Getting Infected

I, like pretty much everyone else, have been spending a lot of time thinking about the ongoing COVID-19 crisis. My particular line of thought tends to drift along two tracks: what the policy response to the crisis should be and what our moral obligation as individuals is.

I’ve come to the conclusion that intentional self-infection, ideally though not necessarily as part of a human-challenge trial, has a lot of strong arguments for it both from a general policy perspective as well as from a moral perspective.

Since many folks respond to this idea with something ranging from disgust to bafflement1My thinking has been shaped by other people who have been talking about this for a while. Robin Hanson ( in particular as well as many others brought the idea to my intention and have shaped my thinking on this. I wanted to explain a bit more about my reasoning and why I think this is an important policy avenue to pursue and why individuals who choose to self-infect (e.g., by going to so-called “COVID parties”) could be acting entirely rationally and morally.


I’m working off a few assumptions that I find reasonable though they definitely aren’t universally accepted. If these assumptions are wrong, then the conclusions I’m drawing will surely be sub-optimal.

  • A widely-available vaccine is at least 18 months away, and likely more.
    • Additionally, there’s a substantial non-zero (i.e., > 1%) chance that we never get a vaccine
  • We will not develop a therapy that meaningfully reduces death or hospitalization rates in the next 18 months
    • This might be overly pessimistic. If you have a good reason to be more optimistic about this, I’d love to hear why.
  • Catching and recovering from the virus confers a meaningful amount of immunity for at least 18 months
  • The US doesn’t have the practical ability to administer a workable test-and-trace policy of sufficient scope. Period.
  • The US population’s tolerance for extended quarantine has already been reached and our days of general maximum-adherence to quarantine measures (absent military intervention) are already behind us.

These assumptions apply specifically to the situation in the USA — there are other countries that have the benefits of being islands, have a generally higher trust in government and willingness to follow government directives, or have governments capable of administering large-scale test-and-trace programs. Those countries have other options.


Assuming the above is true, we’re on the path of having the virus eventually infecting some large part of the population in the US — sufficient to reach so-called “herd immunity”. There is no practical way to prevent that happening — at best we have some control over how long it takes to reach herd immunity and how overwhelmed the healthcare system gets in the meantime.

Conditional on that being true, then what we want to optimize for is:

  • reaching herd-immunity without “over-shooting“ and infecting more people than needed
  • minimizing the number of deaths by infecting those most likely to survive first
  • optimizing for partial-herd-immunity earlier by providing immunity to those most likely to spread to others earlier (i.e., people who have more contacts on average)

So on that path, the advantage of self-infection over just waiting to get accidentally infected from chance contact are numerous:

  • Assuming you’re young and healthy you can contribute to reaching partial herd-immunity faster by self-infecting.
    • This reduces the risk to any other people at higher risk than you who you might come into contact with in the future
  • If you self-infect you can intentionally self-quarantine for two weeks and avoid spreading the virus to higher-risk individuals
    • Since it seems that most transmission occurs from non-symptomatic carriers, this is a real meaningful benefit. With this certainty you can compress your quarantine into two weeks rather than having “light quarantine” over months.
  • You can time your infection to a time when the healthcare system is not overwhelmed, so in the worst-case scenario, you won’t be robbing a hospital bed from someone else who needs it.
  • It reduces the need for quarantine in the future thereby allowing the ability to return to work, be the “dedicated grocery shopper in the household”, etc.
  • It reduces the psychic burden and uncertainty on an individual — there’s a strong benefit of “just getting it over with” that I personally find very compelling.

I find these reasons incredibly compelling. Making the decision to self-infect (assuming you’re young and healthy enough to believe your chance of death is very low) is rational on a personal-interest basis (reducing future uncertainty and psychic burden) as well as on a moral/public interest basis (by moving us toward herd immunity and reducing the likelihood of accidental infection of someone at high-risk).

Even better, if self-infection comes as part of a human-trial challenge trial (i.e., deliberately exposing people as part of a clinical trial to test vaccine effectiveness) we get all of the benefits listed above plus speeding the development of a vaccine. This seems like such a slam-dunk to me from a policy and moral perspective that I’m surprised there’s even a debate about it.


If my reasoning is mistaken here or my assumptions are off, I’d love to hear more about either (please email me). If not, I’d really like to push this conversation forward and into the public domain and get people talking about policy paths that don’t rely on the US government suddenly becoming effective.

In the meantime, if you’d be interested in self-infecting, it’d be best if you did it as a part of a human-challenge trial. You should read more about that (and consider signing up as a potential candidate) here:

About the author

By michael