The Logic of the Mask

A bad decision doesn't become a good decision just because the worst doesn't come to pass.

So most days these days, I wind up reading Andy Slavitt’s daily COVID-19 update thread, which I’ve come to think of as being the closest thing to the fireside chat we might be having in the event of an actual president.

They’re interesting, useful, sometimes reassuring and not prone to either panic nor pulling punches. Andy strives to tell it like it is, as he sees it.

Of course, I don’t always see it as he sees it, and that’s fine, but yesterday he said something I found egregiously wrong, to the point I feel like it must be addressed.

Speaking about the situation in Georgia, he counseled a wait-and-see attitude towards Kemp’s decisions re: opening rather than immediately criticizing before we know the results.

The thing is, he’s correct that we can’t say for certain that Georgia will be materially worse off than other states that did better. This is not a Just World, not in the sense that outcomes are just and not in the sense that anything is just one thing. People can make the wrong decision for the wrong reasons and come out ahead through sheer luck. People can ignore the warnings and take the worst risks and it won’t force anything bad to happen.

The virus can’t spontaneously generate out of thin air just because it senses people without masks playing beach volleyball. Someone has to bring it to the party. That’s why various state restrictions have focused on the size of gatherings: it’s not that the virus will leave if it gets to the house and sees fewer than 10 people are there, it’s that each new person increases the chances that the virus gets there.

And just to be clear, I don’t expect Andy Slavitt will ever personally read this newsletter, so I’m not trying to explain this to him thinking he doesn’t get it. This is not “Andy Slavitt is horrible and cancelled and I’m never going to listen to him again.” He just tweeted a thing I disagree with.

He went on talk about randomness in the same thread:

And this is correct. Saying the virus has an R value of around 3 doesn’t mean there’s a rule that when you get the virus, you give it to 3 people. Some people give it to nobody, and all those people are averaged out by people who pass it on to more than 3 people, sometimes many more. Just where you are and what you’re doing when you get into the invisible spreading phase of it can have a lot of impact, and that’s obviously going to be random, subject to way too many variables for us to predict or figure out.

It’s possible to fire a bullet into a crowded room and hit no one. It’s possible to fire a bullet into a crowded room and hit several people with the same bullet. There’s no guarantee that firing a bullet into a crowded room will hit anybody.

But here’s the thing, and the reason we can and should criticize Kemp and other governors who disregard warnings and open early:

You still shouldn’t fire a bullet into a crowded room. If you do so and no one gets hurt, this doesn’t mean you were right! It means you were lucky! It means everyone was lucky!

Now imagine there are fifty people with guns and for whatever reason many of thema re eager to let off a round indoors and the rest are willing to be convinced, and for whatever reason there’s a great big building-wide conversation about whether this is a good idea or not.

If one of them decides to just test the theory out and gets lucky, that person should still be criticized!

This is the same logic whereby we all wear seatbelts even though most car trips don’t involve any collision at all.

It’s the logic of the mask. It’s entirely possible to wear a mask and not spread the virus — only people who have it can spread it, and we don’t know who has it.

If everybody wears the mask, a lot of people who didn’t have any chance to spread the virus anyway are wearing a mask “for no reason” but that’s the only way to make sure the people who are capable of spreading it wear one.

If Georgia squeaks by without a devastating outbreak after Kemp’s carelessness, then Georgia is lucky. What if all the states around Georgia see how lucky Georgia is and decide that Kemp was right? Will they all be so lucky? If one of them has a bad outbreak, it might spread back to Georgia and that’s the end of Georgia’s luck.

As Andy Slavitt has noted many times, the lag between policy choices and their effects — as well as between infection and illness — is making it hard to connect outcomes to their events right now. If the takeaway from a “successful” Georgia opening is that Kemp was right not that Kemp took a wildly irresponsible risk with a lot of people’s lives and those people, fortunately, made it through okay, then we’re likely to see other states follow suit… and because of the delay, because of the disconnect, the “success” or “failure” of all those decisions won’t be immediately apparent, which is to say they will all at first appear indistinguishable from a success.

What I fear here is a snowball effect of other states following Georgia’s lead, and still more states looking at those states reaping the financial benefits of re-opening and no outbreak in sight so they reopen, and so on.

The more states reopen, the greater the pressure to reopen will become and the harder it will be to get any relief for states (and businesses, and people) that are still locked down. Already the Republicans have all but declared they’re done with relief measures. Already the narrative is about lazy blue states want to leech off the government forever while red states are eager to go back to work.

I don’t disagree with Andy that we should be careful how we talk about the situation with Kemp in Georgia. He’s right that we shouldn’t assume there will certainly be a devastating outbreak, like there’s an inexorable law of physics or a vengeful god of plague that will make it happen because he did the wrong thing.

But we should be clear: he’s doing the wrong thing. It’s unwise, it’s risky, it’s irresponsible, and while we can’t say that any one state will have a devastating outbreak, I think even the most conservative risk estimates would show that if all fifty states do what Georgia is doing, there will be large scale outbreaks.

And because of the delay between events and outcomes, and the mobility of people, those outbreaks won’t stay in the places they happen.

We can’t guess which states will certainly have those devastating outbreaks, any more than we can predict who will live and who will die if they do get the virus and any more than we can know without testing who is capable of spreading the virus and thus in need of a mask.

We all wear masks so that we all have reduced spreading capability. That’s the logic of the mask, and the logic of the shutdowns.