25 years of the Arkansas Poll
For a quarter century, the Arkansas Poll has been directed by Dr. Janine Parry, a professor of political science at the University of Arkansas. Its longevity, transparency, and effort to present relevant statistics to Arkansans in a digestible fashion are just a few of the hallmarks of its tradition. When I asked about her legacy with the poll, Parry was modest. But when coerced to reflect on it, like any good scientist, she points to the data.
Janine Parry: For me, I like to keep tabs on how many total interviews we've done. And my best estimate, it varies a little bit from year to year, but is somewhere around 23,000 interviews over the years. The interviews are typically at least 60 questions long. So, you know, the scientist in me says, well, that's 1.4 million data points. And mostly, I think that we've informed the public discourse in a way that just elections you know, really can't quite do so. Yeah, it feels good.
Matthew Moore: I wonder as someone who is not originally from Arkansas, was starting the Arkansas Poll a way for you to kind of get a better pulse of politics here in the state?
JP: To be honest, it was actually something one of my colleagues — who has since moved on from the university — but he thought as a policy expert, it was something the flagship school should be doing, and I was the new person so he thought I should be doing it. “Well, you start it!” But I came I came to appreciate almost out of the gate, how powerful it was as a tool and yes, how much it helped me understand the state. It's not just that I'm not from here. It's also that within four months of moving here, I taught an advanced level course in Arkansas politics. So every single time we do the poll, and I kind of look at the patterns develop, it helps me in teaching that course. It helps me on commenting on elections and policies here. It's been an incredible benefit.
MM: Over the last several years, we've seen some hesitancy and some apprehension around polling. There's a lot of folks who are maybe in the camp of throw the baby out with the bathwater when it comes to polling, that they're concerned that polling in general can't be trusted. As someone who has done this for 25 years, what would you say to someone who is concerned about the validity of your polling that you're doing?
JP: I have ample opportunity to consider this question. Because inevitably, somebody who doesn't like the results decides to become an expert in sampling or platforms or that kind of thing, margins of error, they try to dive in. There are reasons to be, I think, a mindful consumer of poll results, because there are lots of bad examples out there for sure. So, you know, if your new station says, “Call in and let us know how you feel about traffic conditions,” well, everybody knows that's not a representative sample. And then of course, the technologies are always changing even at polls that try to be scientific and collect a random sample and that can be zigzagging walk.
So while the latest technique, which is also one of the cheaper techniques of internet polling, or I guess a small piece of that which is sending people links to their texts, and so many people text now, and you can hit the link and then you can participate in the poll. That makes a lot of sense in maybe states that have big metro areas where there are a lot of people with a lot of education and more money. But, in a place like Arkansas, it doesn't make as much sense, the same conflicts and zigzagging performance was true when we move from mail polls, you know, through the Postal Service to telephones because certain people were likely to have telephones and certain people were not and that could affect your results. So there's always reason I think, to be a cautious consumer.
But I would say the hallmarks of a good poll, one that deserves a second look would be one in which you can see all of the questions and all of the details about how the poll was conducted, with no more than one click. If you can't see it at all, that's a no. Move on with your life, consume some other form of information. Or you know if you can kind of dive into the details as we've always tried to do. That's something that's probably worth your time because you can see well how representative is it of who they claim to be standing for. So, you should be able to see not just the margin of error, but how the questions were asked, including the answer options, even the order they were asked. We publish our protocol in full so you can see exactly what respondents heard. So I think that's a really useful indicator of what's likely to have been done conscientiously what's likely at least reaching for the gold standard of polling, and what's just noise or trying to actually influence the conversation rather than represent it.
MM: As we look at the poll that you did, the front page here talks about the most important problem and you ask what is the most important issue facing people in Arkansas today? And more than a third of people said the economy and looking at previous years, this has constantly been a concern. So, I suppose you're not surprised at all when you when you see that the economy is the main concern.
JP: I'm not surprised when I see that. It's usually the economy, education and or health care, those tend to be the big three in Arkansas and in other state polls or nationally. But I was interested this year this was a year in odd numbered years, we ask people, and we don't provide them with categories, you know, we don't force them when, of course, their responses into what we think the top five or six things are. We just ask them and then the interviewers bless their hearts, they type it in as fast as they can. So this is a year in which I get to look and see. And it's really interesting to see all that's captured in what we then end up calling the category economy.
