This is part of a series

Mar 22, 2018

Science vs. Politics: Hume’s Guillotine

Setting up her discussion of Snowdon’s Killjoys and Leyonhjelm’s Freedom’s Salesman, Dale invokes Hume’s principle that one cannot derive an “ought” from an “is.”

When I first had dealings with government (on a brief civil service secondment), I was told, bluntly, by a senior colleague that scientists should for the most part be kept out of legislation and policy development because they were “narrowly clever and broadly stupid.” I’d suggested using some expert scientific evidence to resolve certain legal questions about transgender persons; my boss was, rightly, horrified at the thought.

I’m used to people in some right-leaning political traditions disdaining expertise—Conservative minister Michael Gove famously did so during the Brexit referendum in 2016—but my civil service boss was a social democrat, and my office contained a mix of views from across the British political spectrum: everything from Greens to Nationalists to Tories to LibDems to Labour.

Like all lawyers, I knew that law is a system of norms that tells people how they ought to live. It’s a branch of moral philosophy, like theology, and while there is no necessary connection between law and morality, enacting laws with which large numbers of people disagree can produce everything from civil disobedience (conscientious objection) to violent non-compliance (Prohibition, War on Drugs, backyard abortions) to armed insurrection (pick a revolution, any revolution). Needless to say, being brought into disrespect in this way isn’t great for the law, either.

However, I was unaware—until doing “government work”—of the extent to which scientists often insist on “evidence-based policy.” If science proved something, the argument went, certain laws and policies automatically followed. Nowhere was this thinking more common than in the field of public health, although it cropped up elsewhere, too, especially among environmentalists. I witnessed heated arguments between public health advocates and civil service lawyers.

The scientists’ arguments were striking (and sometimes disturbing) because they attempted to dodge around Hume’s Guillotine. The Hume of whom I speak is, of course, David, the great Scottish Enlightenment philosopher and historian, and his “guillotine” runs like so: it’s impossible to infer a moral claim (what ought to be) from a scientific fact (what is). The distinction between facts and values—between “is” and “ought”—is so routinely collapsed in public debate that it’s necessary to provide an illustration of the problem. Consider the following argument:

  1. Human beings reproduce by having children (fact).
  2. Only humans with wombs are capable of bearing children (fact).
  3. Jane has a womb (fact).
  4. Therefore, Jane ought to have children (value).

What I’ll do in this piece—and in the one following it—is sketch out the relationship between scientists on one hand and politicians and policy-makers on the other. In the second piece, I’ll provide some worked examples of the ethical load-bearing politicians and policy-makers do, and how they respond to scientific demands for “evidence-based policy.” I’ll also discuss two books, one by a politician, and one by a policy wonk. The politician’s book is Senator David Leyonhjelm’s Freedom’s Salesman. The policy wonk’s book is Chris Snowdon’s Killjoys.

The object of politics—and the province of elected politicians and their senior advisers and senior civil servants, to the exclusion of all others, including scientists—is to make ethical choices on policy where reasonable people can (and do) disagree. This disagreement is usually expressed in the form of elections and referendums.

Of course, science can clarify the nature of the choices facing us. Environmental science, for example, tells us that solid fuels result in more local deaths from particulates than result from other energy sources. Meanwhile, economics tells us the clearing price of solid fuels is often much lower than alternatives. That introduces the question of how far the state’s legitimate interest in public health extends when intervening in markets. The welfare good of cheap energy, especially for the poor, must be balanced against the public health risks of solid fuels.

Here, science can introduce honesty to moral choices: it becomes more difficult to call up ideological abstractions (free markets, equality, social justice) or other heuristics on which we rely in the absence of complete data.

However, it is rarely the case that scientific knowledge can be combined with a widely accepted moral principle to produce a single, uncontested policy prescription. And it is the tendency to undervalue the moral heavy lifting of politicians and civil servants—to assume if only facts were clear and proven, playground morals would show us the way—that leads scientists to try to dodge Hume’s Guillotine.

