Science, drugs, policy and Hume

In this blog last week Anders Sandberg discussed the widely criticised sacking of Professor David Nutt from the government’s advisory council on the misuse of drugs. Professor Nutt had openly criticised government policy, in particular the decision by government to change the classification of cannabis and ecstasy against the advice of the government’s scientific experts. The government claimed that it was not his job to enter the political fray.

In defence of the government, some commentators have sought to distinguish the role of scientists and the role of the politicians. Expert scientists are there to synthesise analyse and present evidence. But politicians have to ultimately decide policy on the basis of values and ethical judgements. Ben Goldacre, writing in the Guardian this weekend replies to this argument by presenting further empirical evidence that is open to scientific enquiry – evidence about the effects of different drug policies on drug use, for example. Goldacre’s argument is that policy questions, as well as medical questions can be assessed scientifically. On this basis, then, perhaps the correct drugs policy can be determined by scientists?

But this conclusion is vulnerable to a problem identified by the 18th century philosopher David Hume, the famous is-ought problem. The ‘scientific policy argument’ appears to derive a normative conclusion (what we ought to do, which drugs policy we ought to embrace) from factual statements (which drugs policy leads to a reduction in drug use). But as Hume, and many philosophers have argued, it is not possible to derive a normative conclusion from strictly factual premises.

Imagine if scientific assessment reveals that drug policy A leads to a reduction in drug use within the population compared to drug policy B. Is it the case that we ought  to embrace drug policy A? That can only be the case if there is no other difference between the policies and we have agreed that we should adopt the policy that leads to the lowest overall drug use. (This is the necessary normative statement in our argument) If, as is likely, there are multiple different consequences of any individual drug policy, then we will have to decide which consequences to assess and how to weigh them against each other. Drug policy A may be better in these ways, and worse in others. 

If we have agreed which consequences are relevant, and in what ways to our decision, then a scientist may be able to advise which policy ought to be adopted. Perhaps that is indeed what Professor Nutt and his council did? But the weighing up of different considerations, and the reconciling of different values is not a scientific skill. There is no in-principle reason why scientists would be any better at it than politicians. And one advantage of having such decisions made by politicians is that this is what they have been elected to do, at least partly on the basis of the values that they espouse and their purported ability to reflect these values in policy.

This is not to defend the sacking of Professor Nutt, nor is it to defend the current government’s drugs policy. There are good reasons to criticise both. But

1. even if they agree on the facts politicians and scientists often differ in their values. (fact)
2. Policy decisions should reflect both facts and values (ought)
3. Politicians should sometimes make decisions that contradict the advice of their scientific experts (ought)

Comments

[Anibal]

The leap from an "is" to an "ought" is actually surmountable, and i believe, as the late Susan Hurley, in "cognitive politics": where politics is nevermore a matter of beliefs or subjective opinions but based on scientific rationale.

In recent years have been made numerous attempts to conjugate evolutionary biology, aristotelian philosophy, cognitive science, quinean philosophy... to erase, what i believe, is the false dicothomy between "values" and "facts".

One example, is William D. Casebeer´s "Natural Ethical Facts" (others, are Hilary Putnam´s "The Collapse of the Fact/Value Dichotomy and Other Essays")

In Casebeer´s book one simple conclusion arise to the issue at discussion here: morality can be justified from biological facts and is possible to derive a normative judgement from a descriptive judgement.

Politics can be objectivized. Politicians are not arcane oracles with mysterious powers. They are human beings with nervous sytems which create minds with preferences and motivations and other systems, including morals. The buzzword "neuropolitics" is turning a reality.