COVID-19 Herd Immunity

(Last Updated On: February 18, 2021)

A Faster Path to Normal Life?

2/18/2021 – For an update on COVID-19 Herd Immunity, see also this later blog:

COVID-19 Herd Immunity, in which so many people have antibodies that the virus can no longer spread, is where we want to get, to return to a normal life. But just letting the virus run loose is a dangerous and deadly way to get there. New research offers a safer, quicker path, through weaker virus infections.

There’s actually science behind this new idea. When I show you the research, you may (like me) slap the side of your head and say, Of Course! And Why Didn’t I Think Of That? In any case, it’s something that we all deserve to know.

Here are the major sections yet to come:

About Herd Immunity

An Animal Model for COVID-19 Infection

Masks Significantly Weaken Infections (in Hamsters)

Distance & Hygiene Significantly Weaken Infections (in Soldiers)

Weak COVID-19 Infections as a Path to Herd Immunity



About Herd Immunity

covid-19 herd immunity

Click image to see full description of “herd immunity”

We have been reading about the potential for “herd immunity” to reduce the COVID-19 virus to a non-threat. Some US Government officials have been quoted (perhaps inaccurately) as advocating letting the virus spread at its own rate, so that we reach herd immunity more quickly.

If you want to understand herd immunity, Wikipedia has a clear and accurate discussion. And if you click the drawing above, it will show you a great explanatory graphic.

Here’s what we need to know: herd immunity occurs when a large enough percentage of the population is immune to a disease. They might be immune because they have already caught the disease and recovered, retaining antibodies that resist re-infection. Or they might be immune because they received an effective vaccine. Or perhaps, they are simply naturally immune.

How does herd immunity work? Well, any disease can only keep spreading so long as it infects more people than recover from it. If only a small fraction of the population is vulnerable, the disease will die out rather than continue to spread. Basically, the virus or bacteria can’t find an infectible person before it dies.

How does the virus die? Either its host (the infected person) has a strong immune system that kills the virus. Or, the virus kills the host and then dies because it no longer has a host in which to live.

            How Herd Immunity Relates to Infectiousness

Now think about this. If each infected person can infect a huge number of people, then it’s easy for a virus to find a new victim. The number of people who are infectible is called the reproductive number R0, pronounced R nought or R zero. And the fraction of people who have to become immune for herd immunity (the Herd Immunity Threshold, HIT) to protect everyone is approximated by a simple formula:

HIT = 1 – (1 / R0)

You don’t need to dwell on this formula. Just think about these two examples:

  • Measles is very infectious. One infected person can infect 12 to 18 other people, so the reproductive number of measles is approximately 12 to 18. This implies a herd immunity threshold HIT of 92 to 95%, an immense fraction of the population!
  • COVID-19 is more infectious than influenza, but much less so than other diseases. For COVID-19, R0 is approximately 2.5 to 4. This implies a herd immunity threshold of 60 to 75%.

Bottom line: If 60% of the population can become immune to coronavirus, either because they are naturally resistant, or from a vaccine, or from surviving an infection, the virus will stagnate or die out. Some believe that even 50% immunity or less would substantially slow down the progress of the disease, as we will discuss below.

            But Nothing Is Easy…

However, this is science, and science comes with caveats. The simple equation above assumes “that everyone is equally susceptible to infection and has the same chance of bumping into every other person, like molecules of gas in a bag.”

But that’s baloney, of course. In the real world, people interact only with certain clusters of other people. So the population could have a total immunity of, say, 75%, while still having very vulnerable groups of people here and there in the population. Therefore, we have to embrace the idea of herd immunity with a large grain of salt.

In addition, some experts think that people may already have some level of resistance to COVID-19, likely because they have caught and recovered from flu and other similar viruses. This natural resistance could count toward reaching overall herd immunity. A group of researchers in UK, Brazil, Portugal and US have published articles (Aquas, Gomes, Lorenco) suggesting that taking this into account, herd immunity thresholds might be 20% or below for England, Belgium, Portugal and Spain. However, the analysis relies on fitting a complex model to the reported statistics of COVID-19 in each country, after the fact, and is not widely accepted by other experts.

            Don’t Get Too Optimistic, Too Soon!

Could we get herd immunity just in localized areas? Well, if in fact it takes 50%, the most heavily infected places are still short of that number infected. The cities in Ecuador and Brazil with the most infections barely top 40%.

What about the more optimistic numbers suggested by the UK workers, of 10% or 20% as a herd immunity threshold? Well, in the US, New York City estimates that 19.9% have been infected, Boston 9.9%, and other cities much lower.

If herd immunity were kicking in at 10% or 20%, we should see active cases of COVID-19 dying out in these cities. However, New York City is showing a steady rate of infections and so is Boston. This information suggests that wherever the herd immunity threshold may be in Europe, it’s above 20% in the US.

