We’ve seen this movie before.
Media coverage of President Trump’s polling numbers—his allegedly low job approval numbers, to be precise—veers from breathless to over the top.
Does it sound a bit familiar? Of course it does. It harkens back to the coverage they gave the pre-election polls in the weeks leading up to the November election. You know, when the media was unanimously predicting that blowout win for Hillary Clinton.
They were wrong then. They are wrong now.
Some background: I was one of the pollsters for the Trump for President campaign. Among the states for which I had specific polling responsibility were Michigan, Wisconsin and Iowa.
With that in mind, I would offer the following thoughts on why most of the current public polls measuring the President’s job approval should be taken with a grain of salt: they are making the same two mistakes they made during the campaign.
First, they are not asking the right questions; second, they are not including the right people.
Let’s talk about asking the right questions first. During the campaign, we identified a so-called “Hidden Trump voter”—the voter who was going to vote for Trump without stating it publicly. While these voters came in all shapes, sizes, color and ideologies, a meta analysis revealed them to generally be white voters with a college or post graduate degree—the kind of voters who “underperformed” for Trump in the polls but voted for him in November. The kind of voters the Clinton campaign took for granted.
Hidden Trump voters—or “Undercover” Trump voters as we sometimes called them—had their reasons for wanting to stay hidden.
They didn’t want to be called deplorable or racist. They didn’t want to get in quarrels at work, at the gym, or at their dinner parties. They didn’t want to suffer property damage to their shrubbery or cars due to a lawn sign or a bumper sticker.
So, they kept their mouth shut.
Now, if someone is a Hidden Trump Voter, you can’t just ask if they are a Hidden Trump Voter. They aren’t going to say yes, because they are, you know, hiding.
So you need better ways to ask the question.
We asked whether they knew of anyone who was going to vote for Trump, but just not saying so publicly. In survey after survey, in state after state. In some states, “yes” garnered as much as 40 percent. We asked the same question of Secretary Clinton. The yesses were about half as much on average. We did the math, analyzed the delta in the data and the remaining voters and concluded that there was a surge in store. A surprise in the making.
The same pollsters who missed the Hidden Trump Voter are also missing the Hidden Trump Approver.
There is an unmistakable chunk of voters out there who love what Trump is doing. But they won’t say so publicly.
They don’t want to be called deplorable or racist.
They don’t want to get in quarrels.
So, they keep their mouth shut.
As in the campaign, pollsters can identify these voters by asking the right questions. A question such as, “Regardless of your opinion of the President, do you know of anyone who approves of the job he is doing, but refuses to say so publicly,” might yield some interesting results. Over the course of time and poll results, it will be clear that the Hidden Trump Approver, like the Hidden Trump Voter—exists in significant numbers.
Now, let’s talk about point number two: including the right people.
The media, and ostensibly the Clinton campaign, was surprised by the Election Day surge of rural voters in places like the Florida panhandle, western Michigan, the Pennsylvania T counties and northern and western Wisconsin.
These were a different kind of Hidden Trump voter. But just as consequential to Trump’s margin.
Some of these voters hadn’t voted in years. Some of them voted for the very first time in the 2016 Republican primaries (recall the media’s shock at the size and scope of Trump’s primary wins—more evidence of missing the Hidden Trump Voters). Some of them were registered Democrats—but given their anger and ideology, and the fact that they felt the Democratic party abandoned them—they are better described as Democrats In Name Only.
If you read the tea leaves correctly, and you saw the size of the rallies of the Trump campaign (yes, I went there) you swiftly conclude that if the surveys are to be accurate, you need to at least start with a large number of what we call “0 of 4’s”—voters who had not participated in the four most recent general elections—but who were telling us they were counting the days until election day 2016.
The folks who waited in lines for hours in the rain for a lawn sign or a hat.
Some (not all) of the public polls out there are missing these voters—their models are skewed to more frequent voters and they are not taking the time, care or expense to properly gauge the opinions of these harder to reach rural voters—voters who incidentally are far more likely to approve of the job that the President is doing.
Of the two types of Hidden Trump Voters, it is fair to say that first type was hidden because they refused to answer the poll questions factually. The second type was hidden because they weren’t even asked the poll questions.
One final point has more to do with biased coverage of the polls and less to do with the polls themselves.
The media chooses to focus on certain poll findings while ignoring others. Looking at the recent NBC/WSJ poll, certain findings received virtually no attention from the media. Among them:
- 60 percent of respondents feel hopeful and optimistic when thinking about the future of the country. That is the highest number since 2002.
- 53 percent agree that the news media and other elites are exaggerating the problems with the Trump Administration because they are uncomfortable and threatened with the kind of change Trump represents.
- 57 percent believe it is likely that Trump will bring real change in the direction the country. Among those voters, 63% believe that Trump will bring the right change to the direction of the country.
- 41 percent believe the economy will get better over the next 12 months. While not a high number, compare it to the last time the question was asked under the Obama Presidency—only 24% believed the economy would get better.
- Of those predicting it will get better, a vast majority (73%) believe it will be due to Trump’s policies.
You get the point.
To be clear, Trump isn’t the first President to receive biased coverage from the media. They hammered President Reagan for being out of touch and disengaged. Despite the recent love-fest for President George W. Bush, the headlines against him were brutal. Both, of course, were two termers.
But in our current climate, I’m not optimistic for any kind of peace offering. I suspect the media will pick and choose what to report. The polls will continue to skew badly. And just as in the primaries and general election, President Trump will keep shocking the media elite.
Adam Geller (@adamhgeller) is a Republican pollster and the President of National Research Inc. He served as one of the pollsters for the Trump for President campaign.