In writing a reply to Zvi’s recent post on Covid-19 predictions I realized I had been procrastinating on writing out my thoughts on Covid-19 in full. In this post I will explain in detail how I think about the Covid-19 countermeasures for the next few months, and explain why I think all this.
Disclaimer: my predictions are among the most pessimistic (and speculative) of the ones I’ve run into. Having expectations that COVID-19 will be worse than people are thinking so far can and does have negative effects on your mental health, so stop reading, now.
I’d love to be proven wrong on all of this.
What is our long term goal?
There has been a lot of discussion on the impact of countermeasures on the spread of Covid-19. However, most of this has been relatively unhelpful in trying to make predictions on longer timescales—say, until March 2021. For this reason I want to first figure out what plausible scenarios we have for the long term—i.e. what situations would be stable. I think there are three main contenders.
We establish group immunity, and the virus is no longer a great risk.
We reduce the incidence rate to such a low level that we can prevent the spread by actively intervening in every case (for example through contact tracing, isolating only people who have been exposed).
We (permanently) change our lifestyle to one with reduced social contact, and we make sufficiently many countermeasures against Covid-19 permanent to reduce R0 to 1 or below.
Some countries are currently aiming for option number 2 (for example South Korea, as far as I know). I think this places serious strain on globalization, but I must confess I haven’t looked into this all that much. I think experts are saying it is a near certainty that Covid-19 will become endemic, so this approach would require constant vigilance to prevent new cases. If it is possible to turn this approach into a permanent state of affairs I would love to know more about it, but for now I’d like to ignore it. It also sounds very difficult in places like the US, with approximately 1,189,003 active cases today.
Option number 3 is the doom scenario economists are talking about. Entire industries will have to adapt (or, more likely, collapse). We would, as a species, be cutting into our growth to survive this virus. Most of the discussions I’ve had are about the (lack of) desirability of this scenario versus option 1. I personally think a lot of the severe countermeasures we are currently taking are unsustainable, ruling out option 3 (see for example the dramatic US unemployment numbers).
For the rest of this post I will discuss only option 1. For the record: there is some overlap between options 1 and 3, in that the threshold for group immunity depends strongly on R0, which can be brought down by changes in our way of life. In the section on ‘the R0 scale’ below I will go into this in a bit more detail, but for now I am assuming ‘group immunity’ means that somewhere between 20% and 70% of the population has gained immunity to Covid-19.
How do we get group immunity?
To the best of my knowledge, there are currently two ways. Either people get exposed to the virus, thereby developing immunity, or we create a vaccine. In the first case it is likely millions worldwide will die—Zvi gives the estimated IFR (Infection Fatality Rate) at 0.5%-1%, and it plausibly goes up if our medical systems get overwhelmed. But even if this stays under control, 0.5% of over 20% of the population is still 78 million people (or 328,000 in the US).
The alternative is a vaccine. This takes effort, funding and time to develop, although personally I am only effected by the ‘time’ consideration. Some quick googling reveals that an optimistic length of time to develop a vaccine is about 1-2 years. Sticking with the optimistic lower end of that prediction, I started looking for predictions on Covid-19 for March 2021 about two months ago. The question I was now trying to answer is “how do we manage Covid-19 until approximately March 2021, in the hope that we have a vaccine by that time (or can extend our approach a few more months if not)?”.
Quarantine, and the R0 scale
Viewed through the lens of the above question, what we really need is a way to stall Covid-19 for about a year. If we can keep everybody alive and healthy through this period of time we can plausibly return to life as normal. Based on the post The Hammer and The Dance, the most realistic strategy seems to be to implement sufficiently many countermeasures to keep R0 at 1 (or, technically, oscillating around it), which is the point that gives us maximum quality of life while keeping the virus in check. The problem is that I don’t think we can do this, and certainly not for a full year.
