You are currently viewing Who doesn’t want the COVID vaccine?

Who doesn’t want the COVID vaccine?

So you’ve probably seen the midwit meme about COVID vaccine skepticism:

The data given to back this one up is this survey:

Now you might be tempted to question some random plot online, so here’s the table from the reviewed, published version of the study:

Scroll down to education level, and you do find the results to have a U shaped pattern, but the numbers don’t fit. Odd? The preprint numbers differ quite a bit:

So you might wonder whether this is a case of researchers trying to bias their results in the politically correct direction by reducing the sample of PhDs who were hesitant. I couldn’t figure out why their numbers changed across the versions of the paper:

The sampling itself was done via Facebook and it’s representative, and could in theory involve dishonest responding, or activism by vaccine skeptic groups. Still, in general, the proportion of trolls in a survey is normally not sufficient to really stir up the results, unless one happens to be analyzing some small subgroup, such as PhDs. In this case, that’s 2% of the sample, which is within the lizardman danger zone. The survey is absurdly large, I don’t know how the hell they supposedly found 10k+ people with PhDs. I wrote to the corresponding author of the paper, Robin Mejia, and she replied:

Dear Dr. Kirkegaard,
As discussed in the paper, a small group of respondents who appeared to complete the survey in bad faith was identified.
From the methods:
“Respondents who did not complete the questions on vaccine uptake and intent (N = 365,426), or reported gender as, “prefer to self-describe,” (N = 31,664), were excluded, resulting in a sample of 5,088,772; self-described gender (selected by <1% of responders) had a high prevalence of discriminatory descriptions and uncommon responses (e.g., Hispanic ethnicity [41.4%], the oldest age group [23.2% ≥75 years] and highest education level [28.1% Doctorate]), suggesting the survey was not completed in good faith. A sensitivity analysis was conducted including all gender responses.”
From the results:
“A sensitivity analysis including participants selecting “prefer to self-describe” gender is provided in S8 and S9 Tables. Overall, results were similar; however, hesitancy prevalence was higher for a few categories (e.g., age ≥75 years, Hispanic, and Doctorate) where mis-reporting was suspected.”
We did not have space to include more detail in the paper, but among respondents who chose to sell-report their gender, many chose options such as “Apache Attack Helicopter” (a trans-phobic meme), or wrote transphobic or other discriminatory statements indicating they were not answering in good faith. Those individuals tended to describe themselves as vaccine hesitant. On further investigation, those individuals were also likely to select that that they had a doctorate (as well as other specific demographics, primarily being >75 and Hispanic). While the total number of respondents who chose self-described gender was <1% of the sample, because the total percent of individuals with PhDs in the sample (and in the US) is small, their inclusion led to an elevated rate of vaccine hesitancy for that category. We did retain the analysis with that group included in the sample and report it in Supplemental Table 8.

So let’s look for a replication. One can find a lot of studies that report this kind of survey, but finding a study that reports results separately for the PhD group, which is usually very small, is difficult. However, this study seems fairly good:

Objectives Acceptance of COVID-19 vaccination is attributable to sociodemographic factors and their complex interactions. Attitudes towards COVID-19 vaccines in the United States are changing frequently, especially since the launch of the vaccines and as the United States faces a third wave of the pandemic. Our primary objective was to determine the relative influence of sociodemographic predictors on COVID-19 vaccine acceptance. The secondary objectives were to understand the reasons behind vaccine refusal and compare COVID-19 vaccine acceptance with influenza vaccine uptake.
Study design This was a nationwide US-based survey study.
Methods A REDCap survey link was distributed using various online platforms. The primary study outcome was COVID-19 vaccine acceptance (yes/no). Sociodemographic factors, such as age, ethnicity, gender, education, family income, healthcare worker profession, residence regions, local healthcare facility and ‘vaccine launch’ period (pre vs post), were included as potential predictors. The differences in vaccine acceptance rates among sociodemographic subgroups were estimated by Chi-squared tests, whereas logistic regression and neural networks computed the prediction models and determined the predictors of relative significance.
Results Among 2978 eligible respondents, 81.1% of participants were likely to receive the vaccine. All the predictors demonstrated significant associations with vaccine acceptance, except vaccine launch period. Regression analyses eliminated gender and vaccine launch period from the model, and the machine learning model reproduced the regression result. Both models precisely predicted individual vaccine acceptance and recognised education, ethnicity and age as the most important predictors. Fear of adverse effects and concern with efficacy were the principal reasons for vaccine refusal.
Conclusions Sociodemographic predictors, such as education, ethnicity and age, significantly influenced COVID-19 vaccine acceptance, and concerns of side-effects and efficacy led to increased vaccine hesitancy.

