2016 Democrats, 2016 Republicans, Counting Delegates, New York, Predictions, Uncategorized

Forecasting New York: Final Prediction

April 19, 2016

Donald Trump is going to win. By a lot. You already knew that. We still have several things to figure out. How close does he get to sweeping the 95 available delegates? Does Ted Cruz finish third behind John Kasich? Perhaps most importantly, can Bernie upset Hillary after trailing in every poll?

Let’s start on the GOP side:

Donald Trump 53.8% (84 delegates)

John Kasich 24.4% (7 delegates)

Ted Cruz 20.2% (4 delegates)

We’re now about two-thirds of the way through the GOP primaries. Especially in heavily polled primary states, we’ve found the safest approach is trusting the surveys. Trump virtually always winds up to the lower-middle range of his polling. He doesn’t pull a bunch of undecided voters but also doesn’t lose supporters or fail to turn them out.

In New York, a smaller than average percentage of voters were telling pollsters they were unsure. Given how well they know The Donald, for better or worse, that makes sense. Fewer up-for-grabs voters is good for Trump. I’d be very surprised to see him fall much below 50% or reach 60%.

In the great delegate derby, each percentage point matters. Should Trump underperform and wind up at 48%, he’d grab 70, maybe 75 delegates. If he reached the high 50s, he’d get at least 90, possibly close to all 95.

Since I think he’s going to wind up in the middle of those two extremes, the delegates should follow. We don’t have enough district-specific polling data to build a calculation from. Optimus broke their results out by district, but they have Trump with lower numbers than most of their competitors. Their Wisconsin surveys were less accurate than average, so no logic in treating their numbers as an end-all.

Breaking 50% statewide, finishing over that number in about two-thirds of the districts, finishing second in either the heavily Orthodox Jewish district in Brooklyn, the almost completely GOP free district in the Bronx, or perhaps a random spot Upstate would get him to 84.

Kasich is ahead of Cruz in the polls and Ted doesn’t have any momentum in New York.  That should prevent the type of late break towards Cruz that helped him and sunk Kasich in Wisconsin.

We know the Cruz campaign can always locate their voters and turn them out. There just aren’t that many of them here. In places like New Hampshire, Vermont and Massachusetts, where he’s struggled to reach double digits, the normal Cruz finishing kick didn’t exist. He didn’t fall short of his poll numbers, but didn’t pad them by a few points either.

I do think the combination of his data and ground forces will let him steal a district and finish second in another one or two, preventing a delegate shutout.

Meanwhile, although Kasich is bad at closing and almost never beats his final polling numbers, I do expect him to wind up to the middle-high end of his survey range. Unless/until he gains that elusive outside of Ohio traction, his voter base consists of those who can’t stand Trump and are at least very skeptical of Cruz.

That should equal about a quarter of New York GOP primary voters. Based on pure percentages, he would wind up with several more delegates than Cruz. However, his campaign hasn’t shown anywhere near the same competence. As such, I’m thinking his raw vote edge only gets him 3 more delegates.

If I’m wrong here, it’s probably in giving Kasich too much credit, not short-changing him. There’s a scenario where he gets 5-6 points more of the vote, but Cruz winds up with more delegates.

On to the Democrats:

Hillary Clinton 51.6%

Bernie Sanders 47.7%

I don’t believe the polls. The closest survey is from Gravis Marketing. It was taken on 4/13 only (most are over a period of a few to several days) and didn’t include any sort of useful cross-tab info to page through. Clinton led by 6. That’s the only single digit result.

So far, CBS/YouGov, Quinnipiac and NBC/Marist/WSJ have all given Bernie a fair shake in their polling. They are normally more optimistic about his chances in a given state than the average pollster. None have him closer than 10 points. None showed him making any noticeable progress from the beginning of the month to now.

Overall, Hillary has a 12 point edge in the Real Clear Politics average, and FiveThirtyEight is projecting her to win by 13.5 to 15 points depending on which of their two measures you use. This converts to a 99% chance of victory.

RCP is simply a measure of the most recent group of polls, equally weighted. FiveThirtyEight has a far more sophisticated approach, weighting polls for recency (RCP just cares if it’s before or after their cutoff date), pollster house effects, and various demographic factors.

Both approaches get us to a double digit Hillary win. FiveThirtyEight’s algorithms are still dependent on the pollster data. They can adjust, but are still stuck with the inputs. As such, they aren’t capable of projecting a win for a candidate who trailed in every poll and is at a structural (closed primary) and demographic (lots of voters of color) disadvantage on the ground.

The Sanders campaign is dampening expectations. The Clinton camp isn’t cushioning the media with an explanation for a possible loss. Even in Michigan, they thought defeat was possible. Chief Strategist Robby Mook was telling media sources their huge margin in Mississippi would make up for a narrow defeat there.

But polls have consistently understated Bernie’s support in states where both candidates have actively and consistently campaigned. Beyond the 20+ point miss in Michigan, which looks more and more like an outlier as we move through the calendar, both Wisconsin and Massachusetts were off by 6 to 8 points.

The closed primary in New York makes it tougher for Bernie. He can’t have unaffiliated voters make an Election Day decision to turn out for him. On the other hand, he doesn’t need to worry about college voters being out on spring break. That impacted him in many of the March contests.

This is a long way of saying I’m making an adjustment of approximately 8 points from the poll averages. In similarly contested states he’s managed to beat even his best poll. The one survey taken entirely after the debate showed him making a bit of progress from pre-debate, though not a ton.

Put all that together and you wind up with about a 4 point margin. Enough to keep interested parties watching election coverage for a couple/few hours after the polls close rather than a few minutes. Enough for Bernie to declare moral victory, but not enough to change the direction of the race.

He won’t get embarrassed, but this isn’t Michigan 2.0.




Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s