The two Marks created many Euros of value.
Over at Slate Star Codex, psychiatrist Scott Alexander has an ongoing set of debates. It’s called the Adversarial Collaboration Contests. The idea is to have each person start by taking a side that he/she believes in (adversarial) but also have them work together to try to reach a conclusion (collaboration.) In his ground rules, Scott explains:
Remember, an adversarial collaboration is where two people with opposite views on a controversial issue work together to present a unified summary of the evidence and its implications. In theory it’s a good way to make sure you hear the strongest arguments and counterarguments for both sides – like hearing a debate between experts, except all the debate and rhetoric and disagreement have already been done by the time you start reading, so you’re just left with the end result.
The third one in the series was posted yesterday and I found it quite impressive. (I haven’t read the first one because I was not that interested and I haven’t got around to the second one.)
Why am I impressed? For three main reasons:
- I like civil discussion where each side is actually trying to get to the truth. That minimizes the amount of game-playing, the number of gotchas, and the various other distractions
- In trying to get to the truth, each side seemed to learn from the other.
- I learned a lot about the facts of vaccination from the discussion.
As is my wont, I did find one omission somewhat troubling.
First, some highlights of the positives before I get to the omission.
The main goal of Mark Davis, a naturopathic doctor, “was to argue that there is enough reasonable doubt that routine childhood vaccines could contribute to hyper-inflammatory disease, and enough reduced harm from vaccine-preventable diseases from other medical and public health interventions (in countries with greater economic resources) that parents should be given wide latitude to make individual choices re: routine childhood vaccines despite the clear benefits to individual and public health from preventing those diseases.” And the outcome? “He became more convinced in his conversations with Mark Webb that widespread childhood vaccination is in the best interest of public health.”
The main goal of Mark Webb, a clinical researcher with a PhD in immunology, “was to argue that atopy and autoimmunity are likely not driven by vaccination, and that this idea is a distraction from finding the real causes of the increase in these diseases.” The outcome? “He became persuaded that policy should not just reflect the best evidence currently available, but should also reflect a certain degree of humility that there will always be something we don’t know.”
So based on that, I would say that Mark Davis moved closer to Mark Webb’s position.
Some of the things I learned (and I bet they learned) along the way are:
The differences between the United States and some other developed countries:
the argument, “vaccination should be mandatory” must contend with one uncomfortable fact: in many European nations vaccination isn’t mandatory, and those nations often achieve higher vaccination rates than in the US.
How tiny the known risks are:
For vaccination, the most common, well-documented, known risk is the potential for an allergic response to some component of the vaccine. The most common allergic component is egg, and people with severe egg allergies are instructed to consult their physician prior to vaccination. How common are allergies to vaccines? A good estimate is about 3 per one million doses. This would be the equivalent of about 200 people in France, or 35 in the US state of Ohio. This is the biggest recognized, known risk of vaccines.
The hygiene hypothesis (which they probably knew but was new to me):
the hygiene hypothesis is the idea that some of the bacterial and parasitic infections that modern medical technology has eliminated might have been performing an important function in the human immune system – and when you take them away you start seeing problems.
And some evidence for the hygiene hypothesis:
This hypothesis isn’t just high-level theoretical hand-waving. Parasites, such as hookworm, have been shown to induce the same kind of immune mediators that are commonly seen in autoimmune diseases. In fact, some people with diverse autoimmune, allergic and autoinflammatory conditions have started intentionally infecting themselves with hookworm. Based in part on this movement, clinical trials have been conducted, and more are currently under way investigating whether re-introducing parasitic infections such as hookworm can be used to treat Crohn’s and other autoimmune diseases.
