I’ve been putting off answering this, because a proper answer would require diving into a lot of disparate evidence for some background models. But developments over the past few weeks have provided more direct data on the key assumption, so I can now point directly to that evidence rather than going through all the different pieces of weaker prior information.
The key assumption underlying my belief (and, presumably, most other peoples’ belief) in the efficacy of the commercial vaccines is that the data from the clinical trials is basically true and representative. If we believe the data, then the health risks are trivial (compared to COVID), and the effectiveness is decent, and our certainty about these two facts is very high. The vaccines are not “lightly tested” in any epistemically-relevant sense, only relative to the frankly-unnecessary extreme over-testing typically used. (Zvi’s posts frequently provide helpful snapshots of the relevant data/analysis.)
The question which you seem to be interested in is whether the data is trustworthy. The key model here is that, if there were anything even remotely suspicious in there, then regulators and the media would absolutely freak out over it. That’s exactly what we’ve seen over the past few weeks—first with the AZ clotting thing, then with AZ calling their vaccine “79% effective” when the regulators thought it only merited “69 to 74% effective”. The clotting thing in particular is a clear case where there was absolutely no real, important problem, and regulators/media freaked out over it anyway, because they are super-over-sensitive to even the tiniest hint of a problem. That follows directly from their incentives: regulators do not get punished for delaying good vaccines, but if a bad vaccine gets regulatory approval, then there’s a scandal and an Official Investigation and “heads will roll”.
I’ve been putting off answering this, because a proper answer would require diving into a lot of disparate evidence for some background models. But developments over the past few weeks have provided more direct data on the key assumption, so I can now point directly to that evidence rather than going through all the different pieces of weaker prior information.
The key assumption underlying my belief (and, presumably, most other peoples’ belief) in the efficacy of the commercial vaccines is that the data from the clinical trials is basically true and representative. If we believe the data, then the health risks are trivial (compared to COVID), and the effectiveness is decent, and our certainty about these two facts is very high. The vaccines are not “lightly tested” in any epistemically-relevant sense, only relative to the frankly-unnecessary extreme over-testing typically used. (Zvi’s posts frequently provide helpful snapshots of the relevant data/analysis.)
The question which you seem to be interested in is whether the data is trustworthy. The key model here is that, if there were anything even remotely suspicious in there, then regulators and the media would absolutely freak out over it. That’s exactly what we’ve seen over the past few weeks—first with the AZ clotting thing, then with AZ calling their vaccine “79% effective” when the regulators thought it only merited “69 to 74% effective”. The clotting thing in particular is a clear case where there was absolutely no real, important problem, and regulators/media freaked out over it anyway, because they are super-over-sensitive to even the tiniest hint of a problem. That follows directly from their incentives: regulators do not get punished for delaying good vaccines, but if a bad vaccine gets regulatory approval, then there’s a scandal and an Official Investigation and “heads will roll”.