It is currently an open question how big the impact of individual countermeasures exactly are on R0, and therefore also which ones can be relaxed while which ones have to remain enforced to control ‘The Dance’. To still get some rough idea, I personally think about something I call “the R0 scale”. We know that currently in many places across the world R0 is close to 1, that in the brief period of serious public response to the initial outbreak it probably dipped as low as 0.7, and it peaked around 2.5-3 when people lived their lives as normal. This means that we should place our measures on a scale of “a serious quarantine” at 0.7, via “life as you are experiencing it right now” somewhere around 1, to “life as normal” around 2.5. Since our threshold for ‘acceptable’ is at R0=1, this suggests to me that ‘The Dance’ will look a lot more like ‘a full quarantine, but with a few restrictions lifted’ than like ‘restoring social contact, but wash your hands and wear a face mask’.
As I said above this is an open question, and my conclusion is mostly guesswork. Zvi pointed out that risks follow a power law, so maybe we can get almost all of the reduction in R0 with almost none of the measures. But personally I am skeptical of this idea.
That which doesn’t change
...stays the same. In March and April many places across the world have gone into quarantine to prevent the spread of Covid-19. I support that decision (in fact, I would have liked it if everybody responded faster). If, at some point in the future, we have the same number of contagious people, and are not at an appreciable fraction of group immunity, it will at that point again be a solid decision to go into quarantine (or to extend it). There are other considerations that confound this decision (prolonged quarantine has non-linear effects on the economy, a lot of companies only have a few months worth of liquidity, for example), but from a medical point of view the argument is sound. If the number of contagious people is the same as before, and you revert the Covid-19 countermeasures, the exponential growth will pick up right where it left off last time. This is as true in March as it will be in November.
This is why I personally think the idea of a ‘second peak’ is flawed or misleading. Covid-19 growth is subject to a control system made of human response. We can, in theory, produce as few or as many peaks as we desire by modulating our countermeasures. Conversely, relaxing measures will do exactly this (create new peaks, when we go over R0=1). It’s not like the virus will respond differently to increased social contact a month from now than it did two months ago. This is why the timeline for the vaccine is so important—all the other countermeasures we currently have are only stalling for time, they are not exit strategies. In particular, I think this time scale is the only relevant one for our long term plans. Multiple people in my social circle are making plans for the summer holidays, or the fall semester, under the assumption that we will have Covid-19 under control by then and the countermeasures will be significantly relaxed. However, I don’t see what will have changed between the situation as it is right now and as it will be in August. Why and how will things have changed?
So will all this work?
No. Having considered all of the above, I think it is exceedingly unlikely that we will be able to maintain a state sufficiently similar (i.e. R0=1) to our current quarantine for almost a year. The impact on the economy is devastating, unemployment numbers are skyrocketing and people are unable to endure being cooped up for that long. (Again) Zvi pointed out that it is highly suspicious that currently R0 is so close to 1, and that this is more than likely to be a feedback system. I agree, it is a feedback system of people balancing their fear of Covid-19 and wish to comply with government orders against their reluctance to stay at home/cut back on social activities/miss out on work, mediated by the official reports and statistics. But I expect the cost of staying at home to rise dramatically in time (due to mental health, lack of income, etc.), while the risk of Covid-19 stays roughly constant (and the fear of it might even go down). In short, I think the current balance is not sustainable.
Which puts me at a serious impasse. The text above is far from rigorous, and there might be any number of creative solutions that I have not considered. But personally I think “maintain The Hammer and The Dance until we have a vaccine” is one of the most realistic-sounding long-term approaches to Covid-19 (mostly because a vaccine is our only exit strategy, besides letting the virus run rampant). And I think it will fail dramatically, because such a quarantine is not sustainable for that long.
There is a lot more I would like to say about all of the above, most of which I have very low confidence in but would love to have some discussion on (in particular the ‘second peak’ and ‘the R0 scale’). I have much higher confidence in any of the individual steps than in the conclusions, and I hope that the discussion can be kept at the same coarse-grained level of detail as the one that I have tried to keep for this post.