The Amish and Hutterites:
A group of researchers identified a different German religious sect, called the Hutterites, which also engages in regular farming. They are closely related to the Amish, and came from similar parts Germany at similar times. However the norms of the Hutterites dictate a much lower interaction between livestock and children/pregnant women (most hygiene hypothesis evidence suggests prenatal, neonatal, up through young childhood exposure is the critical exposure period). This genetically similar group, who had less early-life exposure to the farm environment than the Amish, gets asthma at a rate 5-7 times higher than Amish farmers (Amish asthma rate is 2-3%, Hutterite asthma rate is 15%). The researchers then took dust from the Amish barns and forced mice to breathe it in. They found the dust protected the mice from developing an experimentally-induced allergic response, and that it caused real, measurable changes in the immune systems of these mice, consistent with what we see in humans who are less allergic, and consistent with the differences they saw in the Amish farmers who had low allergic disease compared to the Hutterites.
Their bottom-line policy conclusion:
We believe this coordination problem may be largely resolved without restricting individual freedom. Individuals who wish to invoke the precautionary principle for themselves and their families should not be penalized for doing so.
As is often true, many of the comments on the post are valuable also.
The one omission: discussion of the possible link between vaccination and autism. Here’s all they write about it:
Much ink has been spilled about vaccines and autism, and it is not our intent to cover that ground again here. Both authors agreed that the evidence does not support a link between vaccines and autism.
I would have liked to see why they think the evidence doesn’t support a link. I’m totally willing to believe that it doesn’t, and, indeed, my prior is that it doesn’t. The evidence convinced both of them. Why? What was the strongest evidence that convinced them? This inquiring mind wants to know.
READER COMMENTS
Denver
Sep 7 2018 at 2:15pm
You weren’t missing much with that first entry. The terms weren’t clearly defined, and I don’t think the authors actually had substantial disagreement, so it didn’t read much like an adversarial piece.
The next two were better, though I agree that this vaccine one has been the most interesting to me personally thusfar.
Dylan
Sep 7 2018 at 3:00pm
Thank you so much for introducing me to these. About a year ago I was trying to formulate an idea along these lines and had even written to Russ to suggest that EconTalk might try to do something similar. This of course appears to be much more fleshed out then my half baked idea was and I’m looking forward to reading the entries.
Phil
Sep 7 2018 at 3:13pm
Only one study ever found a link between autism and vaccinations (based on a whopping dataset of 12 children), a dozen other studies since have found no link. The first study has not only been soundly refuted by the fact there is no evidence supporting a link, it was actually unpublished by the journal it appeared in (The Lancet) because it was uncovered the data was falsified. As a result, the author, Andrew Wakefield, lost his medical license for professional misconduct. Yet, a belief persists.
David Henderson
Sep 7 2018 at 4:28pm
Thanks, Phil. Yes, I’m familiar with the Lancet article. I also had an econometrics colleague at the Naval Postgraduate School (“had” because, although he’s still there, I’m not) who told me that there were huge identification problems with the Lancet study.
AMW
Sep 7 2018 at 5:56pm
David,
If you have ~35 minutes the following series of short YouTube videos will be helpful:
https://www.youtube.com/watch?v=EXtANMp3wok&list=PLpQImBX6Re8JP4HPInBncwW_yQHWJAHtp
David Henderson
Sep 7 2018 at 7:18pm
Thanks, AMW. I started watching the first video you recommend. If the bottom line is that Wakefield was a snake-0il salesman, I have better uses of my time because I already believe that. That doesn’t handle my issue. See my response to Phil above.
Jeremy Arkes
Sep 7 2018 at 11:29pm
For 25 years, researchers (and Nobel Laureates) claimed that the hot-hand-in-basketball was a myth. Now, it turns out that there is strong evidence that it exists. The logical fallacy they made was that lack of evidence isn’t proof of non-existence. Using the correct logic, I’d argue the jury is still out on vaccines and autism … I haven’t looked in a while, but some of the vaccine-autism studies on the CDC website had pretty wide confidence intervals on the coefficient estimate that included large positive values. What really makes me wonder is:
(1) Why is there a vaccine-injury fund that (according to my understanding) pays out to some families of children who are autistic?
