Moving from Face Masks to Transparent Face Shields?


Samantha Godwin, a fellow at Yale Law School, posted this to a discussion list, and I thought I'd pass it along (with her permission):

As universities try to work out if and how to reopen in the fall of 2020, many are working off the view that face coverings are essential to slow the transmission of the novel coronavirus and must be used if in-person classes and activities are resumed.

The default face coverings are the opaque cloth masks that have been widely adopted. These are, however, not the only option available to us, and may not be the option that is most conducive to delivering the core educational and academic functions of universities—including intellectually engaging conversations among faculty and students. These purposes may be much better served by the use of transparent face shields as an alternative rather than supplement for opaque face masks.

When considering best practices to adopt, we ought to  keep in mind that current practices are still largely experiments, both experiments in viral transmission management, and social experiments. We should therefore ask questions of alternative methods beyond 'has this been clinically proven more effective than standard face masks for reducing viral transmission?' although preliminarily there seems to be strong evidence that face shields are very effective at preventing transmission. This is not the only relevant question.

I would instead ask the question: what practices best enable us to live with this virus, which will likely be with us for years, without giving up too much of what we value in society and in our communities?

When put to this test, I think face shields are clearly superior to face masks for a number of reasons:

[1.] Interpersonal communication and empathy: When two people interact, each wearing a facemask, they can see only a small portion of the other's facial expressions. Each perceives the other as less empathetic because they are less able to emote or read the others emotions—a phenomena that has been pretty well demonstrated in pre-COVID-19 clinical studies. There also seems to be some evidence that people mostly read positive emotions from the bottom half of people's faces and negative emotions from the top half of people's faces so a society of face mask wearing is a stressed society.

This is not a problem for transparent face shields though—face shields allow for unobstructed display of emotions and non-verbal communication.

Also I've certainly noticed that it can be hard for people to understand me when I'm wearing a face mask, and hard for me to understand people wearing face masks: they seem to interfere with speech recognition in any non-ideal sound-quality situation, even for people with normal hearing. I suspect this would not be as much of an issue with face shields.

[2.] Compliance: I'm not sure about you, but when I walk down the street, or go to take out restaurants or grocery stores, I see lots of people wearing face masks pulled below their chin, overtly defeating their purpose, or at a minimum pulled below their nose, rendering them a symbolic but totally ineffective statement. The reason for this is obvious: face masks are tremendously uncomfortable for extended periods of time, especially when there's heat, humidity, pollen, or you wear glasses. Pretty much every time I glimpse the workplaces of 'essential workers' (like people in kitchens) I see employees not wearing facemasks despite the state order here to do so. They are just not sustainable and we need to plan for a world with real humans with imperfect compliance.

Face shields, on the other hand, are comparatively comfortable if well made: by avoiding direct contact with the wearer's face, they are 'wearable' for a much longer amount of time.

One apparent advantage of cloth face masks over transparent face shields is the ease with which they can be made at home. However, face shields are also much easier to manufacture at home than might be assumed, and can easily be made at home with readily obtainable materials, as seen in a number of youtube videos, including clear soda bottles as seen here.

[3.] Psychological benefits: The COVID-19 viral pandemic is accompanied by two other health pandemics: widespread depression and anxiety from social isolation, and widespread anxiety about catching the virus, even among people who are very low risk and who do not have much interaction with high risk people. Facemasks are understood (correctly) primarily as a means of protecting other people from the wearer, not the wearer from other people (though there is still probably a significant benefit).

Face shields, however, very clearly protect the wearer as well by covering the wearer's eyes and discouraging the wearer from touching their face—this may provide a more anxiety reducing effect for people worried about contracting COVID-19. The ability to smile and effectively express positive, mutual understanding through facial expressions can alleviate the stress of social isolation that many, especially students and other younger people who lack strong social networks, have been experiencing for months. Face shields beat out opaque face masks on both of these psychological hazards of the COVID-19 pandemic.

There is a possible downside to face shields: they look a bit silly whereas face masks can be fashionable. On balance, I think this can be overcome by explaining that face shields are probably more protective for the wearer and facilitate more natural communication and we all thought face masks looked silly a month ago too—but there might be another way to go:

Perhaps we could consider a third alternative that provides more coverage than face shields or face masks and maintains the social advantages of face shields: mosquito net hats. These translucent net hats almost certainly filter large droplets from the wearer and stop large droplets from others (short of perhaps being sneezed on from very close proximity) and they don't have the medicalized silly appearance of face shields (they also look a little silly but we're in the realm of the imperfect). These might be a viable third option, at least worth further study. Floating Doctors proposed mosquito net hats early on as a viable way to mitigate covid-19, albeit mostly as an easy way to avoid touching your face.