MM: Can you give some examples?
JP: Yes, there are lots of people who said paying the bills or some version of that. So they're looking at it on micro level. There are lots of people who said inflation, they're looking at it on a micro level, but also probably a macro level. They consume a lot of news and understand what's happening there. They'll talk about jobs, they'll talk about wages, a plant closing and their community. It's really kind of an intimate look into the 800 or so people we talk with every year. So it's a broad swath of individuals. Oh, some of them will say like home loan rates. I saw that a few times, since that's gone up. So there's a lot it's kind of a catch all category, but it's certainly one with which almost everyone can identify
MM: The element of the poll that I think certainly got the most attention was around the approval rating of governor Sarah Huckabee Sanders. According to your polling, 48% approval rating for the governor. It's the lowest in 20 years and that dates back to her father, who was governor in 2003. He got 47% Do you think it's significant the timing of the polling around her rating as well? I know that this happened in early October and through some weeks of October this was right after the story started to break about the lectern. This was soon after the special session where FOIA was restricted. There seems to be some opportunity that the timing of the polling played a role in this too, right?
JP: It definitely plays a role and you see that with presidential approval ratings where you can see particular events — because those are tracked, sometimes nightly, but definitely weekly, monthly, quarterly — you can see particular economic shocks or political events, international events, causing a bounce or a sink and a president's rating, so timing matters. And in fact, with the other one that was below 50%, Mike Huckabee, he was struggling with education reform issues, had proposed a major school consolidation plan as a response to the Lakeview cases over equity and adequacy in K-12 funding in Arkansas public schools. So he also was taking a hit right around that time had just come through an election contest that was closer than you would have thought given how strong he'd been before and how strong he ended up being toward the end of his governorship. So, timing matters, yes.
MM: I wanted to take a look — I hadn't looked at this, I don't know if you've looked at this in a minute either.
JP: It’s been a long time!
MM: This is the copy of thevery first Arkansas Poll from November of 1999. One of the things that stands out to me is I think you may have just recently discovered pie charts, because there are a dozen plus pie charts in here.
JP: Yeah, it was so fancy in 1999.
MM: As you look back on the first time you put this together, I imagine you were thinking, “How can I make this accessible to someone who is not interested in polling? How can I entice them to take a look at this and to be interested in this?” Compared to now, I think political appetite has changed and we've seen people are a lot more likely to be forgiving and look at numbers instead of charts. As you think about the first time you were putting this together was a driving force, “I hope someone looks at this…”?
JP: Yes. there was a certain quality of digestibility and, you know, making it accessible particularly to television media. It was never was — and it still isn't — it's a tough story, I think for TV. I really admire that. So many stations still pick it up every year and sometimes they make their own graphics. But yes, originally we were thinking that that would be flashier and it might help people kind of comprehend what we hoped would facilitate more public conversation.
MM: As you think about this being your final year and during the Arkansas Poll, do you think about your legacy? Do you think about the work that you've done to get to this point and you know, the large shoes that will have to be filled by your absence from this poll?
JP: That's sweet of you to say I think about it, I guess mainly in the sense that it's an unusual thing that Arkansas has a scientifically conducted totally public poll, and that it's had one for this long and that the data are collected and stored. Roper keeps the data, the University of North Carolina at Chapel Hill keeps the data. And then, of course, we have it all publicly available on the Arkansas Poll website.
So, Arkansas is unique in a lot of ways, and it doesn't always feel unique good. That's something I've learned from my students over the years, you know, that they seem to feel a sense of, of shame, or at least humility about where we rank in various things. We're, we're high on the things that you don't want to be number one at, and we're low on the things you do, often. But this is actually something that's pretty exceptional, especially in a time of nationalization and polarization and politics, when all the policy is really happening at the state level. We are one of the few states that can really track what people want over time, and I feel good about that. And I hope that it has facilitated and will continue to facilitate some of the conversation.
It's not just the Arkansas legislature. I mean, the Hawaii legislature now is 90% Democrat, 10% Republican. The people of Hawaii are not that skewed. It's the same thing in our legislature, as you can see over and over again, in this poll, whether it's policy preferences or a partisan identification, we're really like a third, a third and a third. Or 60/40 on lots of issues. We can find each other again, if we'll focus on the things we have in common — and I think this poll shows that — and perhaps less on these really hard-edged messages that we get from the two teams competing for our attention and our loyalty.