Those readers who’ve been with us since our January coverage of the SARS-CoV-2 outbreak in Wuhan, China, know that The Prepared has consistently predicted the twists and turns of this pandemic. Our blog has been early on many stories, and our COVID-19 scenario guide has been accurate in detail about how many aspects of the pandemic would unfold.
As the editor responsible for figuring out what’s next with the pandemic so that we can stay ahead of things, I want to lay out the details of the mental model I’ve been using (and actively adjusting/updating!) to understand where we are and where we’re going.
There is no real secret sauce here. Anyone can do what I’m doing if they’re willing to dedicate enough time and energy to it. It’s a pretty basic set of guidelines, attitudes, and assumptions, combined with constant scrolling, talking, arguing, and thinking.
Attitudes and methods
My guiding star is that I pay close attention to experts in order to adjust the basic assumptions of my mental model of the virus and how it spreads. I have a Twitter feed of these people, and I check it regularly. We’re also subject-matter experts who contribute to The Prepared, and I bounce new information off of them when I get it.
Here’s the second most important thing I do: I stick to my mental models regardless of what any particular crowd thinks at any given moment.
I see so many smart people get derailed by political tribalism in their thinking about this pandemic. If their ideological opponents (the media, the left, the right, etc.) say one thing about the virus or the outbreak, they feel obliged to say the opposite. Even experts fall victim to this. So I factor political bias into my thinking in two ways:
- I police my own thinking for this tendency. I have to be totally comfortable publicly agreeing with facts about the pandemic with people whose views I otherwise find abhorrent, no matter the social cost. Likewise, I have to be comfortable telling people I’m normally aligned with, “You’re all totally wrong, and these other folks we don’t like are right.”
- I do due diligence on the experts I follow, so that I can develop a sense of if they’re prone to falling into this trap. If they are, I adjust my evaluation of their output accordingly. I don’t discount what they say, I just adjust it to control for political bias.
I take it as a given that mathematical models are basically worthless for understanding what’s next with the pandemic. I’ve spent a ton of time monitoring the different sophisticated models that track this pandemic, and I’ve learned that they’re best ignored. There are two straightforward reasons why even the best models have consistently failed to produce useful, actionable predictions about where the spread will go next:
- Like weather models, the further into the future you go, the less accurate they are. To the extent that the pandemic models are good, they’re good a few days out at most. By the time you get to two weeks, the error bars are so wide as to be worthless.
- The virus’s spread is entirely contingent on human behavior, and you cannot predict human behavior.
I’ve come to believe that the main role models have played in this outbreak is to convince the public that experts are clueless and that institutions and authorities are not to be trusted. The models have been a massive own goal for public health, and they should have been actively discounted and downplayed from the beginning.
Finally, I constantly keep in mind that random chance plays a huge role in what happens when and where, and how bad it is. There is a lot of randomness to the timing of the outbreaks — both their beginnings and their acceleration — in different locations and populations. Not every city blows up at the same time, and some are hit way harder than others. So you can’t look at a population’s qualities and say, “These people are definitely gonna get hammered in the next six months.” Maybe they will, or maybe they’ll get lucky and it’ll miss them entirely.
How to think about the pandemic
My constantly evolving mental model of the pandemic has three parts:
The virus (SARS-CoV-2) and disease (COVID-19): Qualities like fatality rate, ease and method of transmission, serial interval, rates of asymptomatic/presymptomatic/symptomatic transmission, length of hospitalization, symptoms, etc.
The population: The characteristics of a given population in a specific geography that make it more or less susceptible to the virus. This is mostly age distribution and prevalence of various comorbidities in a cohort.
The society: The larger social factors, which include everything from government mask mandates, to average hospital occupancy rates, to compliance with social distancing measures, to every other form of institutional and behavioral factor that could affect the pandemic.
The pandemic isn’t just a biological thing — it’s all three of those factors interacting in different ways.
