March 31, 2016
Especially on the GOP side, Wisconsin results are likely to have at least some impact on New York. Beyond that, candidates will have a full two weeks to reach out to New Yorkers. It’s the longest gap between events since voting began in Iowa (I’m ignoring the Wyoming Democratic Caucus on 4/9 that Bernie will get 80% in.)
While we shouldn’t use the first round of polls to make firm predictions, there’s enough recentish data to give us a starting point.
Donald Trump is going to win on the GOP side. We’re looking for margin and delegates. John Kasich and Ted Cruz wound up in the mid-40s in their home states. Particularly if he falls short in Wisconsin, a similar number will make him look very vulnerable heading in to the April 26 contests on the Eastern Seaboard.
On the other hand, if he winds up in the 55 to 60 percent range, the idea of having Kasich and/or Cruz taking delegates by winning individual congressional districts starts looking very far-fetched.
For the Democrats, Hillary is leading, Bernie absolutely needs to win in any scenario where he makes a legit push for the nomination. It’s not a question of who is ahead now. We’re wondering how much ground he needs to make up.
On to the data:
We have four polls taken in 2016. The first two are from Siena College. One at the end of January/beginning of February, the next a month later. Both had Hillary ahead 55/34. Another survey is being taken as we speak. If the results are fundamentally similar, Bernie has a problem.
We don’t know that Siena is measuring correctly, but as no polls have shown Sanders ahead, he needs to begin making forward progress. Intuitively, if he’s trailing nationwide, he’s likely trailing in New York. It’s winnable for him in a world where he’s running roughly even.
He lost Ohio by double digits. There’s no getting from there to a victory here without converting some Clinton voters.
Emerson polled while Hillary was winning Ohio. They show her up in New York 71/23. That’s wrong. It just is. Not only is it an outlier compared to Siena and Quinnipiac, but Bernie isn’t losing any state that experiences winter by that margin.
In places like Iowa and Michigan, we saw a series of surveys that consistently showed Bernie 20 points worse off than the more favorable set. There’s no way to adjust upward from minus 48 to get him a win though. As wrong as the Michigan polls were, none had him this far back.
If Emerson polls again, he needs to cut the deficit to 25 or 30 at least a week out from the vote to have a shot. If you see him within 20 points, that’s almost like a tie.
Our most current sample is from Quinnipiac. They show Bernie trailing by 12. So far, they tend to lean slightly towards Sanders, so this is likely a best case reading of where he stands right now. He’s significantly more popular among Independents.
Bernie has a +65 favorability rating with Democrats. Hillary is +60. Close enough. However, Independents are plus 27 with him, minus 26 with her. If this was an open primary like the Midwestern states, Bernie would already find himself in range.
This one is closed, so he still has significant ground to make up. Most of that ground is actually with white voters. He only trails 66/31 among African Americans. That’s similar to where he wound up in Michigan and Oklahoma, both victories. He might pick up a few more points, but there are practical limits until proven otherwise.
They’re roughly tied (48/47) with white voters. Bernie isn’t going to win like that. He needs to catch up a bit among voters between 45 and 64, where Hillary leads 60/35 (that’s all voters, they didn’t break down by ethnicity.) Clinton has a huge lead among voters over 65.
Since Bernie actually has the higher favorable rating, even among Democrats, his mission is to pull some middle aged skeptics his way. We’ve seen Democratic voters who like both consistently opting for Clinton on election day.
Bernie does far better with Independents and has a huge advantage among caucus goers, but in a closed primary he loses. The next one he wins is the first (Oklahoma was semi-closed.) It’s not an insurmountable barrier, but he’s made almost no progress since voting began two months ago.
In Illinois and Ohio, Bernie started picking up in the last round of polling but had too much ground to make up. He appears closer here than he was in either of those states three weeks out. If he closes as strong as he did in Illinois, or starts earlier than he did in Ohio, he can win.
If that’s going to happen, we’ll see some evidence of improvement within the next 7 to 10 days.
We’re starting to build a decent data set here. Most of the national pollsters haven’t weighed in yet, but we have a new poll from Quinnipiac, and Siena is running a monthly survey. A few others have results too.
With the exception of an outlier from Emerson, which had Kasich at 1% (he’s between 18 and 24 in all others) most of the surveys are telling the same story.
