April 18, 2016
We’re now less than 24 hours from voters going to the polls in the Empire State. Unless we see a sudden shift in a survey released at the last minute, there is zero evidence the race is moving. The polls released in the past 72 hours clearly indicate absolute and total stasis.
If you’re Hillary Clinton or Donald Trump, that’s great news. If you’re John Kasich, perhaps at least not bad news. For Ted Cruz, and especially Bernie Sanders, this is extremely sub-optimal.
The campaigns are apparently believing what we’re seeing here. The Trump campaign expects to win the vast majority, if not all of New York’s 95 GOP delegates. He’s sitting at 52.6% in the Real Clear Politics average.
Not only is that almost exactly where he was when we last checked, but each of his last 8 poll results fell between 49 and 55. There just isn’t much variance here. Remember, if he reaches 50% statewide, he gets all 14 statewide delegates.
The remailing 81 are dispensed 3 per congressional district. If he hits 50% in a district, he gets all three. Otherwise, the second place candidate gets a delegate. If Trump winds up exactly where the polls are indicating, you’d expect he’d break 50% in 15 to 20 of the 27 districts.
Unless he loses one of the super majority Democratic districts because one of his opponents got 245 votes while he got 235, Trump should wind up with all but 10 or so delegates. If the campaign is expecting to win 80 to 85, it completely squares with the polls.
Neither Kasich nor Cruz are boasting about picking up significant delegate quantities. They are both spending money in Upstate districts where their internal polling indicates Trump is under 50% and they can at least snag a delegate.
Speaking of the fight for second, Kasich has a consistent, if not substantial lead on Cruz. They currently have a 5 point gap in the RCP average. Kasich is leading in 8 of the 9 most recent polls.
The one exception is CBS/YouGov which recently showed Cruz ahead by 2, 21/19. Two weeks prior, they had Ted ahead 21/20, so no practical difference. This poll consistently shows Cruz with numbers at or near the upper end of his range, regardless of the state.
Kasich usually registers worse than his average. A result like this is consistent with a small overall edge for the governor. The only poll to show movement in their relative position is from Emerson College. Right after Wisconsin, they had Cruz up 5 on Kasich (22/17.)
Over the weekend, Kasich moved ahead by 3. He’s now leading Cruz 21/18, basically in line with the other results. The average survey only has about 6% undecided. There isn’t much room for a last minute shift in a particular direction.
What about Bernie and Hillary?
All you need to know is Sanders is downplaying the necessity of winning. His campaign manager is saying they’ll win plenty of delegates regardless. Bernie says a win would be a historic upset. They’re lamenting the inability of Independents to participate in the closed primary.
The RCP average has him trailing by 12.8 points, almost exactly where he was a week ago. CBS/YouGov is normally a favorable survey for him. The new survey has him trailing by the same 10 points he did two weeks before.
Emerson College has the one survey taken completely after the recent debate/slugfest. It’s the most encouraging poll of the bunch for Berners. He’s trailing by 15. A little over a week before, they showed him down by 18, so at least he picked up a few points.
When a 15 point deficit is your good news……
The day before Michigan voted, Sanders was facing a 21 point polling mountain, so it’s not completely impossible he could win. His better surveys there were similarly showing him down by 10 to 12 points. The difference is no recent New York poll has him losing by more than 17, where he faced deficits of over thirty points there.
Much like New York, there was little evidence of a late break. Polls taken in the final couple/few days, during or after the debate, showed what the pre-debate surveys did. That was an open primary. This isn’t.
Open primaries are harder to survey. Voters can change their mind about which side to participate on in addition to which candidate to support on a side. More variables equals more variance, equals more mistakes.
If the standard deviation between polls is a couple points like we see here in New York, a major error is less likely than in Michigan where it was several points. Bernie does normally beat his survey numbers, but his larger surprises are generally in caucus states, places with fewer polls, or locations with more variation in polls.
All of this is a long way of saying the numbers are pointing towards a Clinton victory. Perhaps not a resounding one, but an important one nonetheless.