Accusations of indoctrination fly in both directions in American education, and both rest on a shared premise: that political intrusion into schooling is wrong. But little research has asked what the public actually thinks impartiality requires, or whether people apply those requirements evenhandedly to allies and opponents alike. In this Heterodox Academy virtual colloquium, Eric Torres — Postdoctoral Fellow at Harvard’s Radcliffe Institute and Senior Fellow at HxA’s Segal Center for Academic Pluralism — presents two survey experiments on attitudes toward political impartiality in K-12 and higher education, with a combined sample of 2,037 Americans.
He finds widespread support for impartiality in principle. But when respondents judged specific acts of political favoritism, that principle bent: partisans rated identical conduct significantly more harshly when the beneficiaries were opposing partisans rather than their own side. Torres discusses what that double standard means for research, policy, and classroom practice — and where his data cannot yet go. A transcript of Torres’s talk follows below.
Eric Torres: So today I’m going to be talking about how the American public thinks about political impartiality in educational contexts. I’m going to present the results of a couple of survey studies that I did over the last year or so that I think shed some interesting light on what to me, what I think is a really important and pressing question in the politics of education. So to provide a little bit of background, as I’m sure everybody attending this talk is well aware, there’s been much furor in American politics about the politicization of education.
And by that, I mean the intrusion of the inappropriate intrusion of political considerations into what should be politics free domains. So a couple of examples of this recently, last year, Donald Trump, the Trump administration issued an executive order accusing K through 12 schools throughout the country of being rife with radical intent to indoctrinate students into left-wing worldviews, PEN America and many and others replied that in fact, the administration’s ostensible, project to undo indoctrination is itself imposing, political ideology on educational processes.
We see similar dynamics in the context of higher education, where critics both inside and outside the university have argued that the well-documented, ideological imbalances in faculty and controversies around policies in higher ed are turning these institutions into political indoctrination centers. On the other hand, initiatives to increase ideological diversity in the academy have themselves been called acts of political interference. And so we have a situation where partisans in America are accusing one another of politicizing education. This raises or rests on the premise that political intrusion in education is a bad thing.
But somewhat surprisingly, there’s very little research on whether or not the public agrees with that premise and how the public thinks about the relationship between political considerations and education more generally. Much of the recent work on this subject has focused on how American partisans think about controversial issues in education, so how issues of race and gender should be thought. But there remains quite a bit that we don’t know and that hasn’t been studied, such as whether the public wants schools to be politically impartial in principle, and whether they apply their standards of political impartiality even-handedly to making judgments about cases that involve their own team and the other guys.
And even though, to just say a little bit more about that first question, whether the public wants schools to be politically impartial. Although I think for many of us that might seem obvious, there is some evidence, especially coming from the field of political psychology, that as the country has polarized and as partisans increasingly see the stakes of education or of political conflicts as existential, that we might expect a sort of abandonment of principles of impartiality or expectations of norms or impartiality on the grounds that these conflicts are so important that we have to use schools to culture students into our way of thinking about things, or simultaneously, because our way of thinking about things is just obviously correct.
Liberal worldviews are right. So of course, it’s okay to promote liberal worldviews in schools. Conservative worldviews are right. Of course, it’s okay to use schools to promote conservative worldviews. And so figuring out the degree to which, where the public falls on these questions, I think, is quite important. So today’s research, I’ll be describing two surveys, first a pilot and then a follow-up study that I’m just now getting into the thick of analyzing that are aimed at helping us understand how Americans are thinking about these things. The first survey is a pilot study looking at support for impartiality in both K through 12, or I’m sorry, just in the K through 12 context.
I won’t spend as much time there because I think the HxA audience is probably more concerned with higher ed. The follow-up study, which is larger, better powered, and it involves hypotheses that I pre-registered looks at these issues in the context of both K-12 and higher ed. So first, very quickly, this first study comes from an opportunity that I had when I was a grad student. So I did my PhD here at Harvard as well, and I was an affiliate of the Center for American Political Studies. I was able to get a couple of items onto a poll in 2024, the Capps-Harris poll, which is a nationally representative sample or a nationally representative poll that’s run monthly in conjunction with Harris polling.
And these responses were collected on March of that should say 2024, not 2025. So they’re a little they’re a little old now. But I was able to get a handful of items on this one of these polls that were distributed to a random sub sample of 528 respondents. It. So not too many and not too many items, but it gave me an opportunity to probe some questions that I had and to see if they were worthy of follow-up. So I’ll just quickly show you what those items are and then describe the results and then jump into the second study, which I think will be more substantively interesting.