After half a semester of online learning, it is clear to many of us that while it might be better than nothing, it is no substitute for in-person, classroom education and meetings. Too many intangibles of face to face communication are lost online, leaving us with a muted, less rewarding, more exhausting mode of teaching and studying.

To get those intangibles back while continuing to practice risk mitigation, we need to think about exactly what kind of face to face interactions work best. On balance, if we decide that face covers are needed in the fall, we should strongly consider insisting on and normalizing transparent or translucent face covers that allow for a full view of each participant's face to make the most of our time in classrooms together.

Prof. Ed Richards (LSU Law) wrote this in response, specifically referring to the paper in the post to which Godwin linked:

The face shield they are working with is interesting. It has much more coverage than the usual face shield. It looks like a good approach. I would like to see more studies that confirm this study's finding about inhaled virus. Earlier tests by NIOSH did not find face masks acceptable PPE for inhaled particles, but they did not find surgical masks acceptable either. That is where the N95 standard and mask came from.

Masks are problematic, no question. Having used a face shield in other, more pleasant circumstances, they do have their own problems. If you are outdoors, they can get hot and uncomfortable, and you cannot put them in your pocket to use when need to. They look much better for indoors. They fog when conditions are right, and the better they fit, as described in the article, the more they fog. That might prevent their use in certain environments, which probably includes some of my classrooms. But outside of fogging, they have real benefits over masks.

The authors note that no studies have yet been conducted to see how well face shields help keep exhaled or coughed virus from spreading outwards from an infected wearer. I would like to see that study done, otherwise adopting face shields turns the entire rational on its head. You predominately wear surgical style masks to protect others. This would imply that everyone would have to shift to face masks, unless studies show that they are equally good at preventing the spread of virus. We are already ignoring this problem by mixing N95 style masks, those with a valve that allows expired air to bypass the mask filter,  with surgical masks. Surgical style masks are designed to protect others, while ventilated masks protect only the user. We are also completely ignoring beards, which makes masks nearly useless for any purpose.

NEXT: Anonymous Reactions to Bostock

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. The benefit of masks is that they render facial recognition less effective.

  2. I haven’t seen face shields for sale so mass availability would be my main concern.

    1. Any hardware store has them – usually multiple versions. See this from Ace Hardware for a single example.

  3. The general answer is no, face shields do not replace face masks.

    A face mask is designed to prevent you transmitting the virus to others, by catching wet droplets you may expel via coughing, sneezing, or even breathing heavily. There is limited protection for breathing in virus particulates. Transmission: You -> others primarily.

    A face shield is designed to stop fluid spray and/or droplets from hitting you in the eyes, nose and mouth. It has an extremely large opening on the bottom which aerosolized particles can be either breathed in, or breathed out. High impact sprays/droplets are prevented. Generally used during procedures which would generate these high speed droplets. For example, dentistry. Transmission: others -> you primarily.

    There are a number of cases within hospitals where both masks and shields were simultaneously used. But shields really don’t replace masks, and provide extremely limited protection in common usage.

    1. Shields are more uncomfortable imo, but I don’t buy your argument that they are less functional in reducing transmission by the wearer.

      It has an extremely large opening on the bottom which aerosolized particles can be either breathed in, or breathed out.

      I’m pretty sure the direction those particles are going is not towards other people.

      The rest of what you wrote is anecdotal or irrelevant.

        1. Our results showed that 0.9% of the initial burst of aerosol from a cough can be inhaled by a worker 46 cm (18 inches) from the patient. During testing of an influenza-laden cough aerosol with a volume median diameter (VMD) of 8.5 μm, wearing a face shield reduced the inhalational exposure of the worker by 96% in the period immediately after a cough. The face shield also reduced the surface contamination of a respirator by 97%.

          18 inches is a different range than anyone is talking about with masks.

        1. Makes no mention of shields.

      1. Trust science, not speculation.

        Face shields do not prevent aerosolization of the virus.

        1. Your initial post was speculation, not science.

          Masks don’t prevent aerosolization of the virus either.

          I trust the heck out of science. But I also know the limits of individual studues – and the limits I have, as a reader of scientific studies.

          1. “Your initial post was speculation, not science.”

            So was your response. But none of this changes the face that the recommendation is to wear masks, and face shields are not masks.

            1. Poking holes in someone else’s speculation is not itself speculation.

              Are you talking about the CDC recommendation, or something else?

  4. You were asking about us living our lives, so masks, mosquito netting, etc make sense.

    But for teaching in a classroom; why not have–for the teacher–a curved piece of glass, so (1) there is a barrier in front of the teacher, while still allowing his/her voice to be clearly heard, and (2) if it’s approx 90 degrees, it would cover the range of students, from left to right, in pretty much every classroom?

    Students would still need their face shields–to protect themselves from each other in class, in the hallways, etc. And professors would need them, to move to/from classrooms and offices.
    But I saw how quickly Target, Trader Joe’s, et al installed glass or plexiglass barriers, in order to protect cashiers. Would be a trivial matter for law schools to do as well, to protect their instructors.

    1. Why don’t we put the teachers and all the students in human scale hamster balls?

      1. I’d pay good money to watch that.

  5. AYFKM?

    Face shields are used to protect individuals from strong chemicals and the burns they may receive from them.

    Since when do people think that life should be without risks? The FACT is that you have a greater chance of death from driving to/from work or class than you do from this Wuhan Flu.

    You people need to get a grip.

    1. They need to get far more than just a grip.

    2. “The FACT is that you have a greater chance of death from driving to/from work or class than you do from this Wuhan Flu.”

      This is not true. Not even close. And I suspect you know that. We have something over 110,000 deaths from COVID so far this year. Deaths from motor vehicle accidents for 2018 were under 40,000. (from WISQARS).

      1. Look at probabilities. Industry probabilities of death at work (all causes) is roughly 1e-6; auto accidents are roughly 1e-5 (data from my HAZOP/LOPA reference).

        Until we cross the threshold of 3.3 million deaths in the US, you have a greater probability of death from accidents at work. Once we cross the threshold of 330K deaths in the US, you have a greater probability of death from auto accidents.

        1. “Look at probabilities. Industry probabilities of death at work (all causes..”

          This is not what you wrote in your original comment.

          1. Just for the record, as far as certain societal elements blowing the COVID thing out of proportion and shutting down a huge portion of the economy, well, I tend to agree with you.

            1. What irrefutable, unassailable evidence is there to support the proposition that there have been 100,000 + deaths due to the Wuflu?

              1. Your evidentiary demands rather indicate an expected outcome, due to ‘irrefutable, unassailable’ being fundamentally impossible for public health data for any population of size.

      2. I’m not sure that’s a valid comparison, Albert. Auto deaths are a continuous risk. COVID-19 (or any individual infectious disease) is a one-time risk. You would do better comparing lifetime risks of auto-accident-deaths to lifetime COVID-19 risk.

        According to, your lifetime risk of dying in a motor vehicle crash are 1 in 114 or 0.88%. The current worldometers data suggests that the total COVID-19 death risk is somewhere on the order or 0.02 – 0.03%

        1. First, you appear to be determining risk of death posed by Covid by dividing the number of worldwide Covid deaths by the population of the world, which is a prosperous way to assess risk. Second, there’s good reason to believe that there’ll be a vaccine for Covid within 2 so the more apt flawed comparison would risk of death in a car crash over two years.

        2. “I’m not sure that’s a valid comparison, Albert. Auto deaths are a continuous risk. COVID-19 (or any individual infectious disease) is a one-time risk. You would do better comparing lifetime risks of auto-accident-deaths to lifetime COVID-19 risk.”

          I agree, but again, that is not what the OP wrote. The deaths from seasonal flu over the last several years average somewhere around 25 to 30 thousand per year, with a low of under 10,000 to a high of about 40,000.

      3. Yes, but a large majority of the deaths from Covid-19 are among the elderly. Prisoners are also a major factor. In fact, I read that the majority of Covid-19 deaths have been among residents of nursing homes, jails and prisons. So people of working age who are not prisoners (and college students) are at less risk than the raw stastistics would suggest. But there is still a risk of getting it, even if it is not as great as many believe. And anyway, as long as people believe the risk is significant, they will insist on some protection from it. The face shields and mosquito netting are interesting ideas. How well can facial expressions be read through mosquito netting?

  6. Regarding the efficacy of masks:
    We now have a long enough incubation period passed to look for spikes in new covid cases. I don’t see any evidence of a spike of new cases in the overall US data or for the statewide data from NY, CA, MA, and IL.
    As for the masks protecting the user, it is highly unlikely that they do given the way that most people wear them, pushing them down and puling then back up numerous times, almost certainly contaminating both the inner surface and out surfaces. The outer surface does have some kill of virus due to ambient UV. Hence the pracices of people in demonstartions should be looked at in films more closely and compared with the absence of any spikes in the contagion data.

