« Blurb-o-Licious | Main | The Shaming of John Thain »

Monday, January 26, 2009

A Question for Readers

Brian L. links to a paper by a group of law professors, writing through the Center for Progressive Reform, raising concerns about Cass Sunstein's likely role as the head of OIRA.  I was interested in this sentence, which the authors offer up in the course of criticizing Sunstein's skeptical views about the precautionary principle: "It is difficult to think of a single public health or environmental threat that with the benefit of additional research has not proven even more dangerous over time."

What do you think, readers?  Can you supply the authors with any examples?   

Posted by Paul Horwitz on January 26, 2009 at 08:28 PM in Paul Horwitz | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference A Question for Readers:


Note: I'm not seeing any other comments here, although there are apparently 19 comments.

Anyway, the article was rather anti-intellectual, I thought, both because it made facially ridiculous assertions such as the one Paul quotes, and because its criticisms of the very notion of cost-benefit analysis amount to disbelief in the very possibility of rational decisionmaking.

Posted by: Stuart Buck | Jan 29, 2009 4:10:06 PM

Two in today's NY Times alone:



Posted by: Bruce Boyden | Jan 27, 2009 11:57:57 PM

Cell phones by your head (I hope)

Posted by: Scott Moss | Jan 27, 2009 10:44:49 PM

The claim is absurd, and I am surprised (and disappointed) that the authors of the report would have made it.

The most obvious example would be pesticide residues in food which some used to believe were a significant source of cancer. The National Academy of Sciences fairly conclusively rejected this view. A related concern would be the threat of DDT to human health (as opposed to its effect on bird species). While Rachel Carson and others made various claims about DDT's effects on people, they were never substantiated. Another biggie would be the alleged threats posed by agricultural biotechnology. There have been all sorts of claims of potential human health risks, none of which have been substantiated.

Jonathan H. Adler

Posted by: Jonathan H. Adler | Jan 27, 2009 12:47:10 PM

Sarah: you are kind of right, but they are not wrong; you can't assume the perspective of a unitary actor. There are lots of species of the principle, but the one I see in the int'l environmental context has that where there's a threat of very serious damage, we should take preventive measures (qualified as cost-effective measures!) to avoid it, even if the scientific consensus is still out (particularly on the causal relationship). Even if the authors are right -- that every threat perceived at time time A is shown by research at time B to have been genuine, so all the more reason to have the courage of our noncourageous convictions -- they can't move the consensus by fiat, but only hope to trick the body politic into acting in the right way on the basis of an inexact account. They know; we do not.

Of course, it remains a possibility that the evidence at time B is confirmatory only because we haven't acted and so permitted the risk to be taken; perhaps when we do act prophylactically we not only impair the production of evidence, but also make the risk vanish, such that we never should have acted. Fortuna works in mysterious ways.

P.S. I would note that the authors merely say that "it is difficult to think of a single" threat; they may have thought of many, eventually, or might be saying that it is hard to come up with only one. This is the Lay's potato chip problem in risk assessment.

Posted by: Ed | Jan 27, 2009 9:31:49 AM

Sarah L. : Per the original post, they are "criticizing Sunstein's skeptical views about the precautionary principle." That is-- they support the precautionary principle, as you note.

I find it ironic that Sunstein's argument about the paralyzing effects of the precautionary principle are summed up perfectly in the implications of their claim that "[i]t is difficult to think of a single public health or environmental threat that with the benefit of additional research has not proven even more dangerous over time."

That is sloughing off the evidentiary burden, which is even more dangerously paralyzing. Imagine if the PP were LITERALLY to give the benefit-of-the-doubt to allegations of risk. Sunstein wouldn't even have to pick up a pen and talk about the POSSIBILITY of paralysis-- it would be the order of the day!

Posted by: AndyK | Jan 27, 2009 9:28:29 AM

Wait a second...their claim isn't just silly--unless I'm misunderstanding the precautionary principle, there's something deeply wrong with their claim. The precautionary principle, at least as I understand it, says something like this:

Action A will result in a bad outcome with a low probability (or, in an alternate version, we have uncertainty about whether it will result in a bad outcome). But the outcome is very bad. So we shouldn't take Action A.

The sentence you've quoted actually rejects a premise of the precautionary principle, not the precautionary principle itself (unless the authors are the worst Bayesian updaters ever). The sentence you've quoted says, look, every time we have thought that something might be bad, it was bad. So the sentence is saying that Action A will result in a bad outcome with high probability (or with little uncertainty). So of course we should reject Action A! We are certain that it has a bad outcome!

So it seems to me that they're not making an argument against the precautionary principle--they're just fighting the hypo.

Am I misunderstanding either the precautionary principle or their argument? (More importantly, am I going to miss the bus to work because I wrote this comment? The probability of that is increasing with every minute I keep writing...and unfortunately with a high level of certainty...)

Posted by: Sarah L. | Jan 27, 2009 7:11:29 AM

Nuclear power plants.

Posted by: anon | Jan 27, 2009 12:44:00 AM

c-sections, breast implants...

Posted by: p.d. | Jan 27, 2009 12:09:57 AM

Alar, satanism in daycare centers, bad humors, miasma, genetically modified food, marijuana use, Pop Rocks, razor blades in Halloween apples, unburied dead after an earthquake.

Posted by: Bruce Boyden | Jan 26, 2009 11:47:26 PM

ICBMs. And, apparently, crack babies: http://www.nytimes.com/2009/01/27/health/27coca.html?8dpc.

Just to be clear, though, I think it's prudent to avoid any combination of the above risks, particularly Swamp Thing and ICBMs.

Posted by: Edward T. Swaine | Jan 26, 2009 11:22:16 PM

1. Foods that studies show to be bad (often contradicted, or at least shown to not be so dangerous)
2. Malthus's population concerns
3. So many other things that it's crazy to write a sentence like that.

Posted by: Scott | Jan 26, 2009 11:16:52 PM

Alar comes to mind. Flouride in drinking water, too. Oh, and swine flu. Definitely swine flu.

Posted by: anon | Jan 26, 2009 10:30:56 PM

Polio, smallpox, heart disease, just to name three more.

Posted by: David Levine | Jan 26, 2009 10:20:15 PM


Microwave ovens.

Going outside in the cold without a hat on.

Posted by: Sarah L. | Jan 26, 2009 9:26:06 PM

The framing of the question seems designed to work on cognitive biases. When we think of a "public health or environmental threat," we'll usually gravitate toward things that we now see as... threats, not the scares that passed by the wayside.

Posted by: Anyglen | Jan 26, 2009 9:00:58 PM

Demonic possession.

Posted by: James Grimmelmann | Jan 26, 2009 8:56:10 PM

The Swamp Thing.

Posted by: James Grimmelmann | Jan 26, 2009 8:50:01 PM


Posted by: James Grimmelmann | Jan 26, 2009 8:48:59 PM

Does y2k count?

Posted by: Paul Gowder | Jan 26, 2009 8:41:17 PM

The comments to this entry are closed.