(2) What other story can explain the large increase in autistic children? I’m not saying there is no other story. Maybe Round-Up in more of our grains? Maybe we just didn’t diagnose autism cases as well before (I have a hard time believing this one). This question/statement that there is no other good explanation, of course, isn’t evidence for the vaccine-autism connection, but vaccines certainly present a theoretical possibility for the increase in autism.
So, Mark and Mark are right that there is no evidence that supports a link. But, that doesn’t mean there isn’t a link.
Alan Goldhammer
Sep 8 2018 at 10:00am
With the disclaimer that I once was involved with a research project on B. pertussis toxin action (whooping cough) and am solidly pro-vaccine (just got my new shingles vaccination and now need my flu shot), I note that there is overwhelming evidence against the anti-vaccine crowd. The best evidence comes from Europe as most nations have a government run health system and electronic medical records that allow epidemiologists to examine adverse reactions to both drugs and vaccines better than we can currently do in the US. All the data that I’ve seen shows that there are no unusual side effects from vaccines other than the allerginicity and Guillam-Barre syndrome ones.
The old whooping cough vaccine (pre-1983 or there abouts) was very problematic as it was a whole cell vaccine and did have an elevated risk of side effects (perhaps do to hyper activation of insuln in infants causing seizures). However, that vaccine was replaced by a sub-unit vaccine that is ‘almost’ as good at providing immunity and has an extremely good safety profile. The old whooping cough vaccine was taken off the market in Europe in the late 1970s if memory serves me correctly. Things were OK for a couple of years because of existing herd immunity. However, there was a major outbreak of whooping cough in infants who were not vaccinated accompanied by significant morbidity and some mortality (can’t remember the exact statistics). Parents started clamoring for vaccine but since all the manufacturers had exited the European market there was none to be had and they had to wait for the new version to be approved.
I tend not to get in debates against the anti-vaxers as they are pretty much locked into their beliefs. Paul Offit’s fine book, “Autism’s False Prophets: Bad Science, Risky Medicine, and the Search for a Cure” is perhaps the best work on the topic of autism and vaccine.
Vaccines are perhaps along with antibiotics the greatest public development of our time.
Rebecca Jaxon
Sep 8 2018 at 10:31am
If you’re interested in the autism issue, the best summary of the evidence can be found in Paul Offit’s book Autism’s False Prophets. It completely answered all those questions for me, a mother of an autistic child. I don’t blame those speakers for deciding beforehand not to cover that ground yet again. The autism research needs to be able to get beyond the vaccination debate (and vice versa). It has dominated the public discourse for years while our children continue to struggle without real answers. I recently attended a talk about autism by a genetics researcher who is doing truly exciting work and has identified several genes already that have a major role in the cause of autism and a couple hundred more that may have a minor role. (See the Spark study, if you’re interested.)
Jeremy Arkes
Sep 8 2018 at 3:53pm
As I’ve discussed with David, researchers on the hot-hand-in-basketball (and, Nobel Prize winners) claimed for 25 years that the hot hand was a myth. Now, it looks almost certain that the hot hand is real. These researchers and Nobel laureates made 2 mistakes: (1) they used the flawed logic that lack of evidence was evidence of there not being an effect; (2) they failed to recognize the inherent biases of measurement error, gambler’s fallacy (see Miller-Sanjurjo, Econometrica-forthcoming), and perhaps other biases (I’m guilty of this one also).
So, I do not consider Mark and Mark agreeing there is no evidence for a vaccine-autism link to be strong evidence one way or the other. From viewing the studies on this on the CDC website a few years ago, what I noticed was:
(1) A few studies had 95% confidence intervals that included what would be strong connections of a vaccine-autism link.
(2) To my recollection, any randomized study was very small, and no study examined the effects of the current ~37-shots-by-age-6 schedule there is.
I’d say the jury is still out, and it link remains theoretically quite plausible.
