How The States Voted On Semi-Super Saturday

That’s a wrap for us, folks. We’ll have more analysis of tonight’s results in the days ahead, but here’s how I’d score the evening relative to expectations on the GOP side. I mean that in a more precise way than when the term “expectations” is usually invoked. Specifically, I mean that if you’d drawn up a list of plausible outcomes this morning and ranked them from best (10) to worst (0) for each candidate, how would candidates fare by that measure? Ted Cruz — 9 out of 10. Huge win in Kansas, unexpected win in Maine, and unexpectedly close to Trump in Louisiana, with results suggesting he might even have won if not for early votes. Cruz’s loss in Kentucky was also narrow — he came much closer to Trump, for example, than he did in Tennessee on Super Tuesday. Not quite a “best case scenario” but not more than one step removed from it. Donald Trump — 2 out of 10. You could equivocate by saying Trump performs poorly in caucuses, and there aren’t all that many of them left, but the huge split in the election day versus early vote in Louisiana suggests that he’s encountering serious problems, perhaps the most serious since voting started on Feb. 1. It also appears as though Cruz will pick up more delegates than Trump did from the night. Marco Rubio — 1 out of 10. Just 17 percent of the vote in Kansas, 17 percent in Kentucky, 11 percent in Louisiana (with a huge drop-off from the early vote to election day votes) and 9 percent in Maine. Not. Good. He’ll presumably hang on until Florida on March 15 and a win there would still be a big deal, but he needs to gain votes to do that and right now he’s losing them instead. John Kasich — 4 out of 10. We thought he’d do a little better in Maine. His results in Kentucky in counties that border Ohio were solid but not stellar. He’s basically running as a one-state spoiler candidate at this point, although the fallout from some of the other developments (like Rubio’s bad evening) could create opportunities for him in the long run. Still, all of those opportunities would seem to involve a contested convention and it’s not clear why he’d emerge as the choice from such an event.

https://twitter.com/NateSilver538/status/706336495892877312

We’re about to shutter this live blog, so let’s take a look at how Semi-Super Saturday played out on the Democratic side. Sanders won Kansas and Nebraska. That’s the good news for him. The bad news is he’s even further from the nomination than he was before the day started: He lost Louisiana, and, in doing so, fell even further behind in the delegate hunt. Let’s take a look at the math. Sanders won 23 delegates in Kansas to Clinton’s 10. He won — preliminarily — 14 delegates in Nebraska to Clinton’s 11. That’s a margin of 16 delegates. In losing Louisiana, however, Sanders only claimed 12 delegates to Clinton’s 39. Combine the three states, and Clinton gained 11 delegates on Sanders. Now you might be saying, but didn’t we expect Sanders to do poorly in Louisiana? Yes, that’s true. But according to our delegate targets, which takes that into account, Sanders is now 3 delegates further behind the pace he needs to win a majority of pledged delegate than he was at the beginning of the day. Considering he was already running 82 delegates behind his delegate goals, he needs to be exceeding his delegate targets. Overall, it was actually a bad day for Sanders by the math, even with his two wins.

https://twitter.com/DLeonhardt/status/706331653577252864

https://twitter.com/AshleyRParker/status/706327280327356417

Our colleagues at ABC News, which did NOT call the Louisiana Republican primary earlier in the night, now projects Trump will win the state.

https://twitter.com/AP/status/706325190184202240

Where are there still returns outstanding in Louisiana?

• About half of Caddo Parish, which has been good for Cruz so far, has yet to report.
• Lots of votes left in East Baton Rouge Parish, which has also been good for Cruz.
• About 70 precincts in Jefferson Parish have yet to report; Jefferson has been good for Trump.
• Lots of precincts in Orleans Parish haven’t reported, but there aren’t a lot of Republican votes there. Trump is slightly ahead of Cruz in the ones reported so far, but Rubio and Kasich also do comparatively well.
• Significant amounts of St. Tammany Parish, which has been good for Trump, have yet to report.

Overall, our guess is that Trump will just barely hold on, but these were not the sorts of margins some networks were expecting when they called Louisiana for Trump just after 9 p.m..

By the way, kudos to the state of Louisiana for reporting its results very quickly and efficiently tonight. The states you might be inclined to stereotype as being being slow to report returns are often among the best instead.

You keep hearing from people like me that Trump has a floor of about one-third of the vote. For instance, he averaged 35 percent in the Super Tuesday contests. Today, it’s the same story. His average vote percentage in the four states that voted today is 33 percent. The big difference from other days is that Cruz was able to coalesce a lot of the anti-Trump bloc, which led to at least two victories.

