Should I stay safe in order to avoid getting covid?
My assumption is yes, but I want to think it through with a cost/benefit analysis.
A better question might be “how much am I willing to do to avoid getting covid, such that the marginal cost of additional avoidance equals the marginal additional cost in the other direction of increased covid risk?”
Three Worlds
So lets imagine three worlds.
In the first world, I totally ignore coronavirus, take no precautions whatsoever, and have a c1% chance of getting covid. If everyone did this, then c1 would be about 70%. If everyone else continues to take some precautions but I don’t, then it’s probably something like 30%. I don’t think it’s a good idea to take no precautions when everyone else is; see discussion of externalities at the end.
In the second world, I take only the precautions which I don’t find very difficult, but do take precautions which seem super effective. I’d wear a mask, avoid going to restaurants, not spend too long in stores, but I’d see friends indoors whenever I wanted. But I wouldn’t go to dances or concerts, because that seems like a big risk. I think the American public at large is basically already doing this. So c2 is probably like 4%. It’s about 10% for Americans overall (200,000 deaths at a 0.7% mortality rate), but I’m not an essential worker, so it’s lower for me. I also wouldn’t choose to go to bars or restaurants, which probably makes a big difference.
In the third world, I take the precautions that I’ve been taking. I don’t interact with people outside, except for a small group. So c3 is like 1% or less.
Costs
In each of these three worlds, I incur a cost.
In the first world, I bear a large cost from getting covid, but a small cost from avoiding it.
In the second world, I bear a medium risk from both covid and avoidance.
And in the third world, I bear very little risk from covid, but a large risk from avoidance.
Which world should I choose?
In order to answer that, I need to estimate the cost of getting covid, and the cost of avoidance. I’ll measure the cost in QALYs (Quality Adjusted Life Years) to try to get comparable units.
How Bad is it to Get Covid?
What is the cost in QALYs of catching covid?
I see three parts: the risk of death, which has been well tabulated for different groups, the duration of the illness itself, which is probably about two weeks of suffering on average, and the risk of ongoing health problems.
Ongoing health problems is a spectrum — lots of people will have a mild cough for a month, and a small percentage will develop long-lasting chronic fatigue. For this calculation, I’ll assume there’s a 10% chance of having bad fatigue for one year, and a further 10% chance of having mild fatigue for the rest of my life. Those numbers are somewhat made up, and someone gleaned from looking at articles and studies about long-haulers, although there’s nothing high quality.
As a person in my early 30s, I have about a 0.2% chance of dying from a covid infection, or maybe less, since I’m in good health. The illness might involve weeks of being bedridden or it might merely be a mild cold, but I’ll say it averages 2 weeks being stuck in bed, which I’ll assume is awful, only 20% as good as normal life.
So the cost in QALYs is:
(0.2% chance of death * 60 years) +
(10% chance of bad fatigue * 1 year * .5 QALY ratio) +
(10% chance of long, mild fatigue * 60 years * .1 QALY ratio) +
(2 years (0.04 years) of sickness * 0.8 QALY ratio)
= 0.802 QALYs
So overall, getting covid is like losing 9 or 10 months of life, through early death or an extreme sickness of that duration. This doesn’t take into account temporal discounting.
Cost of Quarantine
Then the second part of the question is: what’s the cost of covid avoidance? Let’s assume that quarantine will last for another year. Depressing, but probably realistic.
Then compared to the normal world, world one is 85% as good, since quarantine is still going on around me but I do whatever I want, or 90% as good if everyone stops quarantine but the mass death and sickness around us is still sad.
World two is 75% as good, since I still get to see people and do lots of social things, but lots of events are still canceled, and I don’t go into the office.
And world three, the world of maximum precautions, is 65% as good. That’s like saying that I would be indifferent between being in a coma for four months and then have the world go back to normal than to do full quarantine for a year.
Overall Costs
Here are the costs:
World 1 (everybody): 70% * 0.8 + 0.10 = 0.61 QALYs lost
World 1 (just me): 30% * 0.8 + 0.15 = 0.39 QALYs lost
World 2: 4% * 0.8 + 0.25 = 0.282 QALYs lost
World 3: 1% * 0.8 + .35 = 0.358 QALYs lost
So this makes it seem like I should relax my precautions, and accept a 4% chance of getting covid, if it means I get to see people more.
Obviously these numbers are different for different people. Someone with a chronic lung condition probably sees covid as more costly, and married introverts probably see the cost of isolating as lower. And the opposite is true for others.
Externalities
Now all this ignores the risk to other people. Being a disease vector who gets other people sick creates negative externalities. But Americans overall are accepting a 10% risk of getting covid per year. If I change my personal risk from 1% to 0.5%, then that doesn’t affect the overall average very much, since it’s the additive average that matters, not the multiplicative average. So in a general “good citizen” sense, I feel like I’m helping to fight covid by pursuing a personal infection rate that’s less than half the average.
When thinking about the risk that my behavior has on others, I should focus on the more direct risks. Someone who is several degrees of separation away from me is getting most of their risk determined by broader society. The people who are closest to me are much more greatly affected by my risk, since they bear some of it too. I should talk to them about what level of trade-off they prefer.
Conclusion
I think this analysis is convincing to me that I should target a covid infection risk of about 4%, rather than the <1% that I was previously targeting, because I think the cost of isolation is relatively high, even compared to the considerably cost of catching covid.
A lot of the numbers here are guesswork. In particular, the likelihoods of ongoing health problems from covid dominate the QALY cost and they’re total guesses. Also, measuring the cost of living in quarantine in QALYs is total guesswork. You can see from the chart here that I rate the cost of being in quarantine as similar to the cost of living with malaria or tuberculosis, which makes me think I’ve overestimated the cost of quarantine.
In order to get a more fine grained conclusion, it would be helpful to estimate the cost in microcovids and the cost of avoidance for a bunch of different activities.