Thanks for the thorough write-up! I've been curious about that character, and have also been a little dismayed by the knee-jerk representations. Appreciate the care and time you took.
I appreciate your taking him seriously and agree he shouldn't be dismissed out of hand. But I see the discussion of all-cause mortality a little differently. A lot of misinformation operates by saying true things that imply false things. The speaker can defend themselves by saying that everything they said was literally true, while a lot of their audience takes away the false message. Like "There are a lot of Jews in Hollywood and also finance, and there is a Israel lobby, etc." The speaker never said there was a giant Jewish conspiracy but there is an inference there that a lot of people will hear.
The sign of a true statement being misinformation is when it's stated in a way consistent with the misleading implication, and the misleading implication isn't explicitly refuted. The statement "15 patients who received the vaccine died; 14 who received placebo died" suggests (to someone who doesn't know stats) that maybe the vaccine _increases_ all-cause mortality, for example. RFK never said that, but the implication is there. The giant omission is that the study is not powered to look at all-cause mortality. One might turn that into a more valid critique of the vaccine, like "look, if it doesn't affect all-cause mortality in a statistically significant way for 44000 random adults, the health benefits probably aren't huge". That is not what he said.
In this case, the tweet isn't even true on its face. He writes "The pivotal clinical trial for the @pfizer #Covid vaccine shows it does nothing to reduce the overall risk of death". But the trial doesn't show the vaccine does nothing; it shows that the ratio of death rates is probably between 0.5 and 2. That is very different from "shows it does nothing". (caveat: I haven't calculated those numbers carefully. But I think they're in the right ballpark.)
Hi Rajiv, interesting and balanced points. My own research leads me to discount a bit more these prediction markets' prices. We found that political markets are more biased toward 50% than other markets and this is compatible with a pressure of traders motivated by non-monetary incentives (partisans or manipulators). Increasing the demand of contracts with low prices is relatively cheap. So, the same amount of "motivated" money on both sides of the market can push the price up. The result is in "Do prediction markets produce well‐calibrated probability forecasts?" (EJ 2013).
Why would you care what a study says when we have reams of data from the last 3 years?? The vaccines were effective in mitigating severity and appear safe as of today. Hawaii shows a population didn’t need robust natural immunity to get to herd immunity. We are now in the endemic phase and the notion my getting vaccinated should have protected someone else is a red herring because it’s just a dumb notion.
Because it's placebo-controlled. Observational studies can't answer these questions definitively. You talk about Hawaii and RFK talks about Nigeria with minimal vaccination and low excess deaths. There is no doubt in my mind that the original vaccine prevented a lot of deaths in certain age and comorbidity groups, but less so for healthy young subjects, and even less so when it comes to the boosters.
In America age is a much bigger factor for a population on the low end than the high end because in general America is old and unhealthy. So populations like Utah and Travis County and DC would have had low Covid death rates even if they had done very little with respect to mitigation measures. All things being equal % below poverty level is the biggest factor for a population in America because poverty is a proxy for many public health issues and poverty ages people prematurely and our anti-poverty programs are directed at mothers with babies, i.e. young people.
Exactly my point, such confounding factors mean that there are limits to what you can learn from observational data, which is why placebo-controlled RCTs are so important, and there are not many of these, even the Pfizer study discussed in the post was unblinded after six months.
I was able to spot the Tennessee anomaly using my analysis…nobody that I know of spotted that but me. So according to my analysis Kentucky should have had a lower Covid death rate than TN because the Democratic governor was pro-masking but TN had the lower Covid death rate…until one day TN added thousands of Covid deaths to the official tally. I pointed out the anomaly and then shortly after it was no longer the case.
I agree there is tension here between what is pragmatic/effective with respect to countering his views during the campaign and his established pattern of harmful posturing on the vaccine front. Flaming and ridiculing him without counterargument during this campaign will probably backfire and may even garner sympathy for him among those who aren't familiar with his history of dubious advocacy.
On the other hand, for those who've been involved in countering his bad-faith arguments about vaccines and other issues for almost two decades, the reaction to his candidacy is anything but knee-jerk. While I've soured on him over the years, I think Sam Harris concisely describes why RFK Jr. is generally untrustworthy: https://www.samharris.org/podcasts/making-sense-episodes/325-a-few-thoughts-about-rfk-jr. Corporate influence in the FDA, errors on the part of the public health establishment—those are legitimate concerns better articulated by messengers other than RFK Jr.