As more results come in from Kentucky, an update on Ohio border counties: Cruz has 33 percent of the vote from counties that border Ohio, to 30 percent for Trump and 23 percent for Kasich (and 14 percent for Marco Rubio). Ohio looks as though it could be a three-way race on March 15.

To take one example of the huge split between election day returns and early votes in Louisiana: in Acadia Parish, Trump got 47 percent of the early vote, with 26 percent for Cruz and 19 percent for Rubio. In election day returns, however, it’s Trump 44, Cruz 41, Rubio 9. Trump often performs worse with late-deciding voters, but we haven’t seen anything this dramatic yet.

Trump just won Jefferson County, Kentucky, 30 percent to Cruz’s 29 percent. That’s bad news for Cruz’s Kentucky hopes. While Cruz is down just 4 percentage points, Trump’s margin in the overall vote remains fairly steady at around 5,000 votes. There’d have to be a late surge from somewhere for Cruz to win; otherwise, it’ll be close but no cigar.

In less exciting news from Louisiana, Clinton currently leaders Sanders by 47 percentage points. That’s consistent with her 45-point lead in our pre-election polling average.

https://twitter.com/Redistrict/status/706318939949641728

The margin separating Trump and Cruz in Kentucky is also shrinking. Trump is up just 4.6 percentage points, in part because Cruz won Fayette County (Lexington) 29 percent to 25 percent. Unfortunately for Cruz, both Kasich and Rubio got over 20 percent of the vote there, which may have held down Cruz’s margin. We’re still waiting on Jefferson County (Louisville).

Trump’s lead in Louisiana keeps narrowing every time I refresh the secretary of state’s webpage. It’s now down to 5.9 percentage points. Maybe Trump will hold on, but the networks should be rescinding their calls since there’s no way they’d have called it for Trump based on the information we have available now. There’s the risk of serious embarrassment for news outlets who are not picking up on this.

The very earliest returns in Louisiana, which were substantially composed of votes cast before election day, showed Trump at 48 percent, Cruz at 23 percent, and Rubio at 20 percent. Now? It’s Trump 43, Cruz 34 and Rubio 14, according to the Louisiana Secretary of State. The differences suggest a major gap between early votes and election-day returns, with Cruz surging in the past couple of days at the expense of both Trump and Rubio.

If you’re looking for good news as a Sanders’s fan, look no further than Kansas. Sanders won 23 delegates to Clinton’s 10 in the state. Not only did Sanders win, but he outperformed his FiveThirtyEight target of 19 delegates. Of course, he’ll need a lot more wins like that to have a realistic shot at the nomination.

https://twitter.com/Redistrict/status/706312651991744512

Nate, in answer to your earlier question about why Sanders won Kansas, one explanation is likely to be a very heavy turnout in Douglas County, home of the University of Kansas, and traditionally one of the most liberal parts of the state. (In the 2008 general election, 64 percent of the county voted for Obama, while virtually every other part of the state supported McCain.) The Lawrence Journal-World reported a Democratic caucus turnout so heavy that voters had to be counted on a middle-school football field. Sanders won 81 percent of the caucus vote in Lawrence, and much of his support undoubtedly came from students and faculty at KU.

https://twitter.com/alexis_levinson/status/706310466772078593

There does seem to be some movement away from Rubio in the past few days in Louisiana. While Rubio was getting near 20 percent of voters who cast a ballot early, he is now down to just 15.8 percent including election day voters.

As Rubio looks to have had another very bad night, it’s worth thinking about something we’ve brought up before: Is it possible that Rubio is actually too conservative for his own good? Obviously, the Republican Party is a conservative party. (Although Trump’s success suggests that the conservatism of GOP voters is complicated.) But Rubio is not quite in the moderate conservative “lane” that often proves to be fruitful in the GOP nomination process. Instead, Rubio is quite conservative indeed — at least based on his policy positions if not necessarily his temperament. So Rubio has to compete with Cruz for very conservative voters while moderate conservatives have had other options to consider. Kasich, in particular, is a little more conservative than he lets on but has run explicitly toward the center. In some ways, a race with Trump, Cruz and Kasich would create clearer differentiation between the candidates than one between Trump, Cruz and Rubio, which has left Rubio as a lot of voters’ second choice.

Just a pause-for-breath post as results and calls come in left and right. Republicans:

• Kansas — Called for Cruz, who won by 25 percentage points.
• Maine — Called for Cruz, who won by 13 points.
• Louisiana — Called for Trump by AP; he’s up big based on early votes.
• Kentucky — Still counting. Trump up by 9 points, although some of Cruz’s potentially better areas are outstanding.

Democrats:

• Kansas — Called for Sanders, no data on margin of victory yet.
• Nebraska — Called for Sanders, who’s up 10 percentage points with 75 percent reporting.
• Louisiana — Called for Clinton, who will win huge.