    1. Why would you expect spikes in NY and MA, both of which have high rates of mask usage?

  7. re: “strong evidence that face shields are very effective at preventing transmission”

    Sorry, but no. The study you link to as “strong evidence” was an unvalidated computer simulation. Computer models are not data and can not, by themselves, be evidence of anything. Computer models are mathematical restatements of your hypothesis and assumptions. Computer models can tell you if your hypothesis is internally consistent but only after the model results are compared to empirical observations can you say that you have evidence about your hypothesis one way or the other.

    Face sheilds may be great at preventing transmission of coronavirus. The authors of that study make a plausible series of arguments that they might be. But that study doesn’t demonstrate that they actually are great.

  8. Donning a diaper mouth is dangerous to one’s health.

    Russell Blaylock wrote that there are several dangers presented with wearing face masks. According to Mr. Blaylock, “[s]everal studies have indeed found significant with wearing such a mask. This can vary from headaches, to increased airway resistance, carbon dioxide accumulation, to hypoxia, all the way to serious life-threatening complications.”, May 11, 2020. Blaylock noted that while we scold others in furtherance of our virus virtue signaling, “no one is telling the frail elderly and those with lung diseases, such as COPD, emphysema or pulmonary fibrosis, of these dangers when wearing a facial mask of any kind – which can cause a severe worsening of lung function.”

    1. Blaylock has called the American medical system ‘collectivist’ and has suggested that health-care reform efforts under President Obama were masterminded by extragovernmental groups that wish to impose euthanasia

      Not a great appeal to authority to go with.

      1. What a flabby logical fallacy of a dodge.

    2. Russell Blaylock, the birther from Newsmax?

    3. Increased resistance is an inherent property of any filter. If masks didn’t cause increased airway resistance, they’d serve no purpose.

    4. I’d have to agree with you, Libertymike, that surgical type face masks aren’t good for one’s own health over any period of time. They also dehumanize the face by covering the mouth and nose, and give people the appearance of the Iwazaru monkey who must speak no evil, however one interprets that.

      We’ve had the Silent Generation, and now we’re to be the Muffled one. Volokh’s post suggests a better alternative. As for me, I might wear a niqab veil or apiary hat…

      1. (But, I don’t believe in the PandemicTM.)

  9. Samantha Godwin, along with almost every governor in the nation, is trying to compromise with the virus. What a shame for all of them—and for everyone else—that the virus never compromises back.

  10. Why should it be the case, as reportedly it is, that ordinary face masks, are protective of others, not of the wearer? Seems counter-intuitive. Is it that a significant reduction in viral particles that are exhaled and make it any distance from the wearer’s face will serve to protect others, though probably 100%; but even small inoculum of virus particles inhaled by wearer through mask may be enough to infect?

    @Sarcastr0. I disagree with you about Dr. Blaylock as an authority. A look at his CV leaves me with no doubt that he is clearly a serious, hardcore, and thus dangerous quack, and as such not to be believed as to the most basic facts of anything without convincing corroboration by credible authorities and sources. In a perverse sense, he may be seen as an “authority,” so long as the scare quotes are always there. One must wonder why anyone would see him as trustworthy, and what sources let them to him.

  11. As someone who at work at times has to wear both and for projects at home sometimes wears a face shield without a mask: Masks are far more comfortable than face shields to wear for more than a very few minutes.

    Also- face shields protect from the initial spray of droplets out of a person. They do nothing against inhaling virus that is floating in the air.

    There is excellent evidence that masks reduce the risk of infection.
    I don’t know of any studies that show face shields alone protect against infection.
    When healthcare workers are doing things that cause extensive spray of virus-laden fluids they wear face shields with masks underneath. They never wear a face shield without a mask.

    Wearing a face shield over a mask would be helpful but only if you were going to be close enough to other people for the spray to hit you in the face. Since everyone is encouraged to keep their distance, you should not be getting close enough for a face shield to do any good. If you are forced to be in a crowd, then a face shield over a mask could be better than a mask alone. To be effective the face shield would need to be kept in place. In crowds the jostling can easily displace the face shield, making it ineffective.

    Seeing other people’s faces? Over rated. In my virtual meetings I and most over people resort to video only when there is something to show. We have no need to see the expressions on people’s faces. What difference does it make?
    This is a written site. would it be better, or even 1/10th as good, if every article was accompanied by pictures of the authors faces while they were typing?

  12. Leon Russell via George Benson;

    No matter how hard I try
    To understand the reasons that we carry on this way
    We’re lost in a masquerade. 

Please to post comments