At yesterday’s Brookings/WOLA Congressional briefing on cannabis policy, I made my usual argument that (in rough numbers) 80% of the users of almost any drug use it moderately, take no harm from it, and do no harm to others, but that the other 20%, who use more than is good for them, account for 80% of the consumption and an even larger fraction of damage to themselves and others. My conclusion from that was the necessity of regulation, since the industry that sells the drug (or offers other potentially habit-forming services such as gambling) will always be financially dependent on dependent problem users, while the public interest is in serving the desires of non-dependent non-problem users while minimizing the number of dependent users.
Jonathan Rauch, who heads the Brookings side of the project, found that line of argument troubling. He asked me whether the interests of the responsible 80% should really have to yield to the interests of the irresponsible 20%. (Since the two groups aren’t distinguishable at a glance, there’s no way of restricting the consumption of problem users without somewhat inconveniencing non-problem users.)
That question, asked by someone whose intellect and ethical sensibility I have come to respect, led me to reflect on the difference between a moralistic or rights-based approach to a problem such as this one and a policy-analytic or outcomes-based approach. If you think of problem users and non-problem users as different people, it’s natural to ask which group’s interests ought to make way for the other’s. That seems to be a moral or constitutional question. But if you think of yourself as a potential user of a drug (or, as Jonathan suggested to me, the parent of a potential user), unable to know in advance whether your (or your child’s) use will remain controlled or will instead progress to dependency, and ask how much inconvenience in controlled use you want to sacrifice for protection against a bad habit, then you confront a practical problem rather than a moral one.
(Some readers will recognize in this Schelling’s solution to the puzzle of why it’s justified to save a larger rather than a smaller number of lives, when that’s the choice; if you imagine yourself as a member of one of the two groups, without knowing which one, it’s obvious you’d prefer a higher probability of survival to a lower one. Jonathan instead recognized this as a Rawlsian veil-of-ignorance argument, which also seems right to me.)
Of course, this same approach can be applied well beyond drug policy. Asking “How much do the non-poor owe to the poor?” is a moral question. Asking “How much protection would a reasonable person want against the risk of poverty?” sounds more like a computation. Of course, if you think of yourself as naturally immune to the risks of drug abuse or of poverty, you’ll be more inclined to let the drug abusers, and the poor, go hang. But that seems to me compatible neither with the Categorical Imperative nor with the Golden Rule. If we accept arguments from symmetry in physics, why not in ethics?