Is Trump doing well enough tonight to keep on track to win the nomination? Well, he won 9 delegates in Kansas and 9 delegates in Maine. In Maine, he came in on target, according to our projections. Trump fell far short of his 16 delegate target in Kansas, however. Cruz, on the other hand, finished 4 delegates ahead of his target in Kansas and 3 delegates ahead in Maine. What is clear is that Kansas and Maine were disasters for Rubio; he ended up 7 delegates short in Kansas and 10 delegates below his target in Maine. We’ll have to wait to see how Louisiana and Kentucky get split up.

And, as expected, AP has called Louisiana for Trump.

As expected, AP has called Louisiana for Clinton.

We don’t doubt that Trump is likely to win Louisiana, where he was way ahead in polls and is ahead by 23 percentage points in initial returns. According to Decision Desk HQ, however, the results so far are early votes cast before election day. So how much Trump’s margin declines over the course of the evening, if it does it all, could provide some indication of whether the last few days of campaign news have hurt him.

https://twitter.com/RalstonReports/status/706301495679930368

To riff on what Harry said earlier: combined between the four Ohio border counties to have reported thus far in Kentucky, Trump has won 32 percent of the vote, Cruz 31 percent, Kasich 23 percent and Rubio 14 percent. (These percentages exclude votes for Ben Carson and other withdrawn candidates.) That’s interesting for Kasich heading into Ohio, but also perhaps an auspicious sign for Cruz since he’s nearly tied with Trump. There are no results in yet from highly populous Kenton County, home to Covington, a prominent Cincinnati suburb.

Our colleagues at ABC News just called Nebraska for Sanders. He is leading there by about 10 percentage points with about 75 percent of the vote counted. A win is a win, but his margin is smaller than we’d expect if the race were tied nationally. Not only that, but it’s far less than 35-percentage-point margin then Sen. Barack Obama won the caucus by eight years ago.

Maine is officially called for Cruz, who won 46 percent of the vote and 12 delegates. Trump took 33 percent and 9 delegates, Kasich took 12 percent and 2 delegates and Rubio won 8 percent and no delegates. That’s a very healthy victory for Cruz in the Northeast, where he has otherwise struggled. Trump had easily won in both Massachusetts and New Hampshire, which border Maine. Did Trump fall short in Maine because it was a caucus? Did something change in the past few days that allowed Cruz to crush him? Is Cruz finally coalescing the anti-Trump vote? We obviously don’t know, but it was a good night for Cruz in Maine and a bad one for Rubio and Trump.

So far, 16 Republican states have fully reported results (including Kansas tonight) and in none of them has a candidate won a majority of the vote. The nearest misses were in Massachusetts, where Trump got 49 percent of the vote, and in Kansas, where Cruz got 48 percent. We’ve got too much going on tonight to confirm whether this is a record of some sort, but it looks as though it might be. In the contentious 1972 Democratic primary, for example, Ed Muskie got 62 percent of the vote in Illinois, the fourth state to vote. In the 2012 Republican race, Mitt Romney got (just barely) over 50 percent in Nevada, the fifth state to vote. Note, however, that Trump could plausibly get to 50 percent in Louisiana, where polls close in a few minutes. He was as high as 48 percent in one poll of the state.

The Maine GOP has, after considerable preamble, announced Cruz as the winner.

One minor note of caution about Kentucky, where Trump has a 12 percentage point lead with 26 percent of precincts reporting: the share of precincts isn’t the same thing as the number of votes. That’s especially true in Kentucky, which seems to be reporting every county as a single precinct. The counties reporting so far represent about 18 percent of Kentucky’s population, rather than 26 percent.

Kansas was just called for Bernie Sanders by the Kansas Democratic Party. And it’s worth asking whether it fits the pattern of other states Sanders has won. Kansas is white, of course, but is it liberal? Of course Kansas is not liberal, you’re saying. (Even if there is a Liberal, Kansas.) Kansas hasn’t voted Democrat for president since 1964. But Kansas also doesn’t have that many Democrats period, which means the ones who vote may nevertheless be fairly liberal. According to the 2008 exit poll from the general election in Kansas — there was no exit poll conducted of Kansas’s caucus that year, so this the best we can do — 39 percent of Democrats who voted in the general election were liberal. That’s fairly similar to the percentage of Democrats nationwide who identified as liberal that year, 42 percent. Furthermore, Kansas holds a caucus, which tends to attract a higher share of liberal voters than primaries or general elections do.

