Jim Manzi Disappoints on the Devastation of Lead and Crime

And Kevin Drum disappoints in his own defense.

Regular readers will know that (at least sometimes) I like to seek out and highlight the very best, most cogent and convincing arguments countering my beliefs, trying to figure out what might be wrong with those beliefs, and hoping to understand what’s really going with the subject I’m considering.

Jim Manzi — a guy with a decidedly conservative/libertarian bent — has often been my go-to guy on that front. His thinking about global-warming mitigation, for instance, doesn’t try to deny the IPCC conclusions. He accepts and adopts them, building on them for further thinking and calculation about the costs and benefits of mitigating the undeniable — and undeniably human-caused — climate changes that we’re facing. (I’m not saying his reasoning or conclusions there are flawless, by any means. But at least he starts by accepting a large dose of reality that many or most among his conservative cohort prefer to deny, distort, or just blithely dismiss.)

But in his recent response to Kevin Drum’s article on lead and crime (a subject I’ve gone on about), Jim disappoints me utterly. (And Kevin’s response — which limits itself to answering Jim’s cherry-picked objections to one bit of evidence without pointing out all the evidence that Jim inexplicably ignores — is decidedly limp-wristed.)

Short story, Jim looks at one study (Reyes 2007) that Drum talks about, and pokes procedural holes in its analysis and conclusions. “There might be confounding variables! She should have controlled for X!” I can’t tell right off whether he actually looked at the paper, or just responded to Kevin’s description of it.

Well okay, yeah, the Reyes study is in no way definitive. No single study ever is. He’s absolutely right that the Reyes study, all by itself, “is way short of making a convincing case for spending $400 billion of taxpayer money.” No duh.

Which points out the real problem: Jim doesn’t seem to have even looked at any of the the (other) research available on the subject (at least he doesn’t mention it), notably the body of work — robustly highlighted in Kevin’s article — by Rick Nevin. (Meanwhile Jim refers to Reyes’ work as Kevin’s “key econometric source” — which it isn’t; it’s just one of them, and not the strongest source, which is what I would expect Jim to address.) Nevins shows correlations between lead and crime across nine countries over many decades. It’s pretty convincing stuff that you’d think Jim would have looked at and considered if he really wanted to understand what was going on.

Even more surprising: Jim’s main question is whether it’s worth spending $20 billion a year over 20 years (which he presents as an undiscounted figure of “$400 billion”) to mitigate lead in the environment. But he doesn’t seem to have looked at or considered the paper that analyzes exactly that question, in the very terms he prefers when making his own arguments (Zahran, Mielke, et. al. 2010).

I can point out many other problems with Jim’s response.

• When discussing what evidence that might support the lead/crime hyphothesis, he fails to note that a higher concentration of lead in the blood of convicted criminal as adults would be rather convincing.

• He points out that Reyes doesn’t find a correlation between environmental lead and burglary, but he doesn’t note that Nevin finds (because he apparently hasn’t even looked at Nevin) a significant correlation between blood lead and that conservative bogeyman, unwed pregnancy.

These are niggling (though important) details. But they’re representative. There’s a lot know out there, and Jim doesn’t seem to have any interest in knowing it — only in shooting (certainly, potentially, valid) holes in a single study.

I find in Jim’s response, much to my disappointment, a prima facie case for confirmation bias — a small-minded and unconvincing effort to undercut and dismiss information that strongly suggests conclusions contrary to core conservative and libertarian beliefs:

1. Expert research can provide more useful and actionable information than market prices, and

2. Government action — including quite restrictive regulation — can provide hugely positive returns on investment.

He’s making another (in this case, dispiritingly lame) stab at the argument that’s lurking right at the heart of his book Uncontrolled: because the world is so complex,  whatever we’re doing right now must, obviously and inevitably, be the best of all possible worlds, and absent massively overwhelming evidence that’s universally agreed upon, we shouldn’t do anything different because we just don’t know enough. We should instead rely on market forces and the price mechanism — the only capable social calculator — to solve any problems.

Jim will deny he believes that, or hews to that belief dogmatically, but I believe this lead/crime response, with its obvious confirmation bias, strongly suggests otherwise.

I won’t argue all the merits of the lead-abatement issue here, or the evidence supporting it. I will simply state what I’ve concluded by looking at lots of different research on the subject, sifting it, and steadily adjusting my priors based on my evaluation of the how convincing that research is:

1. Lead in the environment (resulting largely from the auto/gas industry’s profit motives) has cost us trillions of dollars in measurable economic damage over the decades, and has had a devastating, unmeasurably destructive effect on hundreds of millions of lives. (Lifetime income doesn’t really quite capture or measure the value of a robust life endowed with one’s full, inherited endowment of cognitive, emotional, and social skills…)

2.  That damage continues, due to the residue of lead in buildings and dusty soil that is still poisoning our children, constantly, every year.

3. A mitigation effort as suggested in the literature on the subject would deliver an excellent, really massive, return on investment before even considering the unmeasurable manifest benefits to humans whose brains would not be stunted and retarded by mitigatable poisoning.

Mitigatable through collective action, and only through collective action. Which means: government action that is often the only counter to destructive market forces.

Cross-posted at Angry Bear.





, ,




3 responses to “Jim Manzi Disappoints on the Devastation of Lead and Crime”

  1. Olav Martin Kvern Avatar
    Olav Martin Kvern

    …but you’re weird. When you are presented facts and/or arguments that contradict your current ideological framework, you *change your mind*. I point this out because it makes you (and me, I hope) unusual. Actually changing your mind due to new information is rare.

    There’s probably some good research to be done on why this happens. I’d expect it would tie into Kahneman and Tversky’s research–re-evaluating previous positions requires a bunch of that expensive slow thinking. Easier to just return to ideological certainties.

    I also have to point out that applying cost/benefit analysis to things like childhood lead poisoning too often depends on it not being *your* child that ends up a criminal, not being *your* salt-water-flooded Bangladeshi farm land, and so on. One would hope that empathy–or, at least, big-picture thinking–would have some role. I don’t know Manzi’s work very well, but I’m reminded of Bjorn Lomborg.



  2. Asymptosis Avatar

    “not being *your* salt-water-flooded Bangladeshi farm land”

    How To Destroy (Almost) Half the Planet for the Low, Low Price of Just 5% of Global GDP


  3. Rick Nevin Avatar

    Great post – thanks. I have also posted a response to Manzi and others at http://www.ricknevin.com – including links to a rigorous cost-benefit analysis for how we can eliminate the remaining risk of lead paint hazards.