People's political preferences come from their identities.
Not from information (or misinformation). The public understanding of science (PUS) model is truly alive and well in politics.
Scholars in my field of science and technology studies (STS) are very hard on a particular model of science and politics that they label the “public understanding of science” (PUS) model. The PUS model is often deployed when policy-makers are thinking about the decline of trust in scientific institutions. Why, when the COVID vaccine in the US was pretty much free, did so many people choose not to vaccinate themselves? Why do so many people—often very educated ones—tend to be suspicious of genetically modified (GM) food even when it has been proven safe over and over again?
The PUS model attributes this to, well, a lack of understanding. The people who distrust science simply don’t know science that well. They have a flawed understanding of it; they do not understand how vaccines work or how scientific research on vaccines was conducted; they may not know the biology of GMOs. If they could understand the science better, the PUS model says, if we could just educate them, then the distrust will disappear and the conflicts will abate.
STS scholars beg to differ. Brian Wynne, whose study of the Cumbrian sheep farmers (on which I hope to write more about), is one of the canonical studies of STS, would explain the disagreements between experts and laypeople differently. Wynne says that the disconnect between the scientists and the distrustful publics is not about public ignorance. It is also not really about the piecemeal scientific question itself (is the vaccine safe? is the vaccine effective? are GMOs safe for eating?). Rather, he argues that the people resisting the scientific establishment have very different outlooks on central questions about meaning and social life; in other words, they have a worldview and an identity that is somewhat different from what the establishment policy-makers or the scientists have.
This worldview is shaped by the history of their interactions with scientific institutions, the government, and much else. When publics reject science or some institution, this is less about them not knowing the facts of an issue; it is about how that scientific institution might fit into their worldview. Vaccine skepticism, for instance, may have less to do with the safety of vaccines per se and more to do with how these groups view the intervention of the government into their daily lives.1
A kind of mirror image of the PUS view operates in progressive politics as well among certain journalists, social scientists, issue advocates and political hobbyists; let’s just call it the “public understanding of politics” (PUP) view. Here, the reason that people don’t vote for progressive candidates in elections is either because they are misinformed about an issue or because they have some kind of internalized, latent prejudice. The rejoinder to the PUP view will be similar: voting preferences are no more explained by thinking about “information” or changed by “education” than people’s attitudes towards vaccines.
Here’s one articulation of the PUP view that I feel like I want to bring up in the light of the recent Presidential Election.
A few weeks ago, I listened to an episode of the political scientist Matt Grossmann’s podcast, The Science of Politics (highly recommended, btw), titled “Are Black voters moving to Trump?” One of the interviewees was Christopher Towler, a political scientist who runs the Black Voter Project at Sacramento State University. The goal of the project is to track the public opinion of black people in America over time.
Grossmann began by asking Towler about the latest findings based on the new round of surveys. Towler began by describing how interesting it was from a social science perspective to have a candidate change (Biden to Harris) mid-campaign. His polling had found that the candidate switch had led to more support for Harris from black voters and that one of the reasons for this was abortion which was a “really powerful [factor] in determining whether or not a Black voter decides to vote for Harris that wasn’t going to vote for Biden in the previous wave.”2
Grossmann then asked Towler what the latest round of surveys had revealed about black voters and their tilt towards either candidate, especially given the analysis that Trump had increased his share of black voters in recent polls. (To be clear, black voters still overwhelmingly vote Democratic; but the margin showed a decrease.)
Towler replied that he was mostly a skeptic on this front and thought that stories of Trump’s improvement with black voters in 2020 were somewhat overstated and were based on samples that “are not very representative of the Black community at large.”3
Grossman pressed him a little bit more on this later on in the interview. If polls showed after the 2024 election that black voters had indeed shifted towards Trump, he asked, how would Towler explain it? Here is Towler’s reply in full:
Christopher Towler: I think from what I’ve seen in survey research and most importantly focus groups, and talking to especially low propensity Black voters, this is where we’re seeing most of the support for the GOP coming from within the Black community. Again, aside from Black Republicans, it’s from independent, low propensity Black voters who haven’t made up their mind, who are on the fence, and might ingest one or two of these GOP talking points. And so there’s been a pretty successful misinformation campaign and campaign to put a few snippets of why the GOP is good for Black voters out there. Oftentimes snippets of information that is not backed in any history or truth whatsoever that have been picked up. And so a lot of sentiment around anti-immigration reform or anti-immigration policy, deportations has shown up in some of the focus groups we’ve done. Sentiment around the stimulus package and small payouts for people has shown up.
But again, when pressed on these issues, when asked more information about these issues, the talking points fall off because that’s all they are is talking points. And so these are mostly things that are picked up amongst, again, low propensity Black voters who are not high information, who are not highly sophisticated when it comes to politics. And so if, for some reason, we do see large shifts, I would chalk it up mostly to a successful disinformation campaign and propaganda that’s gotten out there.
And you can’t write off patriarchy as well. A large part of this increased support, if you will, for Trump amongst Black supporters is coming from Black men, particularly younger Black men who are more susceptible to talking points that require an internalization of oftentimes racism and sexism in ways that cause them to support GOP policies or look towards conservative rhetoric more so than progressives might otherwise do so.4 [my emphasis]
This answer is worth unpacking. First, Towler says, it’s worth asking which black voters might lean towards Trump. They are:
“independent, low propensity Black voters who haven’t made up their mind, who are on the fence”
“low propensity Black voters who are not high information, who are not highly sophisticated when it comes to politics.”