If you’re looking for hints ahead of the Ohio primary on March 15, look at the map of Kentucky tonight. You see that Kasich is pulling decent numbers along the Ohio/Kentucky border. He’s at 21 percent in Gallatin county and 26 percent in Campbell county. That could be good news for Kasich as we head into a must-win primary for him.

We’re excited to see if/when Maine is officially called for Cruz, and whether Trump’s early lead in Kentucky holds up. But remember, that doesn’t matter all that much from the standpoint of the delegate math. Maine and Kentucky are both strictly proportional, subject to a 5 percent threshold in Kentucky and a 10 percent one in Maine. Relatedly, Rubio is currently running below that 10 percent threshold in Maine, which would deprive him of any delegates from the state.

There’s clearly some geographic patterns to Cruz’s and Trump’s support in Kentucky. Trump is cleaning up in coal country in the eastern part of the state, while Cruz has gotten his best percentages from the two counties that have reported in the west. A lot more of coal country has reported than the west, so on that basis one would expect Trump’s 11-percentage-point lead to shrink. We are, of course, also waiting on the vote from Lexington and Louisville, which are big population centers.

Democrats are voting in Kansas, Nebraska and Louisiana today, and we still haven’t seen any results yet. But it’s worth taking a look at our state-by-state demographic benchmarks. They suggest that, if Clinton and Sanders were tied nationally, Sanders would win Kansas by 18 percentage points and Nebraska by 17 percentage points, but lose Louisiana by 28 percentage points. Keep in mind, however, that Sanders has fallen well behind pace in the delegate race and will need to not just match but beat those targets as we go forward through the primary calendar.

The real thing we’re waiting for as the votes come in from Kentucky is Fayette County (Lexington) and especially Jefferson County (Louisville). A strong margin in those counties for Cruz could tip the state in his direction, but Trump will likely win the state if he carries both counties.

Although the margin of Cruz’s win in Kansas is a big deal — with all results reported, he beat Trump 48 percent to 23 percent — his performance in Maine might be even more surprising. Decision Desk HQ has now called Maine for Cruz — in fact, it looks like Cruz could win by some margin — although the networks haven’t yet. When I wrote about Maine a few days ago, I said it was unpredictable (that part looks smart) but mentioned Kasich, not Cruz, as the candidate to keep an eye on (that part doesn’t look so smart). Cruz couldn’t have won Maine just by winning evangelicals, since it has relatively few of them (although more than other parts of New England). Cruz also seems to have made inroads with a relatively broad spectrum of conservatives: libertarian-ish voters who nearly gave Ron Paul a win their four years ago; working-class voters Down East. If this all sounds a bit vague, it’s because there are neither exit polls in Maine nor detailed county-by-county results yet. But it doesn’t make Cruz’s performance any less impressive.

Although Kentucky is officially holding caucuses, they are functioning much more like a primary than caucuses normally do. The polls were open for six hours, and there were no speeches allowed. That might be part of the reason Trump may do better in Kentucky than in Kansas or Maine.

Trump won overwhelmingly in “coal country” in far western Virginia and far northeast Tennessee, so it’s not surprising that he did really well in the Kentucky county most representative of coal country to have reported so far: Clay County, which gave 58 percent of its vote to Trump and 23 percent to Cruz. Coal country might be enough to tip Kentucky to Trump, even if he and Cruz run fairly evenly throughout the rest of the state.

https://twitter.com/DecisionDeskHQ/status/706275347348660225

Micah, here’s what I would have said going into the night:

• 4 wins — great night for Trump;
• 3 wins — status quo, in line with expectations;
• 2 wins — shows vulnerability, although Trump is still the frontrunner;
• 1 win — Trump may have major problems, although keep in mind these are small, quirky states;
• 0 wins — no more excuses; this means he somehow lost Louisiana — catastrophe.

Obviously the margins matter too, however. If Trump loses Kansas and Maine big but only narrowly wins Kentucky and Louisiana, that might be closer to the “1 win” than the “2 win” interpretation.

So let’s say Cruz wins Maine, as Harry suspects will happen. If Trump wins Kentucky and Louisiana, won’t the results today not seem so bad for him?

There are reports that Cruz has carried York County in Maine by a small margin. If that turns out to be true, it seems likely that Cruz will pull out the win. York was a fairly good county for Romney in 2012 and is not as rural, where Cruz tends to do best, as the rest of Maine.

Appropriately, the first results from Kentucky came in from Bourbon County, which gave Cruz and Trump 32 percent each, Kasich 18 percent and Rubio 16 percent. It’s been some time since we had a competitive Republican caucus or primary in Kentucky, so the results are hard to put into context. In the past few presidential general elections, however, Bourbon County — which consists of working-class exurbs in the Lexington metro area — has been a reasonably good bellwether, voting about the same as Kentucky does overall.