So first, I just asked people straightforwardly to select which of the following statements they agreed with most, that schools should promote liberal views, schools should promote conservative views, schools should not promote any specific political views, or I don’t care if schools promote specific political views. These are non-exhaustive, obviously, but they give us sort of a coarse sense of how the American public is thinking about the relationship between political socialization and schooling and whether or not they are, you know, the degree to which they’re comfortable with schools being used explicitly as vessels for what we might think of as either indoctrination or acculturation into a political community.
Second, I asked a question that sought to get at the relationship between what they wanted and what they perceived to actually be happening. So I asked whether they believe public schools in their area were promoting liberal views, whether they believe they’re promoting conservative views, whether they believe they were not promoting any particular political views or whether they didn’t know. And then I also gave them an option in the first two cases to indicate whether or not they thought that was a good or a bad thing, just both as a sort of consistency check and as a way to see if there are differences between abstract principles and principles applied to the local context. And finally, I gave them a, and there was, I included a survey experiment.
So I gave all the participants a vignette in which a teacher tells students in his 10th grade government class who he supported for president in the last election. I chose this because this is an issue about which people disagree. I don’t think there’s necessarily a right answer about whether or not teachers should, in the context of teaching, be able to disclose their own political views. But there were three versions of this scenario, one in which he expressed support for Trump, one for Harris and one for an ambiguous and unnamed candidate.
And these were randomly assigned to participants so that we could look at the effect of matching between their own political ideology and the expressed support in the vignette and see if that affected whether how they evaluated the appropriateness of that behavior, whether they said it was wrong for the teacher to tell the students, whether they said it was fine for the teacher to tell the students, or whether they said they didn’t have enough information. And so here’s what I found. First of all, I saw that a majority of Americans across groups, about 50% preferred that K-12 schools refrained from promoting liberal or conservative worldviews.
I next saw that a plurality of people were unsure about what schools were actually doing this sort of makes sense. Most of us don’t have a lot of insight into our local schools. But unfortunately, twice as many, more than twice as many, believed that schools were involved in some form of ideological promotion, either of liberal or conservative worldviews, mostly of liberal worldviews. And only 17% believed that they were not. And perhaps most importantly, I found that Democrats and Republicans alike were about one and a half times more likely to judge teachers’ political disclosures as inappropriate when they supported the opposing party. So this is suggestive, but this is a small sample.
And so I wanted to rerun this study and dig a little bit deeper in order to understand some of these issues better. I wanted to figure out for one, what it means to promote liberal or conservative worldviews. And I thought I might be able to do that by asking about more discrete practices in I read and K through 12 schools. I wanted to know whether Americans had similar sorts of attitudes about impartiality in K through 12 and higher education. And I wanted to see whether we observed these effects of what we might think of as bias and judgments of others politicized behavior across a variety of scenarios rather than just a single one in which we’re asking about the relatively unique case of telling students who you happen to support politically.
So this second study I carried out in the spring of this last year, not very long ago. This study was funded by my fellowship at the Siegel Center for Academic Pluralism, and the hypotheses and design were pre-registered with the Open Science Foundation. I contracted with Verasight, the survey firm Verasight. They composed a hybrid probability, non-probability sample of around 1,500 Americans. There’s some prior work on their sampling strategy is found to perform similarly well to other probability-based samples, although we should be cautious about inferring exactly what the American public in general thinks from samples like these.
The results that I’ll present of the descriptive items are weighted for representativeness, and these items were collected actually just a month ago. So this is fresh data. Instead of three items, this time I had an opportunity to put together a full survey. So there’s 75 items on attitudes about impartiality in higher ed and K-12 education, including how people define impartiality, perceptions of what’s actually happening, like in the pilot. I asked about their trust in these different institutions and again ran a series of experiments assessing their consistency of judgments about various policies and practices.
Today because I only have so much time I’m going to focus just on a quick overview of their normative attitudes about impartiality in higher ed as well as to review the results of the experiments. The subjects that I explored in the higher Ed context included the deliberate value cultivation. So again, should schools be attempting to promote liberal or conservative worldviews? Should they attempt to promote tolerance for multiple perspectives or a commitment to advocating for one’s own worldview? View. I looked at whether how the public felt about selection, political selection criteria for both students and faculty. Should admissions take students’ politics into account? Should hiring committees take candidates’ political leanings into account?