Their results by demographics:

Their sample is obviously not representative either, with a quite large PhD/JD/MD group. Still, they don’t find any replication of the U shape pattern at all, PhDs were the least COVID-19 skeptical group.

Of interest is that the healthcare worker group is slightly more skeptical. Indeed, what we really want to know is whether expertise on the topic is related to skepticism or not. PhDs in general have more expertise, but most PhDs are not medical experts. Healthcare workers presumably know more about healthcare than the average person, so their increased skepticism is cause of concern. On the other hand, most healthcare workers are low level staff, majority female, so maybe their result is just due to sampling a lot of nurse helpers. So what happens if we look at surveys of healthcare workers in more detail? I found a meta-analysis of such surveys:

COVID-19 vaccines were approved in late 2020 and early 2021 for public use in countries across the world. Several studies have now highlighted COVID-19 vaccination hesitancy in the general public. However, little is known about the nature and extent of COVID-19 vaccination hesitancy in healthcare workers worldwide. Thus, the purpose of this study was to conduct a comprehensive worldwide assessment of published evidence on COVID-19 vaccine hesitancy among healthcare workers. A scoping review method was adopted to include a final pool of 35 studies in this review with study sample size ranges from n = 123 to 16,158 (average = 2185 participants per study). The prevalence of COVID-19 vaccination hesitancy worldwide in healthcare workers ranged from 4.3 to 72% (average = 22.51% across all studies with 76,471 participants). The majority of the studies found concerns about vaccine safety, efficacy, and potential side effects as top reasons for COVID-19 vaccination hesitancy in healthcare workers. The majority of the studies also found that individuals who were males, of older age, and doctoral degree holders (i.e., physicians) were more likely to accept COVID-19 vaccines. Factors such as the higher perceived risk of getting infected with COVID-19, direct care for patients, and history of influenza vaccination were also found to increase COVID-19 vaccination uptake probability. Given the high prevalence of COVID-19 vaccine hesitancy in healthcare workers, communication and education strategies along with mandates for clinical workers should be considered to increase COVID-19 vaccination uptake in these individuals. Healthcare workers have a key role in reducing the burden of the pandemic, role modeling for preventive behaviors, and also, helping vaccinate others.

Associated factors that favor compliance:

Concerning factors associated with lower COVID-19 vaccination hesitancy and higher willingness for COVID-19 vaccines, the male gender was found to be an enabling factor in the majority of the studies (25/35 = 71.4%). Older age or having a doctorate or postgraduate education were associated with higher COVID-19 vaccine acceptance rates in more than half of the studies (23/35 = 65.7%). Compliance with vaccines, confidence in vaccines, or history of influenza vaccination was found to be a predictor of COVID-19 vaccine acceptance in a little more than half of the studies (18/35 = 51.4%). Similarly, direct patient contact/ caring for COVID patients or higher perceived risk and fear of being infected with COVID-19 were associated with lower COVID-19 vaccination hesitancy in more than half of the studies (20/35 = 57.1%). Other factors associated with higher COVID-19 vaccination acceptance were: White or Asian race, higher income/education, medical risk, and chronic disease history, not being infected with COVID-19 in the past, knowledge about COVID-19 infection and disease, working in non-rural areas, and beliefs that vaccines may protect friends, family, and community members.

The male advantage is something we don’t hear a lot about. The media narrative is that Trump/nationalist voters are bad because they are vaccine skeptics. E.g., Pew Research has an article Americans who relied most on Trump for COVID-19 news among least likely to be vaccinated. Danish media ran a story about how members of the New Right party were vaccine skeptics spreading disinformation etc. True enough, nationalist voters are more hesitant, it’s just that the vaccine skepticism surveys show other results that don’t fit the narrative well:

  • Men rather than women are more accepting of COVID vaccines (indeed, any vaccines)
  • European ancestry is associated with more acceptance, the least accepting group are African ancestry people, and Amerindian (Hispanics) are also trailing. Asians are more accepting. You know this pattern
  • Older people are more accepting than younger people.

If you are wondering: no these patterns are not due to the focus on healthcare workers. This meta-analysis of normal people surveys finds “Younger age, females, not being of white ethnicity and lower education were common contextual factors associated with increased vaccine hesitancy.”.

The favorite group of the media is young, non-male, non-European leftists — unfortunately 3 of the 4 demographics here are vaccine skeptics, so these don’t get mentioned so much.

Returning to the expertise question, the surveys of healthcare workers show that more expertise — whether in form of degrees, personal experience with the disease or patients, or tested knowledge — is associated with more vaccine acceptance. This Singaporean survey of graduate students find that studying a healthcare related field is associated with COVID vaccine acceptance. There is no support for midwit model. We begin to suspect something is wrong with the first study. The samples sizes are too large for this to be random sampling error, so it can be sampling bias, dishonest reporting, or research fraud. I can’t think of another factor. I emailed the authors to inquire about details and data sharing.

To be fair, I was able to find one piece of evidence in the other direction. This 2021 Israeli study used contrasting, hypothetical vaccines to see which factors were associated with hesitancy. They find that doctors and life science students were strongly against mRNA based vaccines when there was an alternative:

How come? I guess because it was done before the vaccine roll-out, when mRNA technology was not well known:

Vaccine hesitancy is a global health threat which may hinder the widespread acceptance of several COVID-19 vaccines. Following the collection of 2470 responses from an anonymous questionnaire distributed between October and November 2020 across Israel, we analyzed the responses of physicians, life science graduates (biology, virology, chemistry, etc.), and the general public to whether they would obtain a COVID-19 vaccine with particular vaccine characteristics such as vaccine country of origin, technology, side effect profile, efficacy, and other attributes. Physicians and life science graduates were least likely to accept a vaccine based on mRNA technology (30%) while the general population seemed to adopt any vaccine technology if the declared efficacy is above 90% and the country of manufacturing is the USA/UK rather than China or Russia. However, current inoculation rates in Israel far outpace our predicted rate. Our results highlight the importance of tailored vaccine educational campaigns based on population demographic details and specific vaccine concerns.

Can we find a newer survey of healthcare workers about mRNA? I couldn’t but this study of Italian students, also done pre-rollout found… the opposite pattern with more acceptance of mRNA based vs. viral vector. So I don’t know what to make of this result.

I don’t think that mitwit meme for COVID-19 vaccines is correct, PhDs are probably more accepting of COVID-19 vaccines, and this is true also for healthcare professionals. These are the people with the most expertise, so insofar as we trust the opinions of experts (pro, contra), this should provide some grounds against COVID vaccine skepticism. The results also show that personnel with more personal experience with COVID-19 in their work were more in favor of vaccines. This mirrors the typical media stories of [distraught doctor seeing 100s of dead COVID patients trying to convince public to get vaccinated].

For those looking for personal disclosure: I am vaccinated (Johnson), and already had COVID-19 (delta and maybe omicron). I oppose vaccine mandates. I think it is important to decouple one’s opposition to government hysteria and overreach about COVID (I oppose most measures), and one’s opinion about the vaccines. The vaccines are somewhere between quite and somewhat useful, depending on who you are and what strain is around. Not very useful for people below age 50, not very useful against omicron. I think it’s completely nuts to mandate these vaccines of children.