The Politics of Science: Why Are More Democrats Vaccinated Than Republicans?

From the Left

Republicans and Democrats are not getting vaccinated against covid-19 in equal numbers. The large and persistent enthusiasm gap primarily originates from a partisan divide in trust in scientific institutions.

A recent Monmouth poll released December 15 showed that 96% of Democrats have at least one dose, followed by 79% of Independents, and 54% of Republicans. 30% of Republicans say they will likely never get one, compared to 2% of Democrats.

I compiled data of vaccinations per state, and compared the rates of fully vaccinated with the proportion of the state that voted for Trump. The data was collected on November 15 from the Mayo Clinic Covid Tracker. A theme emerges:

The legitimacy of scientific institutions, such as the ones that manufactured and promoted the vaccine, is a partisan question. Republicans’ relative reluctance to get vaccinated reflects their lack of trust in a scientific elite that they increasingly associate with the political left. Democratic enthusiasm for the vaccine starts with a higher degree of trust for those same institutions, which include colleges, think-tanks, international organizations, and government departments. 

Anyone who perceives institutional science as standing in opposition to political goals or symbols they care about is far more likely to be skeptical of the vaccine than someone who sees institutional science as advancing their political goals and symbols.

Science Respecting

Dr. Anthony Fauci, the director of the National Institute of Allergy and Infectious Diseases (NIAID), is the subject of many friendly and hostile memes, and is essentially the human face of the partisan science trust gap.

But is there any empirical evidence that Republicans trust science less than Democrats? The answer is a complicated and qualified “yes,” and depends in part on what the respondent understands the word “science” to mean.

Last July, Gallup published survey research showing a 34 point gap in confidence in science between Democrats and Republicans. 79% of Democrats said they had a “great deal” or “quite a lot” of confidence in science, compared to 45% of Republicans. This 34 point partisan gap in confidence was exceeded only by President Biden (49 points) and the police (45). Of all the institutions Gallup inquired about, science ranks 3rd in partisan controversy behind only the President and the police.

Honestly, I am not in love with the way this question was asked, or the response options given. If someone asked me, “excuse me sir, but how much confidence do you have in science?” my answer would probably be “huh?”

The question does not specify what is meant by “science.” Are they referring to the scientific method? Or are they referring to authoritative scientific institutions? I suspect that most who answered the question were visualizing the latter, but I can only suspect it.

Science can be understood as an epistemology (the scientific method); as a body of work produced by scientific institutions (“we should follow the science”); a political signifier (“I believe in science”); or as a rhetorical device intended to humiliate others (“that’s science!”). The respondent is free to decide which science they mean when they consider their degree of confidence in it. I do think however, that a 34-point gap speaks for itself.

The different interpretations of the word “science” are crucial. In anti-vaccine media, it is actually quite common to hear pundits assert the authority of “real” science — meaning science that isn’t merely a label to promote a leftist agenda. When we consider the partisan gap in trust in science, we need to focus on trust in scientific institutions and not the process of reasoning from observation, which forms the basis for hardly any political beliefs, and this includes those who enthusiastically took the vaccine. (Seriously, consider for a moment how many of your core political beliefs come from systematic observation, experiment, and empirical evidence, and how many others come from who you decide to trust).

The only proxy available to test this is partisan opinion on colleges. This data does suggest that the science trust gap is driven by institutions.

Research from Gallup in 2017 shows that Democrats have higher confidence in colleges than Republicans, 56% to 33%. When asked why they didn’t have confidence, Democrats’ most common answer was the cost, but Republicans' most common answer was that they think colleges are too liberal and political.

This is important because it suggests that Republican distrust in science is motivated in large part by their perceptions of institutions of higher learning as being politically motivated. This distrust in institutions of knowledge, which rely heavily on their reputation and authority, manifests in other ways too, such as Republican suspicion of the World Health Organization (WHO).

Pew Research Center has shown that the decline in Republican opinion on colleges started sharply in 2015. Pew asked respondents if they believed colleges were positive or negative for their country. In 2015, 54% of Republicans said colleges were positive, and 37% negative. In 2017, those numbers were nearly flipped, 36% to 58%. Democrat support has remained high and consistent.

The brief fixation with hydroxychloroquine and the continuing fixation with Ivermectin also say something interesting about how trust in science is mediated by one’s understanding of the word. Both medications are produced by “Big Pharma” and presumably made in part by liberal college grads, not unlike the vaccine. Why don’t they have the same liberal taint as the vaccine, then? I think it is because these medications, especially Ivermectin, are perceived as being in their control. Ivermectin is more available and established, and doesn’t require deference to Fauci and major scientific institutions in such a flagrant way. By misusing medications that Fauci and the WHO caution against, vaccine refusers are asserting their own agency and independence from the institutions they perceive as politically corrupt and hostile.

The emphasis on fitness and natural immunity amongst those who do not want the vaccine is also an assertion of defiance and independence: “we can handle this disease on our own, without help from liberals like Fauci.” There isn’t anything stupid about bringing up natural immunity by the way, I am only noting that the intensity and frequency with which it is invoked is related to defiance of authority perceived as leftist. The idea here is that scientific institutions are ignoring natural immunity and physical fitness as health indicators in order  to advance a political agenda, but this stems first from the perception  that the institutions are both political and untrustworthy.

The phrase “Do your own research” expresses something similar: don’t be dependent on leftist scientific institutions for information.

Political Bundles and Higher Education

But why is it that scientific institutions are perceived as friendly to the left and hostile to the right?

The short answer could very well be that it is simply the truth. It is hard to answer this question without noting that the available evidence on politically salient issues such as global warming and evolution favor the political left in the United States.

But this is complicated by instances in which Democrats selectively reject scientific consensus. For instance, Democrats are far less likely than Republicans to support building more nuclear power plants, despite the former being more concerned about global warming; or their trouble recognizing the findings of human embryologists about when human life begins because of the perceived threat it carries to their positions on abortion.

I think another way of looking at it requires considering the political socialization that college students experience that separates them from non-college students. Not all, or even most college students study a natural science, but a college degree is highly predictive of vaccine acceptance. I believe formal education, especially higher education, familiarizes its students with “political bundles,” or what beliefs, symbols, and demographics “go together.”

One superficial sign of a politically engaged person (though not necessarily an informed one) is their knowledge of these associations. These associations try to help us to recognize patterns — if someone is a vegan, someone may assume they are also an environmentalist and a feminist. If someone is a gun owner, someone may assume they are a conservative who voted for Trump. 

But such associations can rarely be deduced logically or mathematically. Some political “packages” like simultaneous hostility to abortion and birth control seem to defy mathematical logic, since birth control tends to reduce abortion rates. These associations are learned socially, and the more formal education you have, the more aware of these associations you are likely to be.

Awareness of these associations provokes different trust reactions in partisans as it relates to scientific institutions. If it is true that scientific institutions and their findings are associated with the left, and if it is true that people with college degrees are more likely to be aware of this association through political socialization, we would expect trust in “science” to be positively correlated with Democratic partisanship and more education, and we would expect a more complicated relationship for Republicans with more education.

It is true that those with college degrees are more likely to both vote Democrat and be vaccinated. The Public Religion Research Institute found that four-year degrees reduce vaccine refusal rates and increase vaccine acceptance rates across all racial categories they tested (white, black, Hispanic, multiracial, or other).

Could this be because Democrats and educated people are just smarter, so they just know that vaccines are a good idea? Unlikely.

We should be cautious in assuming that more educated people are “smarter.” I say this not just because many of the biggest dolts I’ve ever met were college educated, but because of the way higher education is specialized. I don’t see why someone with a Bachelor’s in Accounting, or someone with a PhD in Computer Science would know any more or any less about vaccines than a high school dropout. 

Public polling data does support the idea that Democrats warm up to science as they acquire more education, but not so much for Republicans.

A study from Pew in 2019 found that 84% of Democrats with high science knowledge thought scientists should play an active role in policy debates, compared to 58% of low-knowledge Democrats. For Republicans, the relationship was the reverse: 40% of Republicans with high science knowledge responded that scientists should be active in policy debates, compared to 52% of low-knowledge Republicans.

The same Pew study also found that Republicans with high science knowledge were more likely to see scientists as susceptible to bias than Republicans with less science knowledge.

A 2018 study by Michael Tesler of University of California, Irvine found that liberals who paid more attention to political news and who were more educated were more likely to accept the existence of man-made global warming. However, conservatives with more knowledge of political news and higher education were more likely to deny the existence of man-made global warming.

In my view, learning political symbol bundles requires a political education of sorts. For example: highly-educated Republicans “know” they are “supposed to” reject global warming, whereas a less educated Republican might refer to personal experience or gut feeling instead of education, and therefore be less likely to get the answer “correct”.

One problem with my thesis is that Republicans without college degrees are less likely to be vaccinated than Republicans with one, according to data from the Kaiser Family Foundation, although the gap isn’t very large.

This data may be influenced by workplace or city requirements that might affect educated Republicans than less educated ones, but this does provide some evidence that higher education erodes vaccine skepticism among Republicans. It might be fair to say that the effect of education on Republicans’ faith in science and vaccines is mixed.

The relationship that partisanship and scientific knowledge have with trust in science is not the same when you ask explicitly about the scientific method: 

Here we see that Republican trust in the scientific method increases with scientific knowledge, though not as dramatically as among Democrats. This reflects that Republicans, and maybe Democrats as well, really do understand “science” as institutional, and need to be asked explicitly about the scientific method if that is what the researcher is actually curious about. 