            The Death Rate for Coronavirus

Is it practical to get a large percentage of the US population infected with coronavirus, so as to be confident of having herd immunity? Well, let’s discuss that.

On September 20 (“day 200” in the Johns Hopkins data, summarized on, we reached 199,509 coronavirus deaths, out of a total of 6,810,879 confirmed cases. Thus our best data indicates that death claimed 2.9% of the people who got infected.

Are things getting better, as we learn to deal with the disease? Yes, they are. Let’s look at a date when we had one-half that many confirmed cases. That date turns out to be July 14 (“day 132”), and as of that date there were 137,066 deaths. Therefore, as of July 14, 4% of the people who had contracted coronavirus had died. However, the most recent 3.4 million cases resulted in 199,509 minus 137,066 equals 62,443 deaths, a rate of 1.83%. So we are more than twice as successful in treating coronavirus cases as we were earlier this year, or perhaps infections are becoming milder as we will discuss below.

            Herd Immunity for COVID-19: Brute Force

The discussion above implies that the effective herd immunity threshold is somewhere above 20% in the US. So let’s be optimistic and suppose that we aim at 25% immunity in the US. 25% of 330 million is 82.5 million people who would need to become immune, which is about 75 million more than the people who have already caught the virus.

What are our choices for achieving this level of immunity in the population?

  • That many additional people might catch the virus. If 1.83% of them die, that would be 1,372,500 additional deaths in addition to the 200,000 who have already died. So we would have to kill more than a million additional people. Some people think that would be smart public policy, but that doesn’t include anyone whom I know!
  • We could have an effective vaccine, make it generally available, and hope that at least 25% of our fellow citizens accept the shots. However, we don’t yet have a vaccine, and health experts believe that even after we have a proven vaccine, it will take a number of months to inoculate everyone who is willing.

Neither of these approaches seems destined to save many lives, very soon.

            Herd Immunity for COVID-19: Getting Smarter

But there’s another answer. During the last few weeks health scientists have proposed a better choice: for these millions of people to catch a very mild infection. Not strong enough to show symptoms, certainly not enough to put anyone in the hospital. But enough of an infection to stimulate their immune systems to develop antibodies that will make them immune to further infection.

With your permission, I will describe the march of science that has helped this solution make sense to some very smart people. If you don’t care for the gory details, feel free to skip to the Conclusions section.

An Animal Model for COVID-19 Infection

The easiest way to tell this story is chronologically. That gives due credit to those who had the earliest ideas. And, it shows how each new idea made sense in light of previous work by others.

            Animal Testing to Save Humans

Some readers may feel uneasy about running tests on animals, even for the laudable purpose of saving human lives. However, the biological sciences are highly dependent on what are called “animal models” to learn to prevent and treat human diseases.

It’s considered unethical to use humans to try untested vaccines that might harm or kill them. But scientists have come to accept that properly regulated animal testing plus institutional ethics rules provide an acceptable framework for tackling human risks such as COVID-19.

Will we always rely on animal testing? In the future, it’s slightly possible that alternatives may reduce or even eliminate the use of animals:

  • A comprehensive computer model of the human organism and of the threats that challenge it. Thus simulation might replace physical testing.
  • Synthetic organelles are being studied to simulate natural processes. One application might be to create livers and other replacement organs for humans. At some point scientists might create a sufficiently complex biosimulation that could allow them to study diseases outside the body in an inanimate structure.

However, this is pretty far off. For the foreseeable future, researchers need animal models to help them tackle human diseases.

            The Syrian Hamster as a “Golden” Solution

There has already been some limited coronavirus testing on animals whose response seems close to humans. Rhesus macaques, a type of monkey, seem to respond to COVID-19 very similarly to humans. However, expertise and suitable biosafe facillities are scarce. In addition, there are many reasons to prefer an animal model that is small, available and not a primate. (Monkeys seem so humanlike that people would rather not use them for experiments.)

On March 26, Chan et al in Hong Kong reported that the Syrian golden hamster seemed like an excellent animal model to study coronavirus. Additional confirmation came from a May 15 article by Imai et al in Japan.

            Face Masks on Hamsters

Meanwhile, Chan et al followed up on May 12 with a study of how face masks changed disease transmission among the hamsters. I know what you’re thinking: how did they get those active little animals to wear face masks, anyway?

The answer is: very cleverly. Because it’s not exactly a face mask. It’s an air filter made of the same material as surgical masks.

The researchers put two hamster cages next to each other in an isolation chamber. In the left cage was a single hamster whom they had infected with a measured amount of coronavirus. In the right cage were three uninfected (“naive”) hamsters. An air flow passed from left to right, so exhalations from the infected animal would pass by the uninfected ones.