These voters, he argues, by virtue of being low-information and not as politically sophisticated, are therefore highly susceptible to “misinformation campaigns”; these are campaigns that range from “snippets of why the GOP is good for Black voters” to information about “anti-immigration reform.” None of these, he says, are “backed in any history or truth whatsoever.”
If misinformation is the first factor, the second is “patriarchy.” This is because many of these “low propensity,” low-information voters tend to be men; their sensitivity to Trumpian/GOP talking points, i.e., misinformation, comes from their “internalization of oftentimes racism and sexism.” What exactly is meant by the “internationalization” of racism and sexism is not quite defined.
Now Towler is an academic, not a politician, so he is perfectly at liberty to posit explanations.5 But the explanation itself is somewhat cyclical: some black voters might lean towards Trump because (a) they are not as politically engaged, (b) therefore susceptible to misinformation campaigns, and (c), especially, the men, have internalized sexism and racism. All of this makes them more susceptible to “conservative rhetoric” than “progressives.”
But what this boils down to is that these voters are not progressives. They have identities and worldviews that might be better understood as “conservative” which is what makes them open to conservative rhetoric.6 This seems a better answer to me than arguing that these are progressives-at-heart who have been brainwashed by misinformation7 and prejudices.
We need to ditch both the PUS and PUP models.
I think there is however a real danger that scholars and journalists, in trying to explain the identities of voters, can be guilty of identifying them too much with the perspective that the scholars and journalists themselves hold. It can be convenient to think that the voter who distrusts government is really a socialist at heart and will support progressive economic policy. I used my favorite example here but I am sure I’m guilty of this as well. I hope to address this in a future post.
Here’s the full quote:
Towler: It’s only a few times in American history where the presidential nominee has switched in the middle of the campaign. And so, to have data around this moment is really incredible.
And what we found is that support for Harris, both favorability and Black people wanting to vote for the Democratic nominee jumped up a great deal. There was a tremendous increase in favorability. We saw a 19% increase amongst those that were favorable of Harris in April compared to in August. And so, almost a 20% jump in people who were very favorable of Harris since she became the nominee.
And then we also saw a vote share jump, about a 10% vote share jump from Biden to Harris, where back in April, about 62% of Black voters said they were going to vote for Biden. That jumped up to about 72% for Harris, and it’s even higher, 84% if you just look at likely voters. And so, right off the bat, we saw wild differences in Black support for Harris, both favorability and vote share when it came to over that time period where she became the nominee.
[…] so, a lot of what I’ve talked about so far has to do with the what’s happened since we’ve gone from Biden to Harris at the top of the ticket, but we’re also able to model why. And so, we’re able to look at what factors from wave one actually predict these changes in wave two. And not surprisingly, when looking at vote choice, whether you were going to vote for Biden is one of the strongest predictors of whether or not you’re going to vote for Harris. But in addition to that, we also find abortion being really, really powerful in determining whether or not a Black voter decides to vote for Harris that wasn’t going to vote for Biden in the previous wave.
Here’s the full quote:
Matt Grossmann: So in 2020, most analyses suggest that Democrats lost a little bit of Black support relative to 2016, maybe about three percentage points of margin. Obviously, it’s down from the high point of the Obama elections, but even comparing Hillary Clinton to Joe Biden and even when Joe Biden was gaining some white support. So what have we learned since then about the best explanations for that? Or do you view it as just sort of a small change that may not be meaningful?
Christopher Towler: Yeah, I’d stick with the latter here. I view this mostly as a small change that will be insignificant and pretty unmeaningful when it’s all said and done, when we look at the final election results.
I think back in April, as our data suggests, there was a higher proportion of Black voters who were not supporting Biden and were not necessarily attached to the Democratic Party. We did have Black support for Trump at about 14.5%. So it was higher than sort of the historical average of 10 to 12% in the last few election cycles, but nowhere near this 20 to 25% that was coming out and we’re still seeing in some of these Beltway polls that have really small Black samples that I’d imagine are not very representative of the Black community at large. And so, again, from the start, I think this story has been overblown and I think that’s some of what we were seeing in 2020 as well, is that the good data, when we look at these Pew data sets that have really large Black samples that are far more representative and a better depiction of the Black community, […] Black support for Biden in 2020 was right up there with everyone else from years past. And I think we’re starting to see that again now with Harris at the top of the ticket. Black support is definitely coalescing around her. Trump’s support for us in our sample dropped even more with Harris at the top of the ticket from about 14.5% to 13%. And again, 12% among likely voters.
He goes on:
I still believe that when it’s all said and done, we’re going to be right around the historical averages when it comes to support for Trump within the Black community, and maybe above historical average, maybe closer to Obama levels when it comes to Black support for Harris.
A long time ago, I started writing a post called “If all research is politics, then we must learn to act more like politicians.” But it became too long and it’s just sitting in my drafts.
It’s an open question as to what Democrats need to do to get conservative-leaning low-propensity voters to vote for them. I certainly don’t know.
See Dan Williams on why “misinformation” is an incoherent concept.