Where politics is personal not partisan

Election Analysis Isn’t Just About the Polls

Election Analysis Isn't Just About the Polls

If pollsters and the media learn nothing else from this election, they should have learned to not rely on just polls.

“The numbers were a little low in this precinct,” the campaign manager said as he passed the printout of the election results to me.

I looked at which one he was talking about and replied, “Nope. The candidate skipped the little league play-offs there.”

He barely looked up, and passed another sheet along, “Very low here.”

“Well, what do you expect? The man he beat in the primary to make it to the general lives in that precinct,” I shoved the papers away, so I could focus on my own.

“But the polls…”

I put my pen down, took off my reading glasses, and looked at him, “But the polls said what? That our guy was ahead? Of course they did. Do you seriously think that the people we called were going to admit that they were going to skip party loyalty over a kids’ baseball team or the fact that their neighbor lost?”

“Well, no…”

“No, they weren’t going to admit that they were going to vote over anything so petty. But, that’s reality.”

The manager pushed back from the table, and shook his head. “Where the hell do you get this stuff?”

“Didn’t you see my Ouija board in the closet?”

He laughed, but I knew he wasn’t sure if I was joking.

After over twenty years working in various capacities on political campaigns, my final “niche” was in research and post-mortem analysis. The latter initially involved rehashing polling numbers until I couldn’t sleep without numbers flying through my dreams. I bought the concept that all the answers are in the numbers, until I learned that they aren’t. The numbers are just a distorted mirror image of the people, and it’s always distorted because of human nature. Ironically enough, the reason that is true is one of the first lessons you learn as a pollster – people lie to pollsters, period.

Once I figured that out, I stopped paying attention to just the numbers, and started listening to the people. I’m not suggesting that I went out there on the streets to talk to everyone, but I certainly did pay attention to the messages left in the office by voters. If there was something on the community calendar, I knew about it. Morning reads included obituaries, society pages, youth sports reports, and police blotters. When there was a bar in the building at a campaign event, I’d duck out and talk to the drinkers. The people in the back of the hall, or left outside were my target when there wasn’t. The ones in the front row were either already sold on the candidate or determined to disrupt the event, so I didn’t worry about what they had to say. People on the edges or outside mattered to me, because they usually told me what a candidate was doing wrong in the eyes of people who could be swayed.

I’ve had more than a few people say that paying attention to those small details doesn’t work on a national scale, but it certainly does work up to even statewide races. In the best circumstances, a presidential campaign is a centrally controlled collection of state-level campaigns, at least to a very basic extent. That didn’t necessarily happen this cycle, though. With all the talk about the Clinton political machine having a good ground game, it seems quite a few people missed something. Beyond the numbers, there are quite a few clues out there, and CBS News graciously gave everyone a chance to see some of them, even if they didn’t necessarily realize it.

In a perfect world, every major political consultant in the country would commit the following names to memory, as a reminder about this election cycle: Sopan Deb, Hannah Fraser-Chanpong, Erica Brown, Jackie Alemany, Kylie Atwood, Sean Gallitz and Alan He. Those seven people were the digital journalists for CBS during this campaign, and John Dickerson failed to recognize the value of what they said to him and America in the video above. To the pollsters and numbers people, that’s just anecdotal information. People like me, with the political Ouija boards in our closets, see the observations of generally new journalists on the ground as priceless nuggets of information. Young journalists haven’t had time to become completely cynical, and they actually talk to the people.

One of them pointed out that political journalists are conditioned to listen to data instead of people – she is right. She was also right about the fact that she saw the signs of the Clinton loss on the ground as it was happening. The social media meme where no one could point out anything good about Hillary Clinton arguably sunk her on the ground in Ohio. West Virginia coal miners coming out against her in the pouring down rain were a bellwether for Pennsylvania, even in the cities where that coal fuels the power stations. By the numbers, low black voter turnout turned Pennsylvania red, not increased “angry white men.” Also, the claims from Salon that long waits at polls in urban centers were a Republican tactic to suppress black votes doesn’t hold water. A simple Google search for videos of lines at polls in 2008 and 2012 shows that lines didn’t matter. The problem wasn’t with obstacles in front of voters – it was a problem from apathy or just plain dislike for the candidates.

Of course, there’s also a problem with the media not understanding how the numbers work. Seasoned journalists who have been covering politics mostly understand what is happening, and their only fault is relying on the numbers too much. But, they do get it. The running in circles that we’re seeing among pundits is because the numbers that usually don’t lie are not giving them a satisfactory explanation for what happened. It doesn’t help that the infotainment set just don’t get the numbers at all, and the public thinks that they do.

Nate Silver of FiveThirtyEight deserves a medal for doing the interview above with Trevor Noah on The Daily Show. Noah simply didn’t get how Silver and his numbers were right, even though they didn’t predict the Trump victory, per se. There’s a longer version of that interview out on the web, but most of the public will only see the clip offered on air. And that means that there’s a good chance that they’ll think that Silver was just tap dancing to cover for a massive blunder. He wasn’t. Silver’s numbers came closer than just about everyone else’s, because he placed Trump’s chances of winning in the double-digits. Most everyone else grudgingly gave him somewhere around a 3% chance of winning. Yes, it is like the weather, and no it’s not an exact science. Remember, at the root of this is a pile of people who were asked who they intended to vote for in the election.

And that’s where the Ouija board comes in again. The question I asked myself from the beginning was “how likely is it that large numbers of people polled would lie to the pollsters?” Given the high pressure out there, and how Trump managed to depict himself, the chances were very high. Every time he said or did something despicable, he increased the chances that people would lie, and tell pollsters that they weren’t voting for him. Bluntly, both campaigns were an embarrassment. The better question might have been, “how many people are willing to actually tell the truth?”

The results of this election were predictable, and I stand by my previous contention about that. Here, I’m just showing a portion of the rubric I personally used to reach that conclusion. It’s not scientific, but it works. It’s not likely I could ever teach anyone how to replicate it exactly, because it is subjective and relies on a fair amount of instinct. But, this isn’t about numbers only. It is about understanding “group think,” and more importantly, recognizing what really matters to people. I can’t explain why a precinct didn’t come out to support a candidate because he didn’t bother going to a little league play-off game, but I know that’s what happened. Conversely, I also know a Republican Congressman who has easily carried deep blue precincts repeatedly, simply because he’s taken flights home to press the flesh during homecoming and children’s Halloween parades. Yes, you read that right – he came home to see kids in costumes. I don’t know why that matters so much to that group of voters, but it does. I just learned how to recognize the fact that it matters in the first place. There are no polls or numbers out there to explain it, either. And that’s what happened to Hillary Clinton this time around. She simply didn’t recognize what mattered to the voters. She assumed that historical trends would be on her side, instead of paying attention to what the people were probably trying to tell her. All politics are personal. The kids on the bus for CBS said that, and they were right.


Last updated by .

Liz Harrison
About Liz Harrison 66 Articles
Political commentator, former campaign operative, media executive, legal and medical writer, literary editor and publisher. Founder at Vigilant Liberty Radio, podcaster and radio talk-show host, and a sexual freedom activist.

Be the first to comment

Leave a Reply