The Science-Based Medicine blog, for instance, has been posting regularly about RFK Jr.'s activities since 2008. SBM uses a lot of derogatory language to describe him, perhaps unwisely, as doing so can come off as unmeasured when contrasted with Prasad's seemingly measured tone. But it's worth noting that Prasad himself has made a number of ludicrous and unmeasured statements, such as claiming that masking policies are akin to faith in "witchcraft". (Whatever one's opinion on the effectiveness of masking, I hope this statement by Prasad would be viewed as absurd.) Consequently, he may be too forgiving of RFK Jr. because he sees a potentially powerful ally in his own crusades against corporate regulatory influence and aspects of the coronavirus response, where he has not infrequently employed dubious reasoning about RCTs along with calibrated histrionics. Such debates are crucial, as we have much to learn from both the successes and failures of the coronavirus response, but they should be predicated on rigor of thought, not posturing.
I agree with much of this Jason, except that I don't think that RFK's arguments are in bad faith. He seems to hold his expressed views quite sincerely. And some of those views are not unreasonable, even if one disagrees with them (closing down most overseas bases for instance, or strictly enforcing border controls while expanding legal immigration). I have listened to him for many hours now and I think that he is worth listening to and engaging with, and on some issues could bring a healthy dose of dissent to the democratic party.
Prasad can be a bit over the top sometimes but I find it forgivable, given that there is still a lot of denial about policy errors during the pandemic. And I think he is especially upset about the masking of toddlers and the insistence that this was science-based.
Thank you for your measured and thoughtful comment.
The states that found the best balance between public health mitigation measures and a return to normalcy were the red states with Democratic governors. So Georgia could have had the same Covid death rate as North Carolina had it been a little more aggressive with mitigation measures AND the economy would not have suffered. So Georgia could have saved around 10,000 lives had it done what its neighbor did.
Unfortunately Matt Taibbi didn’t appreciate me correcting the #fakenews that he reports and he got Substack to delete all of my posts AND all of my comments. So I had some good Substack posts covering this and Substack just incinerated them to protect their cash cow. Substack’s “moderation guidelines” miraculously appeared a few hours after they deleted my account. ;)
Thanks for the thorough write-up! I've been curious about that character, and have also been a little dismayed by the knee-jerk representations. Appreciate the care and time you took.
Thanks for the comment Aaron, really wanted to get this off my chest
I appreciate your taking him seriously and agree he shouldn't be dismissed out of hand. But I see the discussion of all-cause mortality a little differently. A lot of misinformation operates by saying true things that imply false things. The speaker can defend themselves by saying that everything they said was literally true, while a lot of their audience takes away the false message. Like "There are a lot of Jews in Hollywood and also finance, and there is a Israel lobby, etc." The speaker never said there was a giant Jewish conspiracy but there is an inference there that a lot of people will hear.
The sign of a true statement being misinformation is when it's stated in a way consistent with the misleading implication, and the misleading implication isn't explicitly refuted. The statement "15 patients who received the vaccine died; 14 who received placebo died" suggests (to someone who doesn't know stats) that maybe the vaccine _increases_ all-cause mortality, for example. RFK never said that, but the implication is there. The giant omission is that the study is not powered to look at all-cause mortality. One might turn that into a more valid critique of the vaccine, like "look, if it doesn't affect all-cause mortality in a statistically significant way for 44000 random adults, the health benefits probably aren't huge". That is not what he said.
In this case, the tweet isn't even true on its face. He writes "The pivotal clinical trial for the @pfizer #Covid vaccine shows it does nothing to reduce the overall risk of death". But the trial doesn't show the vaccine does nothing; it shows that the ratio of death rates is probably between 0.5 and 2. That is very different from "shows it does nothing". (caveat: I haven't calculated those numbers carefully. But I think they're in the right ballpark.)
These are all good points and I especially appreciate the tone.
Hi Rajiv, interesting and balanced points. My own research leads me to discount a bit more these prediction markets' prices. We found that political markets are more biased toward 50% than other markets and this is compatible with a pressure of traders motivated by non-monetary incentives (partisans or manipulators). Increasing the demand of contracts with low prices is relatively cheap. So, the same amount of "motivated" money on both sides of the market can push the price up. The result is in "Do prediction markets produce well‐calibrated probability forecasts?" (EJ 2013).
Why would you care what a study says when we have reams of data from the last 3 years?? The vaccines were effective in mitigating severity and appear safe as of today. Hawaii shows a population didn’t need robust natural immunity to get to herd immunity. We are now in the endemic phase and the notion my getting vaccinated should have protected someone else is a red herring because it’s just a dumb notion.