I looked at attitudes about viewpoint diversity in the faculty, how important it was for students to be exposed to a range of views versus to get an education from faculty who all hold similar views, similar questions about syllabi. I looked at norms of social inclusion and exclusion on campus, how important people thought it was for schools to be places socially where if you’re liberal or conservative or neither. You felt like a member of that community. And I looked at treatment of controversial speakers as well. So here’s just a snapshot of some of those items. And these are pretty reflective of the patterns that I saw. So first, I’ll show you this one.
I asked respondents how colleges and universities should try to influence students’ political views. I gave them the options where that they should encourage students to adopt liberal or left-wing views, moderate or moderate views, conservative or right wing views to develop their own independent views, or if they should not try to influence students political views at all. And here’s what we saw. I’ve disaggregated these. I’ll just presenting the results only among people who identified as either Democrats or leaning Democratic and Republican or leaning Republican, just to kind of be concise here. And you see basically no appetite for at least the explicit promotion of any particular ideological or political worldview.
But we do see some differences here in support for whether schools should be helping students develop their own political views or whether schools should be trying to stay out of the politics game entirely. With Republicans being pretty evenly split between those two sort of North Stars and Democrats being more in favor of schools attempting to help students develop their own views. Next, here’s an item looking at just a binary choice, asking whether the public thinks that colleges should teach students to understand multiple political perspectives or teach students to advocate for their political beliefs. And here we have a really striking finding where Democrats and Republicans are in remarkable agreement.
About 80% prefer that schools attempt to promote students’ ability to understand diverse views, whereas only around 20% of both Democrats and Republicans want schools to be teaching students to advocate for their own views as a priority. I also asked whether colleges and universities should take applicants political views into consideration, whether they should give preferences to students with some political set of political views, whether they should try to generate a mix of political views in the student body, whether they should not take political views into consideration at all. Here again, we see a really remarkable consistency between Democrats and Republicans.
It’s a little bit cut off on the left there, but you see again very little appetite for giving preferences to students on the basis of their politics. And surprisingly, against my expectations, very little appetite for attempting to create a deliberate mix of views on campus. Instead, at least in the context of admissions, Americans seem to think that political views shouldn’t factor in at all. I asked the same question about faculty hiring and got a slightly different result, where here there’s more enthusiasm for deliberate attempts to create a mixed faculty on campus. But again, majorities of both Democrats and Republicans are saying that they prefer that political views didn’t factor into hiring at all.
And finally, here’s a handful of items on how issues that cover politically controversial subjects should be taught. So I asked about what readings should be on the curriculum, whether students should read mostly liberal, moderate, or conservative perspectives, whether they should read a mix of different perspectives, or whether they should read the best work regardless of perspective.
And here again we see a little variation, but again, that the general patterns are of consistency where most the a majority in fact of both democrats and republicans would prefer that students be exposed to a mix of perspectives and some amount number of democrats republicans around hovering around 40 percent feel that you know if the best work is all of one ideological camp so be it as long as they’re we’re teaching the best work. Here, a similar item assessing how people think how much how important they think it is to learn from faculty members with different political views. And this one again I just gave them sort of a binary choice.
Is it possible to have a good education from faculty who all the same views, or students must study with from faculty with a mix of political views to get a good education. Here we see a little bit more significant difference between Democrats and Republicans, but the trends very similar. Both Democrats and Republicans largely agree that a good education, at least on subjects that involve these controversial issues, requires faculty political diversity. Potentially at odds with their opposition to deliberately cultivating a mix of faculty perspectives in hiring, but an important finding, I think, nonetheless. Now I’ll just quickly gloss over these handful of experimental items.
So here I gave them a scenario following a similar approach as I did in the pilot. I gave there were these scenarios in which some content, the politically valence, the political valence of the scenario was randomly altered. So there was a control scenario in which the politics of the actor in question are ambiguous. And then there’s one in which the actor in question is sort of has a right wing affinity or a left wing affinity. Here we have a faculty member in the philosophy department saying that they are not going to hire a new candidate despite him being very qualified because of his political views. And again, we now the option categories here are sort of a five point Likert type right to wrong plus a I don’t have enough information to decide.
I’ll talk a little bit, talk a little bit about how I analyze this in a moment. What we see here. So here I’m presenting the results of exclude excluding those who say I don’t have enough information to decide. I actually have a separate analysis to see whether partisan alignment affects that. The likelihood of indicating that you don’t have enough information. But just looking at those who felt like they did have enough information. We see something very similar to what we saw in the pilot, which is that when the actor is aligned with their own political leanings.