My prediction for Covid-19
In writing a reply to Zvi’s recent post on Covid-19 predictions I realized I had been procrastinating on writing out my thoughts on Covid-19 in full. In this post I will explain in detail how I think about the Covid-19 countermeasures for the next few months, and explain why I think all this.
Disclaimer: my predictions are among the most pessimistic (and speculative) of the ones I’ve run into. Having expectations that COVID-19 will be worse than people are thinking so far can and does have negative effects on your mental health, so stop reading, now.
I’d love to be proven wrong on all of this.
What is our long term goal?
There has been a lot of discussion on the impact of countermeasures on the spread of Covid-19. However, most of this has been relatively unhelpful in trying to make predictions on longer timescales—say, until March 2021. For this reason I want to first figure out what plausible scenarios we have for the long term—i.e. what situations would be stable. I think there are three main contenders.
We establish group immunity, and the virus is no longer a great risk.
We reduce the incidence rate to such a low level that we can prevent the spread by actively intervening in every case (for example through contact tracing, isolating only people who have been exposed).
We (permanently) change our lifestyle to one with reduced social contact, and we make sufficiently many countermeasures against Covid-19 permanent to reduce R0 to 1 or below.
Some countries are currently aiming for option number 2 (for example South Korea, as far as I know). I think this places serious strain on globalization, but I must confess I haven’t looked into this all that much. I think experts are saying it is a near certainty that Covid-19 will become endemic, so this approach would require constant vigilance to prevent new cases. If it is possible to turn this approach into a permanent state of affairs I would love to know more about it, but for now I’d like to ignore it. It also sounds very difficult in places like the US, with approximately 1,189,003 active cases today.
Option number 3 is the doom scenario economists are talking about. Entire industries will have to adapt (or, more likely, collapse). We would, as a species, be cutting into our growth to survive this virus. Most of the discussions I’ve had are about the (lack of) desirability of this scenario versus option 1. I personally think a lot of the severe countermeasures we are currently taking are unsustainable, ruling out option 3 (see for example the dramatic US unemployment numbers).
For the rest of this post I will discuss only option 1. For the record: there is some overlap between options 1 and 3, in that the threshold for group immunity depends strongly on R0, which can be brought down by changes in our way of life. In the section on ‘the R0 scale’ below I will go into this in a bit more detail, but for now I am assuming ‘group immunity’ means that somewhere between 20% and 70% of the population has gained immunity to Covid-19.
How do we get group immunity?
To the best of my knowledge, there are currently two ways. Either people get exposed to the virus, thereby developing immunity, or we create a vaccine. In the first case it is likely millions worldwide will die—Zvi gives the estimated IFR (Infection Fatality Rate) at 0.5%-1%, and it plausibly goes up if our medical systems get overwhelmed. But even if this stays under control, 0.5% of over 20% of the population is still 78 million people (or 328,000 in the US).
The alternative is a vaccine. This takes effort, funding and time to develop, although personally I am only effected by the ‘time’ consideration. Some quick googling reveals that an optimistic length of time to develop a vaccine is about 1-2 years. Sticking with the optimistic lower end of that prediction, I started looking for predictions on Covid-19 for March 2021 about two months ago. The question I was now trying to answer is “how do we manage Covid-19 until approximately March 2021, in the hope that we have a vaccine by that time (or can extend our approach a few more months if not)?”.
Quarantine, and the R0 scale
Viewed through the lens of the above question, what we really need is a way to stall Covid-19 for about a year. If we can keep everybody alive and healthy through this period of time we can plausibly return to life as normal. Based on the post The Hammer and The Dance, the most realistic strategy seems to be to implement sufficiently many countermeasures to keep R0 at 1 (or, technically, oscillating around it), which is the point that gives us maximum quality of life while keeping the virus in check. The problem is that I don’t think we can do this, and certainly not for a full year.