This data also shows that although respondents understand the scientific method differently from scientific institutions, there is also a partisan gap in trust in the scientific method, and it is primarily driven by educated Democrats.

It could also be the case that Democrats perceive scientific institutions and the scientific method as trustworthy, whereas Republicans’ relative hostility to science is only slightly reduced when “science” refers to the scientific method. Like with many polls though, it isn’t clear if the respondents understand the terms being used, such as “scientific method.”

New Data Analysis

Using the same state-by-state vaccination data at the top, as well as various demographic data on the state level, I tried to illuminate the partisan vaccine gap beyond what available data already has. Specifically, I am interested in exploring the relationship between education, partisanship, and vaccination status.

I want to examine this relationship in order to test my claim that higher education informs people of the associations among political symbols, and therefore creates a partisan gap in “science” trust.

If you don’t know how to read a regression table, that is fine, I will explain in English what my findings were, and you can just skip over the tables.

My first finding supports the rest of the data: educational attainment predicts higher rates of vaccination at the state level. No surprise there.

But when you account for the proportion of the state that voted for Trump, education loses its statistical significance. This means that a conclusion cannot be asserted either way with confidence when you hold voting behavior constant. This suggests that education does not have an independent effect on vaccination rates, and it only matters insofar as it predicts partisanship. 

Put even simpler, states with high rates of college graduates tend to have higher rates of vaccination, but only because they also have higher rates of Democratic voters. This is important because it suggests that college grads are more likely to trust the vaccine not because of their education as such, but because they come to trust symbols of the left, which includes science.

I also accounted for the percent of the state that is white, as well as some proxies of vaccine availability: poverty rate, and the number of licensed physicians per 100,000 people in the state. The statistical significance of partisanship, measured by the proportion of the state that voted for Trump, holds up. This means that states with a higher percent of Trump votes have lower vaccination rates, even when controlling for those other factors. Not a shock really, but it does demonstrate the power of partisanship on vaccination rates.

Using data from the American National Election Studies (ANES) data from 2020, I also analyzed individual respondent data.

The first finding in my analysis is that Republican distrust of the vaccine is not tied to a broader distrust of big business or the federal government in general. In fact, Republican partisanship is positively correlated with both in 2020 (Trump was still President then). “Big business” was the best proxy for Big Pharma in the data set, but it is not a particularly strong one, I confess.

Unfortunately, ANES does not ask questions that measure confidence in science, so this question can’t be tested further using individual survey questions. But it is interesting to know that Republican skepticism of the vaccine, and Democratic enthusiasm for it, are not driven by gaps in their respective confidence in the Federal Government or big business.

Conclusion

We have theories of information distribution (too much time on social media or watching cable news, etc.) We need a theory of information acceptance. Encountering wild misinformation does not explain vaccine skepticism, at least not on its own. We need to understand what drives political trust, and in this case, political trust in science.

Future public opinion research on vaccine skepticism, as well as vaccine acceptance, needs to interrogate what people think science is. Some questions could take a multiple choice approach, but others could leave the question open-ended, and categorize their answers after the fact. 

When somebody barks “that’s science!” or “the science says,” we really need to ask them what they are visualizing when they say that. I suspect in most cases, it is nothing at all.

One self-congratulatory study by Gallup concluded that Republicans just aren’t used to scientific and rational thinking, and instead rely on emotion and intuition. This is actually how basically everyone thinks, at least most of the time. When liberal science respecters listen to doctors and experts, they aren’t actually thinking scientifically or even rationally. They are intuiting that the person is credible because they are a scientist, and this sensation of trusting someone is an emotional response to identifying with that person.

I do, however, believe that emotional thinking can lead to what I believe is the correct conclusion: that getting vaccinated is a good idea. Liberal political emotions and intuitions are such that they are more likely to accept the findings and recommendations of the scientific community as it relates to vaccines, but that should not be confused with scientific thinking.

Science as a politically operative word, or as a signal of political loyalty, is “left” (whatever that means). This “science” as political signifier needs to be understood as separate from science as epistemology by survey researchers and data analysts.

Not everyone is a Republican or a Democrat, though. This article does not explain why Independents may or may not get vaccinated, nor even why many Republicans have been or Democrats haven’t been vaccinated. It has only tried to address the large partisan gap between Republicans and Democrats, which proved to be hard enough.

Paul Mulholland is a freelance political reporter and data journalist who covers a range of topics including social movements, conspiracy theories, policing, tenant-landlord issues, and public opinion in both writing and video. He has a Lean Left bias.

This piece was reviewed by Managing Editor Henry A. Brechter (Center bias), Director of Marketing and Media Bias Ratings Julie Mastrine (Lean Right), and CEO John Gable (Lean Right).

Paul Mulholland is XXXXX. He has a Lean Left bias.