Does "nonpartisan" even exist?
The Rasmussen kerfuffle illustrates just how problematic the concept can be for media and polling.
I’ve been thinking a lot lately about perceptions of partisanship in media and polling. Some of that is driven by recent debates about whether a pollster that engages in promoting conspiracy theories should be included in a poll aggregate (more on that later). Some of it is driven by considering my own career choices, past, present, and future.
Up to this point in my career, I have worked in the polling industry from largely nonpartisan positions - first in academia, then at the Marist College Institute for Public Opinion, and from two 501(c)3 nonprofits, which are by definition nonpartisan. I skipped an employer in there, because I’m not entirely sure how to classify HuffPost, particularly since I worked there during the time period when we put Donald Trump on the entertainment page instead of politics. Arianna Huffington was still editor-in-chief most of the time I worked there, and her positions are well-known. But in my role - and on the Pollster team as a whole - we worked from a nonpartisan standpoint. Any rules about what polls were included or excluded had nothing to do with partisanship and everything to do with methodology and transparency.
Because of the HuffPost history, the bent of the nonprofits I’ve worked with, and my own writing and social media, I’m aware that I’m perceived to have a partisan lean. That’s not incorrect; I’m not here to tell you I’m nonpartisan and neutral, although my actual placement on the ideological continuum would probably surprise some people. I can be pretty nihilist sometimes. What I’m here to tell you is that pretty much no one working in political fields is nonpartisan.
Nonpartisan neutrality is a laudable goal, but it’s just that — a goal. Organizations can have that goal and uphold it. Individuals will rarely be able to achieve true nonpartisanship on their own, simply because when one engages with politics constantly for one’s profession, one is going to have some opinions. Keeping individual views private is really the only way to prevent the spillover, and some organizations successfully do that — but it’s difficult in the age of social media.
It’s fine that people working in nonpartisan organizations have opinions. That’s called being human. The role of the nonpartisan organization is to make sure those personal biases are balanced and/or eliminated in what it puts out. Personally, I prefer to know the biases behind what I’m reading or looking at so that I know how to think about the information. What I get tired of is seeing people who are themselves extremely partisan lamenting the lack of nonpartisan information. If you openly campaign for conservatives and then lament that others are liberal, my friend, you are part of the problem of which you speak!
Now, I do think it’s true that most people working in “mainstream” media are left-of-center ideologically, in large part simply because many media operations are located in big coastal cities that are typically left-of-center, attracting young college-educated people who are also generally left-of-center to work for them. It’s a simple probability game that doesn’t favor the right. But the right’s response to that — hemming and hawing about no one being nonpartisan or neutral — is kind of telling on themselves, too. What they want is not nonpartisan neutrality; they want to see their views represented, which are decidedly not nonpartisan or neutral. Of course, the left does it too: Anything not sufficiently left is deemed right-leaning partisan; anyone too close to unpopular conservative figures is suspect (see the unrelenting criticism of Maggie Haberman).
My point here is that “nonpartisan” is extremely difficult to achieve because the concept itself is relative. The Republican version of nonpartisan is quite different from the Democrat version, and even within the parties, the progressive view of nonpartisan varies considerably from the just barely left-of-center nonpartisan view. Each person has their own distinct definition of “nonpartisan” based on what they believe and where they fall. It’s no wonder so few organizations can walk the tightrope. Notably, many don’t really even try except in name - I’ve heard the phrase “nonpartisan does not mean neutral” many times. That’s splitting a hair to some degree by defining “nonpartisan” as simply not explicitly championing one party over the other, but it works for legal purposes.
It gets even more complicated to discuss “nonpartisan” when one side is engaging in bad behavior that doesn’t really have a parallel on the other side. This brings us to the discussion of Rasmussen Reports, a polling outfit that has been using its data to promote and spread conspiracy theories and lies about election fraud.1 ABC News/FiveThirtyEight recently reached out to them to ask some methodology questions and some questions about their political affiliations with a warning that they would no longer report Rasmussen in FiveThirtyEight’s models without satisfactory answers. We don’t know who else ABC News has asked these questions of, but when Rasmussen put out the letter, most on the right were very quick to complain that it was a partisan witch hunt.
To be clear, I don’t know what exactly ABC News’s motivations were, but I strongly agree that a pollster engaging in antidemocratic behavior deserves to be questioned regardless of which side they’re on. And the methodology questions are fair — I couldn’t get them to answer methodology questions in the 2014 and 2016 cycles; they’ve always been opaque about how they do their polls. But because it fits with a right-maligned-by-media narrative, it blew up. I suspect we’ll hear more from ABC News on the topic at some point.
The tricky part is that I can’t name someone on the left engaging in similar behavior. There have been pollsters on the left — and still are — that I don’t really think do great work, ask leading questions, and definitely promote an agenda. But none that I’m aware of are pushing antidemocratic falsehoods. Now, I would have put Data for Progress in the very-bad-actor camp due to its founder Sean McElwee’s unethical behavior, but the company fired him and has moved forward. When there isn’t a counterpart to point to and demonstrate even-handed applications of the questions and rules, it’s incredibly hard to fight back against claims of partisan bias.
You might ask “what’s wrong with just putting Rasmussen’s data in models and seeing what happens?” After all, that was FiveThirtyEight’s previous policy. From a data standpoint, yes, modeling should be able to mitigate the known right-leaning biases that exist in Rasmussen’s data, just like it does with left-leaning pollsters. But when you’re running a public media polling aggregate (something I know a little about), you have to worry about more than just the data. You have to worry about the communications side — the message you’re sending out by including their data and putting their name and their sponsors’ names on your site: You are implying that they are trustworthy and a credible source and giving them a platform to get more attention.
If anyone is unworthy of a platform and implied credibility, it’s an organization that uses its power and its data to spread lies about stolen elections that threaten the very basis of U.S. democracy. If that statement means I’m not nonpartisan, so be it.
To illustrate my point about the slipperiness of the concept, just the way I phrased that sentence gets me branded as a leftie, whereas it shouldn’t be viewed as partisan to state the truth. And that sentence got me branded even further left. But then, my very existence as a [happily] childless late-30s woman living with cats gets me labeled, so I really don’t have a chance at all.