Alan Goldhammer
Sep 9 2018 at 8:57am
The cohort of children with autism has expanded over the years as the definition has changed. this makes it very difficult to do retrospective studies on those born prior to 1990. The best meta analysis of extant data is here. Unfortunately, it’s an Elsevier Press paper and gated. There is also a huge amount of data coming in to the FDA and their VAERs (Vaccine Adverse Event Reporting) database and thus far there has been no evidence at all regarding increased autism among vaccinated individuals. Whether something is ‘theoretically possible’ ultimately relies on acquired data. To date there is no evidence to support a vaccine-autism link.
JFA
Sep 9 2018 at 1:45pm
Can you point to the papers on the CDC website that suggested evidence of a strong connection? If those papers do exist, the CDC doesn’t agree with your assessment: https://www.cdc.gov/vaccinesafety/concerns/autism.html.
Also, the jury is generally out when there is contradictory evidence on the issue. For the autism-vaccine connection, there doesn’t seem to be any contradictory evidence of note. As the CDC website says, “Vaccines do not cause autism.”
Sometimes one should be agnostic on an issue. Rejection of the hot hand effect probably didn’t kill anyone. Rejection of vaccines based on a belief that vaccines cause autism certainly caused much suffering, and I would imagine that the probability of someone dying from their parents rejecting vaccines based on belief in the vaccine-autism connection is higher than the probability of the vaccine-autism being true.
I don’t know whether people should be forced to get vaccines, but it seems a bit overzealous to say “Look, scientists made a mistake in this other area, so we have to be agnostic about everything.”
JFA
Sep 10 2018 at 12:23pm
“I would have liked to see why they think the evidence doesn’t support a link. I’m totally willing to believe that it doesn’t, and, indeed, my prior is that it doesn’t. The evidence convinced both of them. Why? What was the strongest evidence that convinced them? This inquiring mind wants to know.”
I think this opens up an important philosophical issue: what explains status quo beliefs and what makes alternative hypothesis worthy of consideration? It seems that pretty much the only reason that people think there might be a link between vaccinations and autism was due to one fraudulent study (https://www.historyofvaccines.org/content/articles/do-vaccines-cause-autism). Vaccinations have been around for a while and vaccination rates have been pretty steady over time, but autism diagnoses has been increasing steadily over time. You would expect much less of an increase if there were actually a connection (you would actually expect very little increase). People’s access to mental health professionals has increased, so maybe there has been that underlying relationship and the increase in diagnosis is coming from increased access to mental health care. It is, however, very strange that the basic data trends of no increase in vaccination usage while autism diagnoses have increased don’t serve as someone’s prior on the likelihood of the vaccine-autism connection or seem to affect how they weigh the alternative hypothesis of vaccines cause autism. (While noting that number of vaccines on the vaccination schedule have increased, the MMR vaccine (the vaccine with a hypothesized link to autism in the Wakefield paper) has been on there for a while and thimerosal (another suspected culprit) is not used anymore in most childhood vaccines.)
Take this as an example. A new home is being built in the neighborhood. People suddenly find that they are getting flat tires due to nail punctures. This is openly discussed at a homeowners’ association meeting. One neighbor suggests that someone has been going around hammering nails into the tires at night. Rather than setting up a night watch group, the meeting concludes with a resolution to request that the road around the construction site be swept clean each night by the construction crew. Later, a homeowner who wasn’t at the meeting asks someone who attended the meeting “What was the strongest evidence against the theory that someone was going around hammering nails into tires at night?” Here we might wonder why that even needs explaining. Just because someone shouts out a hypothesis doesn’t mean it should be considered.
So David wants to know what convinced the two authors of the paper. I think it’s interesting that people think it needs explaining. I’m not saying David is wrong to want to know what convinced them, nor that it is unreasonable to request that info. I just think this is an interesting instance of how to think about when to categorize an alternative hypothesis worthy of consideration and how to allocate scarce epistemological resources.
Comments are closed.