It is currently an open question how big the impact of individual countermeasures exactly are on R0, and therefore also which ones can be relaxed while which ones have to remain enforced to control ‘The Dance’. To still get some rough idea, I personally think about something I call “the R0 scale”. We know that currently in many places across the world R0 is close to 1, that in the brief period of serious public response to the initial outbreak it probably dipped as low as 0.7, and it peaked around 2.5-3 when people lived their lives as normal. This means that we should place our measures on a scale of “a serious quarantine” at 0.7, via “life as you are experiencing it right now” somewhere around 1, to “life as normal” around 2.5. Since our threshold for ‘acceptable’ is at R0=1, this suggests to me that ‘The Dance’ will look a lot more like ‘a full quarantine, but with a few restrictions lifted’ than like ‘restoring social contact, but wash your hands and wear a face mask’.
As I said above this is an open question, and my conclusion is mostly guesswork. Zvi pointed out that risks follow a power law, so maybe we can get almost all of the reduction in R0 with almost none of the measures. But personally I am skeptical of this idea.
That which doesn’t change
...stays the same. In March and April many places across the world have gone into quarantine to prevent the spread of Covid-19. I support that decision (in fact, I would have liked it if everybody responded faster). If, at some point in the future, we have the same number of contagious people, and are not at an appreciable fraction of group immunity, it will at that point again be a solid decision to go into quarantine (or to extend it). There are other considerations that confound this decision (prolonged quarantine has non-linear effects on the economy, a lot of companies only have a few months worth of liquidity, for example), but from a medical point of view the argument is sound. If the number of contagious people is the same as before, and you revert the Covid-19 countermeasures, the exponential growth will pick up right where it left off last time. This is as true in March as it will be in November.
This is why I personally think the idea of a ‘second peak’ is flawed or misleading. Covid-19 growth is subject to a control system made of human response. We can, in theory, produce as few or as many peaks as we desire by modulating our countermeasures. Conversely, relaxing measures will do exactly this (create new peaks, when we go over R0=1). It’s not like the virus will respond differently to increased social contact a month from now than it did two months ago. This is why the timeline for the vaccine is so important—all the other countermeasures we currently have are only stalling for time, they are not exit strategies. In particular, I think this time scale is the only relevant one for our long term plans. Multiple people in my social circle are making plans for the summer holidays, or the fall semester, under the assumption that we will have Covid-19 under control by then and the countermeasures will be significantly relaxed. However, I don’t see what will have changed between the situation as it is right now and as it will be in August. Why and how will things have changed?
So will all this work?
No. Having considered all of the above, I think it is exceedingly unlikely that we will be able to maintain a state sufficiently similar (i.e. R0=1) to our current quarantine for almost a year. The impact on the economy is devastating, unemployment numbers are skyrocketing and people are unable to endure being cooped up for that long. (Again) Zvi pointed out that it is highly suspicious that currently R0 is so close to 1, and that this is more than likely to be a feedback system. I agree, it is a feedback system of people balancing their fear of Covid-19 and wish to comply with government orders against their reluctance to stay at home/cut back on social activities/miss out on work, mediated by the official reports and statistics. But I expect the cost of staying at home to rise dramatically in time (due to mental health, lack of income, etc.), while the risk of Covid-19 stays roughly constant (and the fear of it might even go down). In short, I think the current balance is not sustainable.
Which puts me at a serious impasse. The text above is far from rigorous, and there might be any number of creative solutions that I have not considered. But personally I think “maintain The Hammer and The Dance until we have a vaccine” is one of the most realistic-sounding long-term approaches to Covid-19 (mostly because a vaccine is our only exit strategy, besides letting the virus run rampant). And I think it will fail dramatically, because such a quarantine is not sustainable for that long.
There is a lot more I would like to say about all of the above, most of which I have very low confidence in but would love to have some discussion on (in particular the ‘second peak’ and ‘the R0 scale’). I have much higher confidence in any of the individual steps than in the conclusions, and I hope that the discussion can be kept at the same coarse-grained level of detail as the one that I have tried to keep for this post.