YouTube this week suspended Sen. Rand Paul's account because of a video in which the Kentucky Republican claimed that "most of the masks that you can get over the counter don't work" as a safeguard against COVID-19. "The virus particles are too small and go right through them," Paul says in that video, which shows him speaking with a Newsmax interviewer. "They don't work. There's no value."
YouTube said Paul violated its "COVID-19 medical misinformation policy," which among other things forbids "claims that masks do not play a role in preventing the contraction or transmission of COVID-19." While conceding that "private companies have the right to ban me if they want to," Paul said he was troubled by the fact that the major social media platforms seem to be insisting that users toe the official line on COVID-19, which makes it harder to criticize ill-founded positions and policies. YouTube, he said, is acting like "an arm of the government."
Paul has a point. But in this case, his flat, categorical statements about cloth masks are stronger than the scientific literature supports, relying on a couple of cherry-picked studies with known limitations while ignoring countervailing evidence.
In a video responding to his YouTube suspension, Paul reiterates that "most of the masks that you get over the counter don't work" and "don't prevent infection." He argues that "saying cloth masks work when they don't actually risks lives," describing it as "potentially deadly misinformation." While N95 respirators are effective at preventing virus transmission, he says, "the other masks don't work."
Paul would have been on firm ground if he had said cloth masks offer less protection than N95 masks. But the claim that cloth masks "don't work," meaning they offer no protection at all, is inconsistent with multiple studies suggesting that they reduce the risk of infection, especially when worn by carriers but possibly also when worn by other people in their vicinity.
Paul cites two studies to back up his belief that cloth masks are ineffective: a 2015 study of health care workers in Vietnam and a 2021 Danish study that compared people who were advised to wear masks with people who weren't. Neither study proves that cloth masks "don't work."
The 2015 study, which was reported in BMJ Open, involved about 1,600 hospital employees who were randomly assigned to groups that used "medical masks" (meaning "disposable medical/surgical masks"), used "cloth masks," or followed "usual practice," which notably "included mask wearing." The researchers measured each group's rate of "clinical respiratory illness (CRI), influenza-like illness (ILI) and laboratory-confirmed respiratory virus infection." Here are the results they reported:
The rates of all infection outcomes were highest in the cloth mask arm, with the rate of ILI statistically significantly higher in the cloth mask arm (relative risk (RR)=13.00, 95% CI 1.69 to 100.07) compared with the medical mask arm. Cloth masks also had significantly higher rates of ILI compared with the control arm. An analysis by mask use showed ILI (RR=6.64, 95% CI 1.45 to 28.65) and laboratory-confirmed virus (RR=1.72, 95% CI 1.01 to 2.94) were significantly higher in the cloth masks group compared with the medical masks group. Penetration of cloth masks by particles was almost 97% and medical masks 44%.
The risk of infection was nearly four times higher in the control group than it was in the medical-mask group, but the difference was not statistically significant.* And the finding that infection was more common in the cloth-mask group than it was in the control group certainly does not seem to support recommendations or mandates that portray cloth masks as useful in preventing virus transmission.
But that is not the end of the story. As a "science brief" from the Centers for Disease Control and Prevention (CDC) notes, "the study had a number of limitations," including "the lack of a true control (no mask) group for comparison, limited source control as hospitalized patients and staff were not masked, unblinded study arm assignments potentially biasing self-reporting of illness, and the washing and re-use of cloth masks by users introducing the risk of infection from self-washing."
In March 2020, after the COVID-19 pandemic inspired new interest in the effectiveness of cloth masks, the lead author of the BMJ Open study, Australian infectious disease specialist Chandini R. MacIntyre, likewise noted that "some subjects in the control arm wore surgical masks, which could explain why cloth masks performed poorly compared to the control group." She added that "the cloth masks may have been worse in our study because they were not washed well enough—they may [have] become damp and contaminated."
MacIntyre later did a follow-up study, reported in BMJ Open last fall, that focused on a subset of the original sample: 607 hospital employees who worked in "high-risk wards" and wore "a two-layered cloth mask." She and her colleagues found that "the risk of infection was more than double among [health care workers] self-washing their masks compared with the hospital laundry." They also reported that "there was no significant difference in infection between [health care workers] who wore cloth masks washed in the hospital laundry" and health care workers who wore disposable medical masks.
"The majority of [health care workers] in the study reported hand-washing their mask themselves," MacIntyre et al. wrote. "This could explain the poor performance of two layered cloth masks, if the self-washing was inadequate. Cloth masks washed in the hospital laundry were as protective as medical masks. Both cloth and medical masks were contaminated, but only cloth masks were reused in the study, [underlining] the importance of daily washing of reusable cloth masks using proper method. A well-washed cloth mask can be as protective as a medical mask."
The Danish study that Paul cites, which was published in the Annals of Internal Medicine last March, was a randomized, controlled trial in which both groups of subjects received "encouragement to follow social distancing measures" but only one group was urged to "wear a mask when outside the home among other persons." Each of the subjects in the mask-recommendation group also received "a supply of 50 surgical masks and instructions for proper use." At follow-up, there was no statistically significant difference in positive COVID-19 antibody results between the two groups.
According to an editorial that accompanied the study, "The evidence excludes a large personal protective effect, weakly supports lesser degrees of protection, and cannot statistically exclude no effect." The study did not address the effectiveness of masks worn by carriers in protecting other people, and the editorial emphasized that the study "does not disprove the effectiveness of widespread mask wearing."
In addition to overlooking the limitations of the studies on which he relied, Paul did not address the substantial body of evidence cited by the CDC, which suggests that cloth masks do "work," at least to some extent.
That research, which includes laboratory experiments as well as observational studies, also has limitations. But instead of explaining why he is unpersuaded by the empirical case for wearing cloth masks when better alternatives are too expensive or hard to obtain, Paul simply cited a conveniently narrow slice of the relevant literature.
Furthermore, Paul did not merely say the effectiveness of cloth masks has not been proven to his satisfaction; he said research has conclusively demonstrated that they have "no value" at all. That gloss is reminiscent of the reckless judgment that then–Surgeon General Jerome Adams rendered early in the pandemic, when he tweeted that "masks" (apparently including surgical masks and N95 respirators) "are NOT effective in preventing [the] general public from catching #Coronavirus."
Although the CDC concludes that "experimental and epidemiological data support community masking to reduce the spread of SARS-CoV-2," it acknowledges that "further research is needed to expand the evidence base for the protective effect of cloth masks and in particular to identify the combinations of materials that maximize both their blocking and filtering effectiveness." Paul could have provided a useful alternative perspective if he had focused on the limits of the existing research and the gaps in our knowledge instead of dismissing the evidence without even considering it.
Although Paul presents himself as a defender of scientific standards, it does not seem like he is actually interested in having this conversation. In a tweet responding to his YouTube suspension, he called it "a badge of honor," saying "leftwing cretins at Youtube" were "banning me for 7 days" even though he cited "2 peer reviewed articles saying masks don't work."
YouTube's response to Paul's overstatements, of course, is hardly a model of rational discourse, which requires rebutting arguments by citing contrary evidence instead of treating them as too dangerous for people to consider. "Private companies have a right to ban me if they want to," Paul said, "but I think it is really anti–free speech [and] anti–progress of science, which involves skepticism and argumentation to arrive at the truth."
I am inclined to agree with Paul that social media companies, partly in response to inappropriate government pressure, are constraining online debate about COVID-19 issues in an arbitrary and unhealthy way. But this particular example would have been more compelling if Paul had been more careful in framing his case against the prevailing wisdom.
*CORRECTION: This post originally stated that the Vietnam study suggested medical masks were effective at reducing infection, but that result was not statistically significant.
Start your day with Reason. Get a daily brief of the most important stories and trends every weekday morning when you subscribe to Reason Roundup.