So when a Democrat is exposed to a liberal actor, in this case, a faculty member excluding a conservative on the basis of his views, or a Republican is exposed to the version of the vignette in which a conservative actor excludes a liberal on the basis of their views, they rate that as about one point on the five point scale more acceptable. So like less severely bad. In this case, we see the movement between approximately 1.4 and 2.2, which corresponds to the shift between saying that was a little bad and that was quite bad. And the control sits right in the middle. A second scenario here. We’ve got the college Republicans, the college Democrats, or an unnamed campus group inviting a speaker to campus.
And there we find out that activists of opposing persuasion are planning to disrupt the event. We ask whether how much they approve or disapprove of that disruption. We see again a pretty striking pattern here where those exposed to aligned versions of the vignette are much more likely to endorse the activist disruption. And here a scenario in which a state legislature is passing a law that requires teaching a particular course with a particular leaning on social justice. Here we look at whether or not they approve this law mandate. End date. And again, we see more or less the same pattern, although in this case, there’s the control behaves very similarly to the condition in which they’re exposed to the opposing group, right?
I pooled the results of these three as well as the scenarios in K through 12. So I had three similar survey experiment vignettes in the K through 12 context that I didn’t show you. And As I had predicted in the pre-registration and as we saw in the pilot study, the effects of partisan alignment on judgments about the appropriateness of these various sort of politically inflected behaviors in higher ed are pretty straightforward. This is also actually a pretty large effect that we’re observing here. So again, it’s about one scale point on that five point scale. And when we pool these results, what we see is that the movement is from essentially ambivalence in the case of scenarios in which you are aligned with the actors in question.
So you say, if it’s someone in your team doing the thing, we say, well, it’s neither right nor wrong. But when it’s somebody else, the average response becomes it was wrong. That’s an important qualitative shift in addition to just a shift of magnitude. I should say too, I don’t think I mentioned this is a multi-level mixed effects model here. So we controlled for the content of the vignettes and included a participant level random effect since each participant was exposed to a random assortment of these vignettes. So let me wrap this up quickly here. First, by saying, what are some of the things that this research does not tell us?
So for one, I don’t have any evidence of how, I’m sorry, why the public endorses impartiality in principle on these items and how they’re thinking about the trade-off between using schools as organs for advancing ideological interests versus as organs for education. I don’t know whether they are motivated by pragmatic or principled concerns, for example, whether opposition to schools teaching liberal conservative views comes from a belief that is inherently wrong or because they think it’s politically intractable to do that. If you were to pursue that policy, it would generate a backlash.
We don’t know whether the motivation is primarily a concern with fairness or a belief that a marketplace of ideas is better for student learning or for the country in some way. I don’t really know if these attitudes that we see, this is a survey study. What people say on surveys may, for one, be driven in part by research or demand effects. They might look at this and say, well, of course, the right thing for me to say is that I oppose, you know, politicization of education in these different ways. Whether or not the discrepancies that we picked up in the experimental section translate into differences in behavior, I think warrants further study.
I do have some unanalyzed items that I will be looking at that instead of outcomes asking about moral judgments about the appropriateness of these acts. They are asked to endorse punishments of varying severity. So that maybe provides us a little bit of interesting texture to see if this is just like a right wrong thing or if this also translates into support for more severe punishment for out groups when they do the same thing. What can we make of these results? I think the evidence of shared values is worth taking seriously. At the very least, we see opposition to bald politicization across basically all of the policies and procedures that I inquired about here.
Like I said, it’s not impossible to imagine a world where polarization has become severe enough that people would say, no, in fact, I do want my schools to be promoting my particular worldview. But that’s not what we’re observing. And I also do think this evidence of double standards is meaningful. So these experiments do suggest that partisans are more likely to call foul on politicization when it’s done by the other team. And then that might have really important implications for designing policies that can protect academic freedom, viewpoint diversity, and the rigorous and pursuit of knowledge in a divided America.
So I think there’s more to be asked about here, but it’s my hope that this helps us, those of us who care about this issue, to get a better feel of how the public is making sense of these controversial issues. So with that, I’ll pause. I’ll just say thank you, everybody who’s here for listening. I look forward to talking with you. Thank you to HxA, the Siegel Center, my colleagues here, both for funding this research and for this opportunity. So with that, I will stop presenting and turn it back over to my colleagues.