Because it's placebo-controlled. Observational studies can't answer these questions definitively. You talk about Hawaii and RFK talks about Nigeria with minimal vaccination and low excess deaths. There is no doubt in my mind that the original vaccine prevented a lot of deaths in certain age and comorbidity groups, but less so for healthy young subjects, and even less so when it comes to the boosters.
In America age is a much bigger factor for a population on the low end than the high end because in general America is old and unhealthy. So populations like Utah and Travis County and DC would have had low Covid death rates even if they had done very little with respect to mitigation measures. All things being equal % below poverty level is the biggest factor for a population in America because poverty is a proxy for many public health issues and poverty ages people prematurely and our anti-poverty programs are directed at mothers with babies, i.e. young people.
Exactly my point, such confounding factors mean that there are limits to what you can learn from observational data, which is why placebo-controlled RCTs are so important, and there are not many of these, even the Pfizer study discussed in the post was unblinded after six months.
I was able to spot the Tennessee anomaly using my analysis…nobody that I know of spotted that but me. So according to my analysis Kentucky should have had a lower Covid death rate than TN because the Democratic governor was pro-masking but TN had the lower Covid death rate…until one day TN added thousands of Covid deaths to the official tally. I pointed out the anomaly and then shortly after it was no longer the case.
I agree there is tension here between what is pragmatic/effective with respect to countering his views during the campaign and his established pattern of harmful posturing on the vaccine front. Flaming and ridiculing him without counterargument during this campaign will probably backfire and may even garner sympathy for him among those who aren't familiar with his history of dubious advocacy.
On the other hand, for those who've been involved in countering his bad-faith arguments about vaccines and other issues for almost two decades, the reaction to his candidacy is anything but knee-jerk. While I've soured on him over the years, I think Sam Harris concisely describes why RFK Jr. is generally untrustworthy: https://www.samharris.org/podcasts/making-sense-episodes/325-a-few-thoughts-about-rfk-jr. Corporate influence in the FDA, errors on the part of the public health establishment—those are legitimate concerns better articulated by messengers other than RFK Jr.
The Science-Based Medicine blog, for instance, has been posting regularly about RFK Jr.'s activities since 2008. SBM uses a lot of derogatory language to describe him, perhaps unwisely, as doing so can come off as unmeasured when contrasted with Prasad's seemingly measured tone. But it's worth noting that Prasad himself has made a number of ludicrous and unmeasured statements, such as claiming that masking policies are akin to faith in "witchcraft". (Whatever one's opinion on the effectiveness of masking, I hope this statement by Prasad would be viewed as absurd.) Consequently, he may be too forgiving of RFK Jr. because he sees a potentially powerful ally in his own crusades against corporate regulatory influence and aspects of the coronavirus response, where he has not infrequently employed dubious reasoning about RCTs along with calibrated histrionics. Such debates are crucial, as we have much to learn from both the successes and failures of the coronavirus response, but they should be predicated on rigor of thought, not posturing.
I agree with much of this Jason, except that I don't think that RFK's arguments are in bad faith. He seems to hold his expressed views quite sincerely. And some of those views are not unreasonable, even if one disagrees with them (closing down most overseas bases for instance, or strictly enforcing border controls while expanding legal immigration). I have listened to him for many hours now and I think that he is worth listening to and engaging with, and on some issues could bring a healthy dose of dissent to the democratic party.
Prasad can be a bit over the top sometimes but I find it forgivable, given that there is still a lot of denial about policy errors during the pandemic. And I think he is especially upset about the masking of toddlers and the insistence that this was science-based.
Thank you for your measured and thoughtful comment.
The states that found the best balance between public health mitigation measures and a return to normalcy were the red states with Democratic governors. So Georgia could have had the same Covid death rate as North Carolina had it been a little more aggressive with mitigation measures AND the economy would not have suffered. So Georgia could have saved around 10,000 lives had it done what its neighbor did.
Do you have a reference for this?
Unfortunately Matt Taibbi didn’t appreciate me correcting the #fakenews that he reports and he got Substack to delete all of my posts AND all of my comments. So I had some good Substack posts covering this and Substack just incinerated them to protect their cash cow. Substack’s “moderation guidelines” miraculously appeared a few hours after they deleted my account. ;)
It is best to judge a candidate by how he thinks, not what he thinks. RFK Jr. does very poorly by this metric.