Along with most others, I watch the MMR/autism controversy with appalled fascination. I am grateful to have received the MMR vaccination. I am glad that I gave it to my children. I have seen no evidence for a correlation between the MMR vaccine and autism, and I know exactly why I would need to be skeptical should such a correlation be demonstrated.

Nonetheless, I feel that the public response to the MMR/autism controversy has been poorly handled. I read a lot of articles like this one, which basically take a paternalistic attitude toward anyone who doubts the current scientific or medical consensus. The anger-producing message of such articles is, “You’re not entitled to an opinion.” It’s a tone that will rally the troops, but harden the (ahem) opposing side.

And, in what is only a minor digression, such writing also does not show a lot of humility on the part of the medical establishment. These are the people, after all, who gave us thalidomide. They produce the pharmaceuticals that are withdrawn from the shelves after a decade because of serious side effects. These are the people who, in living memory, were sedating women during childbirth, and insisting on formula thereafter. To pretend that science and medicine can’t get it wrong ignores history, and insults the intelligence of anyone with even the most modest historical perspective. (As I’ll suggest below, a more productive approach would be to teach people how to ask responsibly, “How would we know if the medical community were getting this wrong?”)

But to return to my main argument, the problem with invoking the God of Science to solve this question is that science lacks the tools to solve the problem. This is true in two ways.

In the first place, scientists cannot prove the absence of an effect. We giggle about astrology because everyone knows that it’s bogus — but in fact science cannot prove that the stars (or the planets, or constellations, or whatever) don’t influence my life. A very clever experiment could show that such effects must be weaker than many other effects. But the complete absence of an effect cannot be proved. If everyone believed in astrology, people who doubted it would be hard pressed to construct a case against it, and indeed, it could only ever be a heuristic argument. The implication here is that science cannot say that the MMR vaccine doesn’t cause autism. It can say that it looked very hard and didn’t find anything, but that’s it.

The entrenchment of the status quo is one of the weaknesses of the scientific endeavor, which makes it all the more of a miracle when the status quo is successfully challenged (cf. the work of Michael Polanyi and Thomas Kuhn). It’s frustrating for any scientist who embraces a minority position, but it becomes somewhat more serious when we have a hardened community of anti-vaccinators on our hands. Suddenly the inability to prove the absence of an effect ceases to be a theoretical difficulty and becomes a public health issue.

The second way in which science cannot solve this problem is that dying from measles and having autism are incommensurate quantities. I’m certain that dying is worse than being autistic, but I can’t quantify that. Every person–or, depending on one’s political views, every society–has to decide about how to balance those risks.

Let’s compare this to Ebola. If my son contracts Ebola, then I’m looking at perhaps 50-50 odds of losing him. If you offer me a treatment that has a 75% chance of curing him, but a 25% chance of killing him, then mathematically I must accept that treatment. When I compare the probabilities, the decision is almost made for me. But how would I compare a 1% chance of his dying with a 5% chance of his becoming autistic? The math suddenly doesn’t help me anymore, because those quantities are incommensurate. I need to weigh those risks for myself–or again, depending on one’s politics, have them weighed for me.

(Just to reinforce here: I’m not assuming a correlation. Even if the risk is hypothetical, the quantities are incommensurate.)

I don’t see either of these observations being made in the media, though they don’t rely on any particularly profound insights, nor on particularly controversial ones. I don’t expect to see them in the media, however, because the subtlety of the points would inevitably be lost on any but the most loyal readers of this blog. 😉

Here are some suggestions for going forward, for any high-ranking public health officials who might stumble across this post:

  • Adopt a more humble tone. Short of compulsory vaccination, the way through this is to woo the crazies, not to provoke them. Medical science has gotten it wrong before. And as I’ve written here, it can’t exactly crack this one. Help people to understand how to make responsible decisions, given the available evidence.
  • Continue popular-level education about the difference between correlation and causation. In the last twenty years vaccination rates have gone up and autism rates have gone up. How could that not be meaningful? The aforementioned web site explains how in a fun and winsome way.
  • Continue popular-level education about why anecdotal evidence often carries more weight than statistical evidence, and why that is problematic.
  • Continue popular-level education about herd immunity. I’ve no doubt that many of these people assume that their children are safe because everyone else is vaccinating.
  • Explain the science rather than the sociology. Hearing that the author of the 1998 Lancet article has been tarred and feathered by the scientific establishment makes it sound like a conspiracy. Explain why his work was flawed.

I get that basic science education is not the easiest or most effective way to do public health. Public health campaigns rely on the techniques of advertising, which in turn rely on the (deeply problematic) ways that people usually make decisions. But if I’m a parent encountering personal testimony after personal testimony about the ill effect of vaccination, and if I’m further shown graphs demonstrating correlations replete with r-squared values, then an advertising-type campaign is not going to change my mind. The only way out of this one, so far as I can see, is through reasoned discussion.