I have my own internal understanding of each of these aspects, and I don’t adjust that understanding without significant, persuasive evidence. Again, I don’t factor in what most other people whom I tend to agree with or feel an affinity with are saying. I stick to my own understanding and adjust it carefully.
Now I’ll zoom in on each of the above, in turn, in order to get into specifics.
The virus and disease
Here are the things I currently believe about the virus.
SARS-CoV-2 is a novel coronavirus. It’s wild that this needs to be said, but this virus is new in the human population, so we are not immune to it. Prior to a vaccine, the only way to get immunity to it is to catch it and recover from it.
It spreads primarily via synchronously shared air pockets. Transmission is mainly the result of a spreader being in a closed space with other people at the same time. I’m sure surfaces also play a role, and there could be some amount of transmission via lingering aerosols that stay in the air after the spreader has left the room. But in the main, you catch it when you’re in the same indoor space at the same time as a person who’s contagious.
Outdoors is mostly safe, unless you’re on top of each other and/or yelling and singing. This is the flip side of it being mostly transmitted via shared air pockets.
We are nowhere near herd immunity, which is something like a 60-percent attack rate. People talk themselves out of this idea with claims about T-cell cross-immunity and so on. And there is some evidence that such things are happening and are important. If we learn more about this I’ll lower my assumed threshold for herd immunity. But for now, I consider any talk of herd immunity at, say, a 20-percent attack rate to be wishful thinking, so I ignore it.
The virus spreads unless we intervene. The virus does not “get tired” or otherwise magically pack up its bags and go home once it reaches some magic threshold — one popular theory suggests that the virus always peaks at 40 days, regardless of population. There are other theories of “immunological dark matter” and so on. I currently take all this to be wrong. If we don’t intervene via distancing and masking up, the virus will spread.
The virus stops spreading when we intervene and change our behavior. The inverse of the above is that we have control over the spread, and can alter our behavior in such a way as to stop it from spreading. Again, there are people who do not believe this to be the case — they think lockdowns and masks have no impact. These people are wrong. Behavior matters.
A lot of spreading happens asymptomatically or presymptomatically. In other words, people are spreading it either without ever showing any symptoms or before they develop symptoms.
The virus’s dispersion factor is low. What this means is that most people who catch SARS-CoV-2 do not spread it to others. Only some smaller percentage, maybe about 20 percent, are actually spreading it to others. This has the following very important implication: transmission chains are fragile and easy to disrupt via social distancing. Again, behavior really matters a lot, and social distancing is extremely effective.
Here’s a concrete example to help you understand how important this is: Imagine a church with 1,000 members who meet every Sunday in a large building. On Monday, 100 members catch the virus. If every one of those 100 church members is contagious, then to stop the virus from spreading this coming Sunday all 100 of them have to stay home. But if only 20 of them are contagious, then only 20 need to stay home to stop the spread.
Masks work, but they’re not a one-stop cure-all. Consistent with my early hypothesis that COVID-19 is a disease of adults sharing indoor air pockets synchronously, I have been a proponent of masks since the start of the outbreak. I still am, but they are not a cure-all. This virus is very easily spread, and while universal masking is an important part of stopping it, it is not by itself sufficient to halt the pandemic. I know there are many news stories about this or that country beating it entirely with masking alone, but those are all gross oversimplifications. We need both masking and distancing.
The fatality rate of COVID-19 is dramatically higher than that of seasonal flu. Since the beginning of this outbreak, we at The Prepared have stuck to the evidence that COVID-19 has a roughly 1 percent fatality rate in most representative population cohorts (i.e. average age distribution, but more on this below). This fatality rate comes from studies where we know how many people were infected via testing and how many subsequently died.