Often scientists do not appreciate that as soon as they make policy prescriptions—thereby entering world of “oughts”—an ethical claim is in play, a claim as value-laden as any made by a Christian or Muslim about, say, the wrongness of homosexuality, or by a feminist about fighting patriarchy even in developed countries, or a vegetarian about the cruelty of eating meat. That the scientist knows a truth about the world, an “is,” that the monotheist, feminist, or vegetarian likely doesn’t is irrelevant. Moral claims have to be undergirded with moral reasoning.

In a now infamous VICE interview, clinical psychologist Jordan Peterson pointed out that because wearing lipstick and high heels developed as sexual signals, women in the workplace ought not wear them. Their presence, Peterson suggested, may even mean “men and women can’t work together.”

I am not an evolutionary biologist or a psychologist or any other sort of scientist. I don’t have the expertise to evaluate Peterson’s “is” claim. However, even if his evolved-sexual-signals claim is true in every particular, his proposal as to what people ought to wear in the workplace is, well, about as useful in policy terms as the conservative Islamist claim that heterosexual married couples should refrain from period sex because menstruation is yucky.

By the same token, scientific truths should not be devalued. Scientific information is a form of specialist knowledge that politicians (and indeed most people) do not have. It may be imperfect, but part of living in a modern society is developing the social trust needed to accept the skills of those who have a cognitive advantage over you, even if the person holding the knowledge is otherwise not as clever or capable. I have watched counsel reduce expert witnesses from many disciplines to intellectual rubble because forensic argument is a cognitive capacity peculiar to the law. It does not mean a probing lawyer with a flair for cross-examination has suddenly become expert in an unrelated field.

This means climate change is real and man-made. GM foods and vaccines are safe. Intelligence is largely heritable. The gender “wage gap” in developed liberal democracies cannot be attributed to discrimination in the labor market. I accept these things despite not having the specialised knowledge to justify or test them. I trust that our scientific institutions and the scholars in them are self-correcting at least to the point that major errors causing real world harms are rare.

The other issue of which scientists are typically unaware is that hierarchy and non-spontaneous order often work in science but fail elsewhere. Soviet and Nazi research, when not infected with Lysenkoism or “German Physics” or similar cobblers, often gave Allied scientists a run for their money. There is a reason the US pinched Nazi Germany’s rocket scientists. There is a reason Sputnik was first in space. Authoritarian states can produce excellent science.

It was only Soviet failure in particular to translate scientific advances into products enhancing quality of life for the masses that meant the US and UK took the lead. A case study: Communist authorities showed Grapes of Wrath as a propaganda exercise. However, Russian cinemagoers were amazed how in America people fled poverty in a car. In Soviet Russia, you hoofed it. The film was withdrawn.

The reason scientists make this basic mistake is they often don’t know that human systems—especially markets—do not respond well to hierarchical structures or central planning. There are exceptions, of course—like the military—but most people understand that going on yomps in the Brecon Beacons with the Royal Marines is not for everyone.

Some time before 1989, a Soviet official visiting the UK asked Oxford economist Paul Seabright who was in charge of London’s bread supply. Seabright gave him an answer that is comical but true: “nobody.” The bread we eat turns up on our tables thanks to an incredible team effort (bakers, machinists, electricity suppliers, distributors, etc.). Seabright’s answer, although true, is also dizzying, especially to someone used to operating within stable hierarchies.

So scientists who seek to apply the same mental heuristics to policy as they do in their own discipline can run into trouble. A good example of this is Britain’s indoor smoking ban, which—as Snowdon argues in Killjoys—was a case of doctors applying a pathological style of reasoning (diagnose the disease, isolate it from the body, starve it of resources) to policy prescriptions in a liberal society. Its main effect was to close a lot of pubs.

The proper role for doctors is to give advice within the confines of their field, and the proper role for politicians and policy wonks is to accept that advice and make one of a range of ethical choices based on it. You can do both of course: it’s just that a doctor who wants to decide on policy also needs to be a politician, and he needs to be good at it. And ditto for everyone else, really.

I’m happy putting anyone into office: actors, scientists, businesspeople, escorts, whatever. As long as they’re good at the task we give to politics alone, which is ethical reckoning. Politics is the art of telling other people how to live and getting away with it because the instructions are of a type to which we have freely consented at the ballot box.

More on Snowdon and Leyonhjelm’s books (in particular) in the second half of this piece.

This is part of a series