DANIELLE KURTZLEBEN, HOST:
Let's turn back the clock. It is a week before the New Jersey governor election, and Monmouth University's Polling Institute has just released a survey indicating a win for Democrats by 11 percentage points. And they weren't alone. Other polls gave incumbent Governor Phil Murphy similar odds. Well, now we know Murphy's margin of victory was razor-thin over Republican Jack Ciattarelli. So what went wrong with the polls? Patrick Murray is director of the Polling Institute at Monmouth, and he joins us now. Patrick, thank you so much for your time.
PATRICK MURRAY: It's my pleasure, Danielle.
KURTZLEBEN: You wrote an article for nj.com with the headline, "I Blew It - Maybe It's Time To Get Rid Of Election Polls." That is a big statement. Tell us what prompted that this time around.
MURRAY: It's something that I've been thinking about for quite some time, and one of the things that worries me about polling where it is right now - we've always had our share of misses. And I've been fortunate to have many more hits than misses in my career. But if the misses start coming a little bit more frequently and that makes people sit up and say - on both sides of the aisle I've heard this - I don't trust polls anymore. Then the stuff that we do that's really important - because the horse race stuff isn't all that important, but the stuff that we do - it's really important, such as tracking public health, you know, understanding where public policy is, what the behaviors are, what people's needs are at the kitchen table, which is stuff that polling does really well - is if we undermine trust in the whole polling operation, then we're not doing a very good job for ourselves or for the country when polling is a very good mirror to hold up to what's going on in the world. If election polling can't do the same as regular polling or public policy polling, then that's where my concern is.
KURTZLEBEN: Well, and as we are talking specifically about election polling here, tell us a little more about that. Why is that so much harder than taking the public's temperature on approval ratings or on a particular issue area?
MURRAY: Yeah, the first rule of sampling and polling is that you know who your population is before you even start. Oh, I'm going to poll this group, and I know exactly who they are. Well, when you say, I'm going to poll the people who are going to vote in the election coming up in a couple of weeks - well, you don't know who they are until they actually show up. So you have to make models, and the polling models have generally been good. But in this much more volatile electorate right now - and by volatile, I just mean that you've got this angry partisan divide in the electorate that makes it very difficult because many of those voters are saying, I don't want to participate in anything. I don't want to let you know who - I don't trust any of you. The other part of election polling is there's this focus on the horse race, and I wonder if we asked every single question that we asked but didn't ask the horse race question or didn't report the horse race question - whether cable news would care about polling and reporting it or not. And I think that's a question that we have to ask ourselves.
KURTZLEBEN: What would we lose if pollsters stopped asking horse race questions and stuck only to more substantive policy questions and attitudes? And what would we gain from that?
MURRAY: Yeah, I think we would challenge the media. Would they - without the public polling, would they then just be more susceptible to spin that comes out of the various parties that are involved? I don't know what that would look like. But then again, that's their responsibility. It's the media's responsibility. And my responsibility is to make a decision about, what information can I provide that I feel confident that I can stand by?
KURTZLEBEN: So in terms of looking ahead then, what do you see as next for the Monmouth Polling Institute? How will these results change the work that you do? Or are you still deciding?
MURRAY: Yeah, we're still deciding. I think the horse race question is the one that I'm going to be looking at closely of whether that's even worth asking in that same way anymore.
KURTZLEBEN: And how long do you have to decide? Midterms are really coming up quick now.
MURRAY: Yeah, I was already giving this much thought before we even had Tuesday's election. So this is not something that spur of the moment on my part. This is something that I've been accumulating evidence.
KURTZLEBEN: Sure.
MURRAY: And what happened this past week with the elections is just simply more evidence that now I'm going to take a look at and, you know, make a decision on.
KURTZLEBEN: That's Patrick Murray, director of the Polling Institute at Monmouth University. Patrick, thank you so much.
MURRAY: My pleasure. Transcript provided by NPR, Copyright NPR.