Trump’s Internal Polls Were Spot On
By Adam Geller
I want to give a gentle pushback against the narrative that “all the polls were wrong” on the Presidential race. Many of them were wrong. Maybe most of them were. But the internal polls for the campaign were spot on.
You may recall that the pundits laughed when Kellyanne mentioned that we spotted a fairly large hidden Trump vote, or as she called them, “undercover Trump voters.” Our polling always showed these folks in fairly large numbers. And they fell off of their chairs on their bright shiny studio sets when we started actively campaigning in states like Wisconsin and Michigan and Pennsylvania. But again, the polling provided evidence that there were cracks in that big blue wall. And we listened to the polls, and trusted them.
Here is one more important point: Simply reading a top line number and reporting it as a result is not analysis. It is not even close. Analyzing a poll when millions of dollars and votes depends on it, is a laborious process. You dive into the cross tabs. You play with different models. You take a holistic approach: who wants change? Who are the less frequent voters and how intense are they are various issues? What did the ’12 model look like and what does a ’16 model look like? At what point do we conclude that a ’12 model is no longer relevant? Where are the key differences in the likely electoral makeup? You get the point. You dive in. It takes time.
If you are down one point in a state where Romney lost by 12, the poll isn’t “predicting” a one point loss; rather it is BEGGING you to spend some time and resources there, because it is a jump ball, not a double digit loss. Remember the movie “The Horse Whisperer?” Well, in this case you need to be a Poll Whisperer. You need to listen to what the poll is trying to tell you, not adhere to models that have no relevance, not to “herd” like so many pollsters do, so that their poll conforms to a narrative.
The polling team for the Trump campaign had some smart people; some of whom you heard of and some of whom you haven’t: Kellyanne Conway, Tony Fabrizio, John McLaughin, Mike Baselice, and yours truly. It was a good, smart group. We worked well together, we played nicely, we agreed, we disagreed and we came together. And we got it right.
Adam Geller is the president of National Research Inc., Holmdel, and a pollster for President Elect Donald J. Trump
Adam, THANK YOU.
A big thank you to the state of Pennsylvania.
I am intimately familiar with 2 of the 3 counties in PA that Trump flipped, which delivered to him the state. These are counties that haven’t voted Republican since Ronald Reagan. They’ve watched their communities be devastated almost overnight by Obama’s policies, and are really suffering and under siege. Very, very sad to witness.
As a reward for their loyal Democrat vote, their plight has been ignored. When some did speak up, they were challenged and told to get back in line. Again, very very sad to witness.
It was only a matter of time before the flood gates burst wide open. If Trump does well by them, I can assure you, they won’t be turning back.
We can only hope against hope that someday New Jersey will wake up and see that same forest through the trees.
Correction, my mistake,
the PA counties I spoke of haven’t voted Republican since 1988. That was George Bush Sr., not Reagan, although they did vote solidly for Reagan earlier in 1984 & 1980.
Apologies.
of pollsters, news people, reporters, and papers/ outlets, who have allowed their personal biases to color, shade, and tilt results, and try and shape outcomes: I trust that folks like yourself, who sample fairly and honestly,will set a better example for future elections. I know that is a tall order, but I believe when our new president and cabinet make it plain that b s won’t be tolerated/ believed, that it will no longer be “cool” to set tone and opinion!
[…] at the post – it emphasises that in looking at the polling data, you can’t just focus on the top […]