A theme of my recent BloggingHeads with Yglesias (and some related IRL conversations) is his claim that people overemphasize the importance of differing political philosophies in driving pratical political conclusions, relative to straightforward empirical disagreements. I’m not sure how sharp a line can be drawn between these categories, since people often talk loosely about their political “philosophy” to mean a general worldview that includes mid-level beliefs about how the world works, including how people typically respond to incentives, how effectively we can predict the consequences of large-scale undertakings, and so on. Even many works of relatively abstract moral and political philosophy—which I think is the sort of thing Yglesias means to refer to—embed casual assumptions that some eminently contestable economic or sociological claim is just obviously true… though that may, in itself, be a point in Yglesias’ favor. (See Samuel Freeman’s book on Rawls for an extreme example of this.)

In general, I tend to think it’s the other way around. People who imagine themselves to be utterly practical and non-ideological are usually just doing their philosophy tacitly, which usually also means badly. Most often, the philosophy in question is a potpourri of mutually incompatible forms of consequentialism seasoned with unexamined, inherited cultural prejudices (dignified by the label “common sense”). But for self-conscious libertarians, who often do make a point of stressing a distinctive theory of rights as the basis of their policy views, I think there’s probably something to this.

So, for instance, libertarians tend to think they are distinguished from others by adherence to some version of the Harm Principle—the idea that individual liberty should be coercively regulated only when it entails harm to unconsenting third parties (as opposed to for paternalistic reasons). But even if libertarians are unusual in regarding this as a necessary condition of regulation, almost everyone regards it as among the clearest and strongest justifications for intervention, and so quite a lot of policy disagreements are nevertheless driven, not by dispute over the (unique) validity of the Harm Principle, but about which types of harms count (which sounds philosophical, but usually has an empirical component) and which activities are, empirically, harmful enough to restrict.

Drug prohibition may be the classic example of a “paternalistic” policy that libertarians reject on the grounds that we all have a right to go to hell in our own way. And certainly, there’s a strong paternalistic strain in most prohibitionist arguments. But it’s also worth looking at how much Drug War propaganda—especially the stuff that circulated aroundm the time of the passage of the Harrison Narcotics Act—involves exaggerated claims about how drugs almost invariably turn once-decent people into addiction-crazed thieves, rapists, and axe-murderers. Think Reefer Madness: It’s viewed strictly for laughs now, but around the time we were actually banning the stuff, many educated people’s beliefs about it were not much less ridiculous. And if it were really true that some drug reliably transformed almost everyone who used it into some berserk, violent beast, then (bracketing pragmatic concerns about black markets and so on) it’s not clear a libertarian could make a very strong moral case against prohibiting (or very strictly regulating) it. Less dramatically, there might be a range of views about which sorts of harms—such as being prevented by addiction from meeting the obligation to care for one’s children—would justify intervention if they did follow inevitably (or nearly inevitably) from drug use. Again, this isn’t to deny that there’s a powerful paternalistic component to prohibitionist arguments, but it’s a mistake to understimate how much work is done by (often confused) beliefs about the magnitude of third-party harms that would result from decriminalization or legalization.

Libertarians also tend to say that taxation to support a minimial government that enforces individual rights against violence, theft, and fraud is legitimate—but that taxation for redistributive purposes is not. But this doesn’t actually get you to a determinate set of policy positions without further empirical premises. Suppose, just for the sake of argument, that some amount of subsidy for education, or for job training programs, did more at some margin to reduce crime, at lower cost, than additional tax-supported policing and incarceration (in part by turning potential inmates from a tax-drain into part of the productive tax-base). It’s true, of course, that such programs are typically defended on other grounds—such as the obligation to help people for their sake, and not as an indirect form of self defense—but if we assume those facts, it’s not clear what a philosophical libertarian would have to say against these policies as a mechanism for rights protection. “No, tax me more for less effective cops and jails!” doesn’t sound like much of a rallying cry, and while there’s an obvious incentives argument against a policy of directly bribing people not to commit crimes, it’s doubtful it applies to education.

Obviously, someone with a libertarian view about what types of justification for state action are valid will admit a case for tax-supported programs of this sort under far narrower sets of circumstances than someone who thinks they’re justified by all sorts of benefits beyond preventing rights violations. The libertarian may also have a variety of views about economics or the dynamics of bureaucracy that make him less likely to believe such programs will actually be effective. The point is just that the policy view doesn’t fall straightforwardly out of a set of premises established by rights theory or moral philosophy. Empirical beliefs are doing a lot of work, both for those who accept a libertarian position and those who reject it. Sometimes the philosophical differences would lead people to different policy views even if we agreed on all the facts. But there are many fact patterns where the plausible moral views all (or mostly) converge on a common normative result, even if for different reasons.

Supposing that’s right, what follows? The bad news (for those who find this sort of thing congenial) is that this reduces the number of questions that can be decisively answered from an armchair by contemplating the nature of rights. The good news is that many disagreements that appear to be grounded in divergent and incompatible values are at least partly concealed empirical disagreements, and therefore perhaps less intractable than they sometimes seem.