Siena has Trump at the lowest number, 45%. It’s also the oldest survey (almost a month old now.) He was at 34% a month prior, so if his progress continued for March, they’ll have him in the 50s when they publish in the next few days.
The three most recent surveys have The Donald at 50, 55, and 56. If you toss out the 64% from Emerson, it’s safe to assume Trump is somewhere in the 50s. This leaves him both on track for a great outcome and at some risk.
All polls (excepting Emerson) are showing Cruz and Kasich splitting the anti-Trump vote relatively evenly. Unless they strategically break for one or the other, it’s virtually impossible to keep Trump from winning the vast majority of congressional districts if he retains his current level of support.
Even if they go all in on Kasich, at least in the New York City districts where Cruz is likely poisonous, Trump won’t drop many delegates if he’s at 55-58% statewide. If the election was today, he’d clean up.
But the vote isn’t for almost three weeks. Home states are easily influenced by how well their candidate is doing elsewhere (see: Rubio, Marco.) Both Kasich and Cruz saw their poll numbers at home move based on their results on the road. After losing South Carolina and Nevada, Ted needed to spend several of the final days before Super Tuesday shoring up his base in Texas.
Of all the candidates, Kasich is the most popular at home. But he’s also got the worst success rate in other contests. He trailed in Ohio until he managed a semi-decent result in Michigan. That was enough to put him over the top.
In the Siena poll taken partially after Trump’s defeat in Iowa, he was at 34%. A month later after winning New Hampshire, South Carolina, Nevada (and for some respondents plenty of Super Tuesday states) he was at 45%.
Further success has him in the 50s. If he loses Wisconsin on the heels of finishing third in Utah, it introduces a slight amount of doubt. It might not make enough difference to put him at real risk of losing delegates, but a 10 point loss there might well drop him below 50% in New York.
To really cost him, those votes would need to head to Kasich, which is questionable in a scenario where Cruz is the big Wisconsin winner. Either that or current Kasich supporters would need to find themselves #NeverTrumpy enough to go for Ted.
We also don’t know how heavily Cruz and/or Kasich will contest this one with Pennsylvania up a week after. That’s a lot of ifs. If The Donald has anything much to worry about, we should know when the first round of post-Wisconsin polls are out.
If you adjust for New York being bluer than Wisconsin, everything is in line. That’s even with Clinton and Trump having a bit of a home field advantage. Bernie still does better than Hillary against each of the Republicans, though the margin is thinner.
Trump still does worse against the Democrats than his opponents, though Cruz is almost equally inept instead of a rung higher. Nobody is assuming the GOP candidate needs to do well here to win the election, though The Donald brags he can put his home state in play.
At the moment, that’s an utter fallacy. He’s trailing both Democrats by over 20 points. Perhaps more damning, in New York City, which includes his birthplace of Queens and current home in Manhattan, voters are 24% favorable, 70% unfavorable.
Those 24% are a big enough proportion of NYC Republicans to get him plenty of delegates in April, but that’s where it ends. I’m assuming those who know him best aren’t overly susceptible to a fall campaign pivot.
Quinnipiac splits the state into three regions. New York City, NYC Suburbs (Long Island, Westchester County, etc.), and Upstate. Cruz and Trump get obliterated in the city, trail narrowly upstate and by a bit in the suburbs.
Once again, Kasich doesn’t appear a threat in the primary, but matches up very well in a general election. He trails Clinton by 5 (46/41) and Sanders by 10 (47/37). Against both candidates, he’s ahead outside of the city.
This isn’t worth worrying about until he wins somewhere or we start exploring the idea of Kasich as a nominee after losing 49 of 50 states, but the favorability gap with Cruz and Trump is staggering.
Keep in mind plenty of New Yorkers are still undecided on Kasich, a full 38% don’t have an opinion yet. Disclaimers aside, he’s plus 27 statewide and net favorable even in New York City.
Plenty of New Yorkers aren’t ready or willing to vote for a Republican, but he’s more popular than Hillary. She’s 45% favorable statewide, he’s 43%. The difference is her 49% unfavorable rating compared to his 18%.
Even if the majority of undecideds turn against him, he’s still in good shape. Think of it as a more extreme version of the Rubio Conundrum. The candidate most popular with the wider electorate is getting clobbered in the primary.
Moderate voters are even more open to Kasich than Rubio, but Kasich still has fewer delegates than the departed Marco.