Public opinion polling and political backlash: “It goes with the turf”

Lee Miringoff is the director of the Marist Institute for Public Opinion at Marist College

Public opinion polling is a staple of political reporting. And campaigns love to second-guess them. So it wasn’t unusual for President Donald Trump to tweet his disapproval of a CNN poll this week that had him trailing Democratic opponent Joe Biden by 14 points.

What was different was the Trump campaign’s demand on Wednesday for a CNN apology and a letter from campaign lawyers calling the poll defamatory while urging CNN to cease and desist. CNN rejected the demand and stood by its poll.

But the dispute highlights the role polling plays in political coverage — and, particularly, in this election because Trump is an inveterate poll watcher. It seemed inevitable that in the nation’s sharply divided political environment, the public opinion industry would become a target as well.

“The whole industry is not immune from what spills over from the polarization that exists,” said Lee Miringoff, the director of the Marist Institute for Public Opinion at Marist College. “It goes with the turf.”

Marist did not conduct the CNN poll, but has carried out surveys with media partners including NBC and the Wall Street Journal in the past.

We interviewed Miringoff this week to get his insights about the current state of polling, the challenges it faces and how financial pressures could affect the frequency of polling during this election cycle.

What was your reaction upon hearing that a presidential campaign had sent a cease and desist letter to a news organization over the results of a poll? In your experience or knowledge, has this type of legal threat over a poll ever occurred?

Miringoff: The short answer is no.

The longer answer is, as they say, I’m old enough to remember 2012…I was in the middle of polling for NBC and The Wall Street Journal in the battleground states and had Obama leading in Ohio, which, at the time, seemed somewhat blasphemous. And we took a lot of criticism. [Obama won Ohio.]

The whole industry is not immune from what spills over from the polarization that exists. … It comes with the turf.

Having said that, Donald Trump with his hundred plus tweets a day taking on CNN for a poll and then hiring (Republican pollster John) McLaughlin to substantiate his position is unusual. And threatening legal action … is part of the threat that he likes to put on to his charges. McLaughlin’s response has a lot of questionable parts to it, in terms of substantiating the president’s position. A lot of quality pollsters would say that’s not valid.

Does it have an impact on how pollsters carry out their work, or how the public reacts to pollsters?

Miringoff: The answer to the first question is not to my knowledge, and certainly we would not in any way react to that kind of charge.

But there are certainly valid answers to these charges about why you don’t have “likely voters” at this point. Or why the margin may be as wide as it is. Why this is different than an exit poll in terms of the party spread — because exit polls at elections tend to have people with greater party identification than you would during the off season like now. So there are more independents in that poll. …

There were six or seven other polls last weekend of varying quality, but they all had double digit leads for Biden. So, the president can talk about the enthusiasm his candidacy is getting, but that’s not measuring it scientifically in a poll.

So are you going to adhere to the scientific principles or not? And he chooses not to, and that’s his choice.

What about from the standpoint of the public? Are they more reluctant to participate because of the polarized environment?

Miringoff: Ironically, the cell phone started to make it harder, but now it’s making it easier. … We can get you where you are. You don’t have to be at home, which was becoming a problem for pollsters literally reaching people. The cooperation rate has not been terrible. It’s still hard to get to people. People sometimes don’t answer.

The proof is in the results. The national polls last time had Hillary Clinton winning the national vote from 2 to 4 points, and clearly they were getting the right people.

Last time we had the introduction of the “shy Trump voter” — the sense that polls were off a little bit because Trump people didn’t think it was socially appropriate to say they were for Trump. 

We know all about the echo chamber. If you are for Donald Trump, you more than likely watch Fox News. The people you work with and your friends and relatives are probably more likely to be for Trump. So you are not socially out of step because you are getting the reinforcement that you want. 

So, yeah, the phone rings, and you pick it up, and you’re doing your interview, and it’s way way too complicated for a person to say, ‘Well, I’m really supporting this guy, but I don’t want to say I support this guy because I don’t want to look like I’m doing something wrong.”

Have you found that, as more people stayed home during the coronavirus epidemic, participation has been a little bit better?

Miringoff: Yes, particularly on the topics that are so important to people right now. Between the 110,000 deaths, and 40 million people unemployed, and cities having demonstrations, people want to have a chat.

When things are uncertain but people have opinions, that leads to people being eager to talk.

When during an election cycle will more news organizations start to partner with poling outfits such as yours?

Miringoff: It’s somewhat shaped by the environment, and I’m referring to cost considerations. Good, quality polling, which has live interviewers talking to people on landlines and cell phones, has gotten a lot more expensive because people are harder to reach.

Obviously the political news cycle affects what we’re doing. So we’ve pivoted from our coronavirus polling to race relations polling.

Now we also have a whole slew of varying quality polls. … Some show up for election time and might not really be bonafide public opinion organizations. We saw that in 2012. There were polls that sort of showed up to try to affect the averages. They don’t have a track record and they were gone after the election.

Unfortunately, the media does not filter that out.  

Are the financial troubles in the news industry affecting the desire for polling?

Miringoff: Unfortunately, yes.

College-based polls have taken over a bigger share of the market —the Marists, the Monmouths, the Siennas, the Quinnipiacs. And given that colleges are under some additional challenges financially right now, I wonder whether there will be a pulling back of some of those college polls.

That could be important because they fill the vacuum as other organizations, like Gannett and others, pull back. So I’m not sure what’s going to happen this fall.

Will you have students on campus to do the poll? We’ve just developed a whole remote system with our software that the interviewers actually don’t have to be on campus … our last poll was done by 55 students in 10 different states. The whole thing was remote.

What’s the biggest challenge facing polling in this particular year?

Miringoff: Coming off 2016, I think the biggest challenge is we have to constantly and convincingly draw distinctions between people conducting scientific opinion research and the prognosticator-aggregator community. They were the ones who last time said there’s an 88% chance or 72% chance Hillary Clinton was going to win.

They use polls, but then they put them through their own special sauce. Members of the public are not probability theorists. To say someone has an 80% chance of winning, they still have a 20% chance of not winning.

If the forecast says there’s an 80% chance of rain tonight and you take your umbrella and it doesn’t rain, you’re not going to care too much. If they say there’s a 98% chance Hillary Clinton is going to win and she loses, well, you’re going to have a bit of a sore spot.

And the answer to that is not that the polls had a bad year. Because they didn’t.

A lot of the state polls, because it takes longer to do state polls, they end early and the relationships with media partners require that. In South Carolina this primary season, in 72 hours the whole thing changed. So if you were not in the field during that time, your numbers would not look like what eventually occurred.

And that was the case in 2016 for the state polls, too…All we know from the exit polls is that it broke late toward Trump from people who disliked both Clinton and Trump.

But this time four years ago, it was a close race. And the Biden lead right now is substantially larger than we’ve seen at this point in a lot of elections. … I don’t know what it means. There’s a lot of things in play here. Who knows what’s going to happen between now and November.

Notify of
Inline Feedbacks
View all comments