* It is more likely that MMR has few significant long-term health effects than if COVID-19 does
* COVID-19 effects are unlikely to be binary outcome
* Effects need not be universal for them to affect parental behaviour
However, none of that agreement is important to the discussion I was hoping to have. My comment was solely on the epistemology. I suspect you also use the same methods I was talking about in my previous comment.
If that is the case, then perhaps some other day we can argue about what the priors are. If we're good about it, we'll end up with the Aumann Agreement Theorem (+ corollaries) result, and rapidly agree (through likely both of us shifting our priors as we exchange information).
I think we have no disagreement and am happy to conclude this here.
It also doesn't have to be a universal effect for it to generate effectively universal concern among parents. If it turned out that 1/20 kids suffered permanent kidney damage due to infection, most parents would keep their kids home. Look at the stats on school shootings and note that we pointlessly traumatize kids to assuage the concerns of parents. That's your figure of merit: that's how powerful the long-term C19 effect would need to be to change the behavior of the median parent.