Masks Significantly Weaken Infections (in Hamsters)

They first tested five pairs of cages, containing 15 uninfected hamsters. Without any filtering, 10 of the 15 naive animals became infected, that is, 66.7%. Not too surprising.

Next, they simulated the case of an infected person wearing a mask. They set up four pairs of cages, housing of course an additional 16 hamsters. Each pair of cages was separated by a sheet of mask material with the blue exterior side (which repels fluids) facing the uninfected hamsters. Only 2 of 12 naive hamsters were infected, 16.7%.

In the third experiment, there were four pairs of cages with an additional 16 hamsters. In this case, the mask material had the blue side toward the infected hamster. The experimental setup simulated the case of an uninfected person wearing a mask. In this experiment, 4 of 12 hamsters were infected (33.3%), a number that was below statistical significance.

The total of the two masked experiments showed 6 infections in 24 hamsters, or 25%. So the masks measurably reduced infection.

However, here’s a very important result: Every hamster infected through the mask had very substantially less viral load in nose, trachea and lungs compared with those infected with no mask filtering. When I say “substantially,” I mean hundreds to thousands of times less! (See Figures 4 and 5 in the article.) In other words, the masked hamsters acquired a much weaker COVID-19 infection.

Distance & Hygiene Significantly Weaken Infections (in Soldiers)

Around the same time (June 4), Bielecki et al in Switzerland reported a study of two groups of soldiers. Nine days after a coronavirus case appeared in one group, hygiene measures were introduced across both groups: social distancing; mask wearing in situations where distancing was not possible; and twice-daily cleaning and disinfecting.

The first group had 354 soldiers, of whom 30% became ill from coronavirus. Apparently this amount of spread occurred during the first nine days, before hygiene rules were imposed. The second group of 154 soldiers had no symptomatic cases at all.

Does this mean that the second group just got lucky and was never exposed? Not at all. The soldiers in this group had viral RNA in their noses, and virus antibodies in their blood. That is, there were mild virus infections that never showed visible symptoms.

            Conclusions from the Swiss Soldier Study

This experiment is only a one-off, and involved an atypical population: young (age 18 to 28 years with a median of 20.4), healthy, active men. Nevertheless, what do these results suggest?

  • That once people are exposed to infection for a few days, they may acquire a full-blown symptomatic case of COVID-19.
  • However, things are different if they follow distancing or masking, plus hygiene, before they are infected. In this case, when they get infected the disease is mild, so mild that they may never know they have it.

One of the authors contributed this additional suggestive comment in Reddit:

One of the first observations that triggered us commencing this study was that when moving patients from single isolation to cohort isolation we noticed their symptoms worsening again! So the amount of “initial virus dose” and “additional” virus dose once you have contracted it seems to matter.

In other words, the severity of the disease depends both on the initial dose of virus that you receive, and any additional doses that you get after you are already infected.

Weak COVID-19 Infections as a Path to Herd Immunity

At this point you may see where we are headed. There seem to be “safe” ways to catch coronavirus and build up antibodies that can give us some immunity.

Several articles build on this idea and take us to its logical conclusion:

On July 17, Van Damme et al in Belgium, India and Australia reported that wearing face masks has two effects: (1) When you catch COVID-19 you get a milder case; and (2) When you have a milder case, you don’t spread the virus to other people as much.

On July 31, Gandhi et al in San Francisco and Baltimore reported that face masks not only reduce the likelihood of spreading coronavirus; they also reduce its severity and its death rate. They quoted examples of two food processing plants (seafood in Oregon, chicken in Arkansas). The companies issued every worker a mask each day and required them to wear it. Of the workers who became infected with coronavirus, 95% were asymptomatic – that is, their infections were very mild.

            “Variolation” as a Path to Herd Immunity

On September 3, Gandhi and Rutherford in San Francisco pulled together the evidence with a specific proposal.

They start by mentioning one of the hamster studies, plus the food plants. They then discuss coronavirus infection on cruise ships. Usually only 20% of such cases are mild, that is, asymptomatic. However, an Argentinian ship that issued masks to both crew and passengers found that 81% of virus cases were mild. The masks greatly attenuated the disease and made matters much safer for those infected.

The authors describe “variolation,” which was a process used to protect against smallpox before the smallpox vaccine was available. It consisted of giving people a mild infection that would confer immunity. The authors propose that we should, similarly, take actions that weaken the effect of coronavirus.

                COVID-19 Weak Infections Get Us Closer to Herd Immunity

Gandhi and Rutherford certainly favor having an effective COVID-19 vaccine as soon as possible. However, they point out that while we wait for the vaccine to exist and to become effective, there’s a relatively safe way to increase population immunity. And that way is to take steps that insure that when coronavirus spreads, that it is a weaker disease and not as fatal.

