A January 1, 2021 article by Zeynep Tufekci reported in The Atlantic talks of a new variant of SARS-CoV-2 which causes COVID-19.

It states:

A more transmissible variant of COVID-19 is a potential catastrophe in and of itself. If anything, given the stage in the pandemic we are at, a more transmissible variant is in some ways much more dangerous than a more severe variant. That’s because higher transmissibility subjects us to a more contagious virus spreading with exponential growth, whereas the risk from increased severity would have increased in a linear manner, affecting only those infected.

To support this, it cites a Dec 29, 20202 Twitter thread by Adam Kucharski (@AdamJKucharski), a professor at the London School of Hygiene & Tropical Medicine:

Why a SARS-CoV-2 variant that's 50% more transmissible would in general be a much bigger problem than a variant that's 50% more deadly.


The above is just an illustrative example, but the key message: an increase in something that grows exponentially (i.e. transmission) can have far more effect than the same proportional increase in something that just scales an outcome (i.e. severity).

However this is an unreferenced claim and is not in a peer reviewed medium.

Is this claim correct?

  • I'm wondering how we're supposed to use empirical science to answer the hypothetical question in the title. Then in the body we see a mor general claim we probably can answer ("a mutation with increased transmissibility leads to massively more deaths than a mutation with increased fatality"). I think we should edit this to be about that claim, not the hypothetical title question. – fredsbend Jan 9 at 3:44
  • 8
    This is going to be difficult to answer, because Prof Kucharski is already explaining a basic exponential model to a lay audience. (Hint to answerers: Simply repeating that explanation using your own back-of-the-envelope calculations would be repeating the claim, not answering the question.) What could an answer do? Show that infectious diseases do indeed follow approximately a exponential model at the beginning of an epidemic? – Oddthinking Jan 9 at 4:50
  • 1
    The conclusion obviously depends on some modelling assumptions. Are you disputing their (fairly trivial) math or their assumptions? The hypothetical alternative in which a Covid-19 variant had 100% presymptomatic transmission but then people suddenly died with 100% probability after two weeks of no symptoms, it could be worse, but it might not be realistic. Kucharski is only considering an alternative in which the CFR increases by 50% which gets it from 0.8% to 1.2%, in absolute terms. – Fizz Jan 9 at 10:39
  • 7
    I don't think this is a good Skeptics.SE question. It's explaining a basic mathematical principle, not making a verifiable claim. The way to understand its veracity is to take a math course that covers things like exponentials. – Bryan Krause Jan 9 at 16:57
  • 2
    I don't think a math course is necessary - this could be explained at a bit more length in a blog post etc. - but I agree that it's unlikely to appear in a peer-reviewed medium, as it's fairly trivial from a scientific point of view – Ben Bolker Jan 10 at 17:46

As was noted in the comments, this depends what you actually compare it with (in a hypothetical alternative scenario). But hypotheticals aside, if you're just looking for peer-reviewed statements on the implications of faster spread, it can obviously lead to more deaths. CDC's MMWR, in a technical/modelling article on the spread of the UK "Kent" lineage in the US, has this "for dummies" popsci figure:

enter image description here

  • Thank you. Appreciated. Up-voted and accepted as answer. – Nigel J Jan 16 at 9:21
  • 1
    For what's worth it, Fauci spoke about the issue at press conference youtu.be/UNDDM0m3fgw?t=253 – Fizz Jan 25 at 10:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .