In defense of the inconveniently accurate Rasmussen polls

By Fran Coombs, Rasmussen Reports

Here they go again.

The anti-Trump media in its typical pack fashion has begun criticizing Rasmussen Reports in recent days. Why? Because President Trump likes the job approval numbers we’re reporting.

‘We were branded “outliers” because our findings didn’t show Clinton leaving Trump in the dust on a fast train to the White House.’

It’s true that our Daily Presidential Tracking Poll often finds Trump’s public job approval several points higher than other national pollsters do. The same thing was true during the latter years of Barack Obama’s presidency, but for some reason the big media didn’t have a problem with that.

Now it’s true that if we only surveyed the newsrooms of The New York Times, The Washington Post and CNN, Trump’s job approval ratings would be lower than low. Fortunately for our readers, however, we survey the real America outside the newsrooms and beyond the Beltway that girdles the nation’s capital. They give the president higher marks for his job performance out there.

The criticism from the anti-Trumpers is nothing new. Rasmussen Reports was constantly criticized throughout the 2016 election cycle for showing that Trump and Hillary Clinton were in a neck-and-neck contest much of the time. We were branded “outliers” because our findings didn’t show Clinton leaving Trump in the dust on a fast train to the White House.

Two days before the election, a prominent Democratic operative sent us an e-mail asking when we were going to apologize for being so wrong all year.

But a funny thing happened when Americans actually got to vote. Trump defeated Clinton in perhaps the greatest electoral upset in U.S. history, and our polling nailed the exact margin between the two candidates.

Who got it right? The three daily tracking polls – Investor’s Business Daily, the Los Angeles Times and Rasmussen Reports. We’re the ones who were taking the temperature of the electorate every single day, not dropping in for a handful of days like the others — usually after a controversy — for a snapshot of popular opinion.

The rest of the polling industry was way off on the 2016 election results and spent the next several weeks apologizing for its worst performance since Dewey-Beats-Truman. Despite all the mea culpa-ing and public breast-beating, though, nothing much changed in the way they count their numbers.

The Democratic operative who wanted an apology from Rasmussen Reports, by the way, neglected to follow up the election results with an apology of his own.

The big media quickly shifted from trying to stop Trump’s election to sabotaging his presidency, and the polling industry quickly fell in step. At Rasmussen Reports, we shifted some of our demographic margins based on the election results, but it also largely remains business as usual: The difference is, we were right in 2016.

So the anti-Trumpers counting on voters to have a short-attention span are attacking Rasmussen Reports once again for being an outlier and leaning Republican. This is the same stuff we heard two years ago.

Just as we’ve done in previous administrations, we ask pointed questions about most of the major comments and actions of the president and the GOP-led Congress, and quite often the results are not what they’d like to see. But Trump is doing better in our Daily Presidential Tracking Poll than other surveyors say he is.

Why? Well, for one thing, since Gallup left the field earlier this year, Rasmussen Reports is the only public opinion organization that is tracking the president’s job approval on daily basis. The rolling number we post every morning at 9:30 Eastern is based on the responses of 1,500 Likely Voters. We pick up 500 new responses every night Sunday through Thursday and drop the oldest 500 off at the other end.

We go through additional screening questions to survey Likely Voters for the obvious reason that they’re the ones whose responses count: That’s where the electoral rubber hits the road. These people intend to vote. Many other pollsters survey registered voters, but think how many Americans don’t bother to vote in every election. A survey of registered voters doesn’t really tell you much about how an election is going to turn out.

Other companies poll an even wider field by surveying “Americans” which is basically anybody over 18 who answers the telephone. A survey like that on an upcoming election is largely worthless, although it often spawns a big headline.

Then there’s the whole issue of so-called “robocalls.” Rasmussen Reports is often criticized for using an entirely electronic surveying method. We record our questions, and then everyone we call hears that exact same question in the exact same voice with the exact same inflections. Many of our competitors use human callers and think that gives them an edge. But especially in an age as politically mean-spirited and divisive as ours, our results suggest that people are more likely to privately tell the truth to an automated voice than to a real person, especially one who asks the question in a little different tone of voice each time. Maybe respondents are more likely to tell the truth if they don’t think they’re being judged by someone on the other end of the line.

Interestingly, Rasmussen Reports did a survey in late August of 2016 in which we asked Likely Voters: Compared to previous presidential campaigns, are you more likely or less likely this year to let others know how you intend to vote? Or do you feel about the same? Seventeen percent (17 percent) of Republicans – nearly one-in-five – said they were less likely in 2016 to tell someone how they intended to vote. That compared to just 10 percent of Democrats. If you’re looking for why Trump was the upset winner contrary to the pollsters, this is a good place to start.

Like many other pollsters, we now draw a sizable percentage of our survey results from special demographically balanced Internet panels to capture the growing number of Likely Voters who no longer have landline telephones. These panels are increasingly important because they both reach younger Likely Voters and do so while maintaining their personal privacy – just like in the voting booth.

Unlike many of our competitors, there’s a link to each survey’s questions within the first two or three paragraphs of every analysis we post. We don’t load the questions at Rasmussen Reports, so we have nothing to hide.

Each pollster brings their own analysis to the number crunching, what we refer to as our own “special sauce.” We tried to explain this to voters just before the 2016 election in a piece we called, Why Pollsters Disagree .

So it’s 2018, and the forces that oppose Trump are calling us “outliers” again. The president, not surprisingly, likes our numbers and boasts about them which makes our competitors even angrier. We’re not on the Trump team; we’re not on the anti-Trump team. At Rasmussen Reports, we’re just trying to do an honest job telling you what America thinks.

Fran Coombs is the managing editor of Rasmussen Reports.