Science and Law
posted by Gerard Magliocca
During the Bush Administration, there was a lot of concern expressed about how religious faith was distorting policy judgments on issues such as stem-cell research. I think these fears were overblown (with apologies to Andrew Sullivan). What surprises me, though, is that people do not talk much about the opposite problem — how science can distort the law.
There are two reasons why this should be of concern. First, science can be . . . er . . . wrong. It is more accurate to say that science is always incomplete, but sometimes that incompleteness can be so serious that the conclusions drawn are unreliable. Second, science cannot answer ethical issues. Something could well be scientifically valid but morally wrong. Here are a few examples:
At the turn of the twentieth century, there was a broad consensus that racism was scientifically valid. This finding was used as a justification for colonialism, segregation, the denial of voting rights, and all sorts of other practices. In effect, if you supported racial equality at that time, you were an “anti-science” person making decisions based on something irrational like faith. Later, of course, this science on race was discredited.
A few decades later, many serious scientists were doing research on eugenics. This was also embraced by policymakers who derided those who fought for disabled rights as “anti-science.” Holmes was the most notable example of this attitude in his opinion in Buck v. Bell upholding the sterilization of what he called “imbeciles.” The only dissenter in Buck (without opinion) was Justice Pierce Butler. Why did he dissent? Probably because he was a devout Roman Catholic. Did that make him an irrational “theocrat,” to use the modern phrase?
This brings me to the debate about global warming. While there is no faith-based objection to doing something about climate change, the question is what level of confidence should politicians have in that science before they take significant policy decisions? I don’t have a good answer to that, though my own conversations with people who study climate change suggest that that field of study is more incomplete than most. Consider the small sample size. The Earth is billions of years old, but we only have climate data over a period of thousands or maybe tens of thousands of years.
The lesson is that caution should be exercised when wrenching changes on energy or regulatory policy are based on current scientific trends. Sometimes the “anti-science” people are right.
May 29, 2009 at 7:52 am
Posted in: Jurisprudence
Print This Post
Responses (6)
Xanthippas - May 29, 2009 at 10:49 am
Okay, but it would be more helpful to mention concrete examples of where a scientific principle underlying some policy change has been discredited. Yes “scientific” racism and eugenics are examples, but they are old examples, and from a time when science in general was less advanced in its methodology and peer review was less rigorous. I think it’s fair to say that there may be more uncertainty in climate science than we believe, but climate science-the conclusions of which are broadly agreed upon by scientists the world over-is hardly in the same ballpark of outright falsehood as eugenics.
Gerard Magliocca - May 29, 2009 at 12:23 pm
What about the “food pyramid” based on the best nutritional science of the 1960s and 1970s? (Eat lots of carbohydrates). How’d that work out?
Mark McKenna - May 29, 2009 at 12:51 pm
“Holmes was the most notable example of this attitude in his opinion in Buck v. Bell upholding the sterilization of what he called “imbeciles.” The only dissenter in Buck (without opinion) was Justice Pierce Butler. Why did he dissent? Probably because he was a devout Roman Catholic.”
So you’re saying that Justice Butler was a lawless judicial activist who allowed his personal beliefs to influence his decisions?
Dave Roth - May 29, 2009 at 1:26 pm
It seems to me there’s a pretty enormous difference between your other examples and climate change. Specifically, the consequences of action or inaction are not symmetrical and this has implications for what level of scientific certainty we should require in order to make policy.
Take eugenics. You’re a policymaker at the turn of the century who believes the scientific “consensus” behind eugenics. If you act on this scientific “consensus” and sterilize a lot of people, and it is later revealed that the eugenics science is false, then you’ve done something enormously unjust and hurt a lot of people.
However, if you don’t adopt any policy measures in light of the science, and the science turns out to be true, then there doesn’t seem to be a huge amount of harm. I guess you’ve slightly delayed the evolution of the human race (or something; I’m not quite sure what eugenicists would say).
So in this case, it seems that policymakers should have required a much higher level of scientific certainty on the question of eugenics, since there are enormous harms if they act on the basis of this science and the science proves false, but much smaller harms if they don’t act on the science and it proves true. The consequences of acting and being wrong are much worse than the consequences of not acting and being wrong. So skepticism and caution seem justified.
But on climate change, the possible outcomes are quite different. If we act on climate change science and restrict emissions, and it turns out that the science is wrong, then we’ve slowed global growth by a few percentage points of GDP. That sucks, I guess.
But if we don’t act, and it turns out the science is right, well then we’ve got a huge disaster (if you believe the scientists’ projections).
So on climate change, the consequences of not acting and being wrong are much worse (at least so it seems to me) than acting and being wrong.
In light of this, I think policymakers should require less scientific certainty to take action than in, say, eugenics.
I don’t know what to say about your food pyramid example, since I can’t really speculate as to what harms were caused by adopting it in error, nor do I know what would’ve happened if we hadn’t adopted it. I think it’s safe to say, though, that the harms of adopting an erroneous food pyramid are probably much less than the harms of failing to act on climate science will be if that science turns out to be correct.
Matt - May 29, 2009 at 4:11 pm
At the turn of the twentieth century, there was a broad consensus that racism was scientifically valid.
This partly shows how hard this sort of thing is. Why this _might_ be true of a majority of researchers at the time, it certainly wasn’t the only respectable view. Franz Boas, for example, one of the founders of modern anthropology, was arguing quite explicitly at just this time that scientific racism was bunk. He may not have had the majority view, but he had a strong and important one. His view was made less influential for some time not for “scientific” reasons but more or less nakedly political ones- hatred of Germans and “leftists” by old WASP types in science, the academy, and politics at the time.
A.J. Sutter - May 29, 2009 at 8:47 pm
There was an even bigger problem during the Bush Administration than that of relgious faith distorting policy decisions: it was that the Administration’s religious proclivities, business interests, etc. were leading to censorship or suppression of scientific information on which policies might be based, or at least against which the policies could be evaluated.
You’re right that science doesn’t give guidance about morality. But I don’t think any of your examples show that “anti-science” people were right — only that some scientists have been wrong. Matt’s example is a good one (though apropos of the Boas case he could have added anti-Semitism to his list of hostilities).
Dave Roth is correct to distinguish the eugenics and racism issues (were these ever distinct?) from climate change. In this case, among others, the precautionary principle deserves more respect than it gets from the champions of cost-benefit analysis. It’s a flexible principle requiring judgment (rather than a rule, like CBA, based on a superficially objective algorithm that conceals judgments). In this it’s no different from most other practical judgments, especially the moral ones you correctly emphasize.
Moreover, the small sample size issue that you raise is specious. This issue is nothing new in daily life. E.g., all questions of political policy, and most decisions in business, are based on the need to make decisions in the face of new circumstances. Consequently, they rely more on Bayesian than frequentist foundations (or readers who are soi-pensant economists can use Frank Knight’s fuzzy distinction between risk and uncertainty if they prefer).
You were the guy who posted that terrific quote from Yes, Prime Minister, weren’t you? You seem to be forgetting it’s satire here.
Leave a Reply