5 percent from a peak of more than 11 percent in 2009.
In economics as in culture, Illinois is really two states: Chicago and its suburbs, and everywhere else. Chicago, like most big U.S. cities, is marked by inequality: high median income, but also high poverty. The rest of Illinois looks more like other Midwestern states, with significant manufacturing and agricultural sectors. Overall, the state experienced a deeper recession than the U.S. as a whole and has struggled with a weaker-than-average recovery.
No one state is a perfect proxy for the U.S. economy, but Missouri comes pretty close. Its poverty rate is a bit higher than the country’s; its unemployment rate is lower. It has two relatively big cities but also a substantial rural population. In terms of income and education, it’s right about average. Missouri wasn’t hit as hard by the recession as many states, but its recovery has been relatively weak.
North Carolina is a bit richer and more educated than South Carolina, where Trump and Clinton both dominated. But both Carolinas have experienced relatively strong economic recoveries, particularly in their manufacturing sectors. And North Carolina’s economy, like South Carolina’s, varies substantially from one part of the state to another; the wealthier, more educated counties that surround Raleigh and Charlotte are thriving, while Greensboro and Salem are struggling.
It’s tempting to lump in Ohio with its neighbor and fellow industrial powerhouse Michigan, where Trump and Sanders rode economic anxiety to victory last week. But like the Carolinas, Ohio shows significant regional variation. Western and central Ohio are faring well, with county unemployment rates often at or below 4 percent. Cleveland, where Republicans will hold their possibly contentious convention in July, is also faring well. But eastern Ohio is struggling, with county unemployment rates still above 7 percent.
Marco Rubio desperately needs a win in his home state tonight, but the polls don’t look encouraging. While it’s not much consolation, Florida is easily the senator’s strongest state in the Facebook primary:
Rubio holds 14 percent of Florida’s Facebook likes (among the remaining candidates in both parties, plus Jeb Bush and Ben Carson), which is double his national share. He has the most Republican likes in populous Miami-Dade County, as well as in Leon County, home to Tallahassee, the state capital.
The only other county nationwide where Rubio leads is Arlington, Virginia, which he won with 50 percent of the vote on Super Tuesday. Unfortunately for Rubio, Donald Trump leads in Florida’s other 65 counties and easily has more likes statewide. See more states!
Sanders got more support among black voters in Michigan than he had in other states with large black populations that have already voted. Most of those are in the South. He’ll need to continue to attract more black support in order to catch Clinton, and some of his supporters are hoping Michigan signals that the electoral map’s shift away from the South will help Sanders. Polling data from online media and polling firm Morning Consult suggests that it will help only marginally, unless he can make further gains — and might not help at all in the Midwest. Aggregating 10 national polls since January, Morning Consult found that Clinton led Sanders by 53 percentage points in the Midwest, 50 points in the South, 47 points in the West and 38 points in the Northeast.
|REGION||CLINTON||SANDERS||SOMEONE ELSE||DON’T KNOW|
For months now, we’ve seen a packed schedule of debates and plenty of television advertising promoting or denouncing various candidates. But few of these ads have been aimed squarely at the Republican front-runner, Donald Trump. Only now, as he has come closer to clinching the Republican nomination, have the gloves come off.
The Political TV Ad Archive, a project run by the nonprofit Internet Archive, has been recording broadcast television in selected U.S. markets, indexing political ads and counting how many times they run. (Frequency can indicate the intensity and stakes of the race and the willingness with which ad buyers invest.) We looked at data from the Archive over the three weeks leading up to the March 15 primaries in Florida, Ohio and North Carolina and saw clear evidence of a steep rise in attack ads against Trump. (The Archive doesn’t track media markets in Illinois or Missouri.)
All told, we tallied 3,365 times that anti-Trump ads have run in these states over the past three weeks (through noon on March 14). But we found different patterns of anti-Trump advertising depending on the state of the race in each state. Here’s what Florida’s ads look like:
Florida had significantly more anti-Trump advertising than North Carolina or Ohio, most likely because Florida is a delegate-rich winner-take-all state in which Trump is leading. (It may also be Marco Rubio’s last stand.) Eighty-eight percent of all Trump attack ads in these three states ran in Florida, 9 percent in North Carolina and just 3 percent in Ohio, where Gov. John Kasich leads in most polls.
Read more of this analysis in my article from earlier today.
It’s already been a good day for Trump. Earlier today, he won all nine delegates in the Northern Marianas Caucus, where he had the support of territorial governor Ralph Torres. Turnout was tiny — 471 people — but Trump won 73 percent of the vote, marking the first time he’s achieved a majority in any state or territory.
He’s been unable to do that so far. In the contests where Clinton has beaten her targets, she has done so, on average, by a larger number of delegates than Sanders has in contests where he’s beaten his.
Sanders has met his targets in three states and exceeded his targets in five. But in those five, he picked up only 11 delegates more than his targets called for, an average of about two delegates per contest.
Clinton, on the other hand, has also matched her targets in three states but has exceeded her targets in 15. And in those states, she’s netted 101 delegates above her targets, an average of about seven delegates per race.
These are each candidate’s targets for today:
John Kasich hasn’t won a primary or caucus yet, but he has his best chance tonight in his home state, Ohio. The governor does not have tremendous Facebook support nationwide — he captures just 1 percent of likes among the remaining candidates from both parties, plus Jeb Bush and Ben Carson — but he’s a player in Ohio, capturing 12 percent of likes.