Flu fatality rates, in contrast, are mostly just estimates. We don’t measure them the same way we measure COVID-19 fatality rates, i.e. by testing for it as much as we possibly can. If we did, experts say that the real-world flu fatality rates would be even lower than the ~0.02% numbers you see cited everywhere. So people who look at the low fatality rate of COVID-19 in the younger population, and then compare it to some overall flu fatality rate of ~0.02%, are making apples-to-oranges comparisons and are wrong.
COVID-19’s fatality rate is heavily age-dependent. Like pretty much every other disease or medical condition, COVID-19’s mortality is dramatically affected by age. Given this, it’s not worth focusing too much on fatality rates, because what a measured fatality rate really tells you is the average age of the infected population.
COVID-19 is a nasty illness with some serious non-lethal effects. This disease isn’t “just the flu, but deadlier”. When you get over the flu you’re done with it, but when you get over COVID-19 you may have lingering problems and permanent internal damage.
There are some areas of the world where most of the old people are already dead, and where people are mostly young, have a low BMI, and spend a ton of time outside. I don’t expect to see many COVID-19 fatalities in these areas, nor do I expect to see the kinds of crippling outbreaks we saw in Wuhan, Italy, New York, and other areas.
In areas where people skew more elderly, or where they have different body shapes and comorbidities associated with increased COVID risk, I expect things to be much grimmer.
In general, though, I try to keep in mind the particular demographic characteristics of a population in a given region when I’m trying to think through how the pandemic may unfold there.
The social and economic forces that push people together into shared air pockets are powerful. So unless we’re actively trying to keep people apart with distancing measures, or unless they’re afraid enough by what they’re seeing spread in their own networks to stay home, then they will cram back into closed spaces and breathe on each other.
Some people don’t believe it’s a threat to them personally until it happens to enough people they know. I’m always shocked at the level of denial people will live in about threats to their routine and way of life. Until they know enough people in their extended network who either have a nasty experience with this virus or who die from it, they’ll mostly assume it’s all a bunch of hype.
People will change their behavior to avoid catching COVID-19 when they see enough of it first-hand. The flip side of the above is that people eventually do grasp the nature of the threat and modify their behavior accordingly. Once they see it up-close, they decide they don’t want it and begin taking precautions.
Government policies and actions matter, but they’re not determinative. What ultimately matters for this pandemic is what people do in large numbers. To the extent that a government’s policies and pronouncements can affect what large numbers of people do, governments can make things better or worse. But there is no simple relationship between, say, the announcement of a lockdown’s start or end and people’s actual distancing behavior. The latter is driven by many factors other than government fiat.
Everyone is gaming and juking everything, everywhere, all the time, for all sorts of reasons. If you think the case counts aren’t reliable, you’re almost certainly right. There’s a lot of weirdness going on there, and I don’t even pretend to understand all of it. I just know that since the beginning, when we expressed strong skepticism about the Chinese numbers, case counts have been unreliable in most countries, most of the time.
Death counts have some similar problems. They’re delayed by weeks, and it’s sometimes hard to know what people died from specifically. You can look for larger “all-cause” mortality spikes, but in the end, I think we will never really know what the true body count of this virus was in many countries (the US very much included).
Hospitalizations and ICU load numbers have their own issues related to our for-profit medical system, the particulars of the hospital chain in an area, and the state and federal apparatus they’re reporting their data to.
In short, I keep a close eye on all the relevant numbers we’re using to measure the pandemic, but I also generally believe many of the widespread reports of this or that type of shenanigans and incompetence in regards to collecting those numbers. It’s all a hot mess, and very hard to make sense of.
Second-order effects and feedback loops are both guaranteed to be dramatic and impossible to predict. I consider the protests that have rocked the nation a direct result of the lockdown and the pandemic. I don’t think they’d be anything like the scale we’re seeing without COVID-19.
Well, there are yet more dramatic crises coming. I have no clue what they are, but things that are as big as the George Floyd protests are in our near- to medium-term future. Our large, complex social and economic system is in an unstable state, and more big parts of it will break suddenly and dramatically before this pandemic is over.