Those steps seem clear from the hamsters, the soldiers, the food plants and the cruise ship. The very steps that reduce the spread of coronavirus also insure that when it spreads, leaking through our defenses, it is much milder and therefore less deadly. One of the Reddit commenters (thiedenkapp) describes the situation like this:

If you start with less [virus] and propagate at the same rate it will take longer to get to the same number. This extended disease timeframe could give the adaptive immune system (antibody producing B-cells and cytotoxic T-cells) just enough time to gain its footing since it takes usually 7 – 10 days before your body mounts a full response to a novel pathogen.


Thus, a weak coronavirus infection primes your body’s defenses and gives them time to protect you. And you make sure the infection is weak by taking steps to reduce the number of viruses you receive. A gigantic input of virus can overwhelm you, leading to a serious and possibly deadly infection. However, if the virus has to get past your defenses (mask, distance, hygiene) you will receive much less of it and will be better able to resist it.

Thus you can catch, survive and recover from coronavirus. We might call this “COVID-19 Mask Immunity” because masking is an important ingredient in the equation. And if enough people acquire these weak infections, we will be much closer to “herd immunity” even before we have a good vaccine to help us all become immune.

            Advice to Take, and to Not Take

Does that mean we should all rush out and try to capture a weak infection? No, that’s a risky approach. You would be running an uncontrolled experiment with yourself as the guinea pig. Because you don’t know how vulnerable you are, or how much virus would be safe rather than dangerous, and you don’t have a way to measure how much you are receiving even if you knew the safe level.

However, this information does give us reassurance: that by following precautions, we can make it much more likely that if and when we are infected, the infection will be mild and nonfatal.

Do your local health and government officials tell you to follow precautions, or not? That is IRRELEVANT. What is important is that it’s a smart idea to wear a mask, and keep some distance from people you don’t know. Because that improves your chances that when the virus finds you – which it eventually will do – it will arrive weakly enough that you can vanquish it, with or without medical assistance.

As we can see, COVID-19 Herd Immunity is not an unreachable goal. An effective vaccine will help us to that goal. And between now and when the vaccine arrives, we can help ourselves, and also advance the cause of immunity, by simple actions: masks, distance, hygiene, minimizing close contact with strangers and, of course, clean air.

Image Credits:
– Herd immunity figure, as is and cropped, from TKarcher under Creative Commons License
– Syrian golden hamster cropped and color-adjusted from Adamjennison111 at English Wikipedia under Creative Commons License
– Cows by flickr on
– Soldier in a flag mask & wearadamnmask by j4p4n on
– Cow by MabelAmber on pixabay


COVID-19 Herd Immunity — 8 Comments

  1. Art,
    Excellent analysis of complex medical experiments and results.
    Your conclusions and recommendations make complete sense if only we can get enough of the US population to abide by them.

    Historically, we are not a people who readily adopts restrictions to our behavior.

    Stay safe! Paul

    • Thanks, Paul! What you say is true. To me, the most positive result is that regardless of what other people do, our own actions greatly reduce our personal vulnerability. When we see photos of large numbers of people mingling, those few people who wear masks are not just setting a good example, they are also reducing their own risk. – Art

  2. Thanks Art ! I’m not a scientist… just somewhat curious & logical. I very much appreciate your excellent, informative presentations (especially this one), that use a minimum of advanced math. We everyday people need the good unbiased and simple to understand information that you provide. Hope you and yours stay healthy.

    • Thanks for your appreciation, Mike! Unfortunately some folks make claims using selective data, claiming that they are being scientific. When someone presents only one side of an argument it raises my suspicion that they made up their mind before they looked at the facts. I try to minimize my own built-in bias (which I suspect that everyone has) by presenting the limitations as well as the results. Faith and morality may offer absolute rights and wrongs, but science doesn’t generally claim an extreme position. Good health to you too! – Art

  3. I absolutely do not disagree. We wear masks and advocate others do, and use other precautions. ‘Viral load’ is another term that has been used in the media that effectively supports the (wonderful) analysis you’ve done. Anyway, no argument from me, just a piggyback concern is all. Thank you!

    • Cheryl, good for you, for making sure you and yours stay safe! Viral load is a good term because it describes something that can be measured. How an individual responds to a given viral load may vary considerably, but obviously less is better in any case. – Art

    • Cheryl, I agree that the data of infected percentages will certainly change. However, I believe that my conclusions will still be valid: that a person who takes precautions, regardless of what other people do, is less likely to get an infection; and if they get one, it is likely to be milder and less spreadable to other people. – Art