Kasich’s support is especially strong in the Columbus metro area; he leads all remaining Republicans in Delaware County, with 23 percent of likes, and and in Franklin County, with 17 percent of likes. Donald Trump takes Ohio’s other 86 counties and has particularly strong support in the state’s Appalachian counties toward the east and south. It should be a close primary, but we can expect Kasich to perform well close to the state capital. See more states!
If Donald Trump wins Florida, and its 99 winner-take-all delegates, he’ll stay on track for getting a majority of delegates by the end of June, based on FiveThirtyEight’s delegate targets. But his win would knock the other three Republican candidates even further off track — putting them a combined 23 percentage points below their targets. That’s because although only one candidate can win any of Florida’s delegates, each candidate needs those delegates to stay on track.
The same goes, to some extent, for Ohio, where 66 delegates will go to the candidate who gets the most votes (Marco Rubio’s target for Ohio is 0, so his cumulative total, relative to his target, wouldn’t be hurt by a loss there). FiveThirtyEight’s forecasts show that to be a closer race, with Ohio Gov. John Kasich favored. A Trump loss in Ohio would put him below his cumulative target, even if he were to win Florida. There are three other GOP contests today:
- Missouri will give five delegates in each of its eight congressional districts to the candidate who gets the most votes in the district and 12 delegates to the candidate who gets the most votes statewide. If any candidate gets a majority of statewide votes, he gets all 52 of the state’s delegates.
- Illinois has a “loophole primary”: In each of its 18 congressional districts, voters will cast their ballot for up to three delegates; delegates may be bound to a presidential candidate or uncommitted (the ballot will indicate which candidate, if any, a delegate prefers). Voters will also vote for a presidential candidate directly; that vote will be used to determine the statewide winner, who will get 15 delegates.
- North Carolina will award its 72 delegates proportionally among all the candidates based on statewide results.
These are each Republican candidate’s targets for today, based on polling, demographics and social media data:
Harry mentioned that in addition to our polls-based forecasts, we now have demographics-based forecasts for the Democratic race. If you want to use those projections as benchmarks to see how Clinton and Sanders over/under-perform tonight, here’s a handy table:
|FORECAST BASED ON DEMOGRAPHICS AND RESULTS IN PAST PRIMARIES||“POLLS-ONLY” FORECAST|
|DATE||STATE||CLINTON||SANDERS||SANDERS WIN PROB.||CLINTON||SANDERS||SANDERS WIN PROB.|
I’ve had a lot of tweets come my way questioning our models’ certainty in the Democratic primaries in Illinois and Ohio today. Specifically, both our polls-only and polls-plus models give Clinton at least an 88 percent chance of winning Illinois and Ohio, and some people think that she is much less of a favorite. They point to how the same models had Clinton at a greater than 99 percent chance of winning the primary in Michigan, which Sanders won. I happen to think Clinton is a favorite in Illinois and Ohio, but not by anywhere near the margin the models calculate.
The models are, as their names suggest, heavily influenced by the polls. Even the polls-plus model’s major component, especially close to an election, is the polls. Both models take into account past poll accuracy and produce a candidate’s percentage chance of winning. Historically, candidates who lead in every poll and by at least 5 percentage points in every poll (as Clinton is in Ohio) win a lot more often than they lose. That’s even more true when it’s a one-on-one matchup like we have on the Democratic side.
Normally, I would put my faith in the models, but there are reasons to doubt them. In Illinois, the polls have been all over the place. A Chicago Tribune survey put Clinton’s advantage at 42 percentage points, while YouGov had Sanders ahead by 2 percentage points. Our models try to reconcile those differences, though I’m not sure they have done so adequately. In Ohio, the demographics suggest it should be about a 3 percentage point Clinton win rather than the 11 percentage point average lead Clinton has had in the polls. The demographics of Michigan suggested a far tighter race than the polls did before Sanders won.
Indeed, my own gut lies more with what Nate’s demographic models calculate as the probability of a Clinton win in Illinois (66 percent) and in Ohio (58 percent) than the polling-based models. That’s not to say Sanders will win both or even either of these states. It’s just that in a year of craziness and with competitive open primaries on each side of the aisle, I’m not taking polls to the bank when the demographics suggest caution.
Today could end up being more super than the original Super Tuesday! Sure, fewer states are voting, but the results tonight will go a long way toward determining whether Donald Trump can reach the 1,237 delegates necessary to clinch the Republican nomination. And although Hillary Clinton’s place atop the Democratic contest is more secure, a bad night tonight could presage a bad month, as the next few weeks on the Democratic primary calendar include mostly pro-Bernie Sanders states.
Here’s how the night should unfold: We’ll start getting exit poll data from all five states voting today — Florida, Illinois, Missouri, North Carolina and Ohio — at 5 p.m. EDT. Then we’ll spend 2.5 hours both discussing what that data looks like and telling you to mostly ignore it.
At 7:30 p.m. EDT, polls close in Ohio and North Carolina, followed by Florida, Illinois and Missouri 30 minutes later (some polls in Florida close at 7 p.m., but 8 p.m. is the earliest networks will call the state).
For the next several hours, we’ll be diving into all the results, trying to figure out what it all means, and — hopefully — talking to you. We want to hear what you think. So leave a comment. And enjoy!