Monday, December 1, 2008

Libertarianism: Through thick and thin

For some time now, I've been doing most of my blogging at the Western Standard. However, this debate between Timothy Sandefur, Jason Kuznicki, and Todd Seavey piqued my interest.

The debate is strictly intra-libertarian, which is why I decided to give it attention here rather than at the WS's blog. For a while now, some libertarians -- including my friend, Peter Jaworski -- have defended a "thin" version of libertarianism. In this debate, Seavey is the representative "thin" libertarian. Sandefur and Kuznicki both criticize the thin view, in somewhat different ways.

My perspective on the debate is a bit odd, because I used to be a thin libertarian. Now I'm unsure about thin libertarianism, for many of the reasons Sandefur identifies.

First, some general remarks about the relationship between politics and morality.

Politics always proceeds from some more basic moral ideas. For example, if X is good, and ought to be promoted, then (it might follow) the state ought to promote X. As Jeremy Waldron puts it, "Catholics...have a particular conception of the good, and for many that conception issues in a particular vision of law and justice, expounded (say) in the jurisprudence of Thomas Aquinas. Muslims proclaim a comprehensive religious vision, and this generates for them a particular vision of the well-ordered society."

On this model, politics are "tied to" or "generated" by particular conceptions of the good. When it comes to justifying one's vision of the political order -- and liberalism, in particular -- Ronald Dworkin calls the strategy of linking morality and politics the strategy of continuity.

The advantage of the strategy of continuity is that one's arguments about the good do most of the heavy lifting. If one can make the case that X ought to be promoted, or that X is the way people ought to live, etc., then the appropriate political order just sort of falls out as an implication of the argument. In addition, one does not need to motivate people to adopt this political vision; the motivation springs from prior moral argument about the good.

The problem with the strategy of continuity -- as Rawls saw -- is that it does not seem that disagreements about the good are always resolvable through rational argument. Instead, "reasonable disagreement" reigns: you say people ought to value this, I say people ought to value that, and that's where the argument stops, with neither of us being unreasonable for our insistence on one conception of the good over the other. Or so people who reject the continuity strategy will say.

The alternative to the strategy of continuity is discontinuity. One sharply distinguishes between conceptions of the good and politics. On this model, according to Waldron, "Particular theories of justice are not seen as tied to or generated by particular conceptions of the good. Instead, they stand apart from competing religious and philosophical conceptions. They present themselves as solutions to the various problems which disagreement about the good generates in society." Emphasis added.

According to the first strategy, every conception of the good comes with a companion conception of justice, a particular vision of what the ideal city/state would look like. According to the second strategy, one attempts to form a conception of justice that is independent of any particular conception of the good. A conception of justice with this form can be called a strictly political conception of justice.

People who have different ideas about the good will tend to disagree about politics. Therefore, to gain the consensus of all reasonable people, a conception of justice must be formulated as a separate "module", one that does not depend for its justification on any particular conception of the good. This is the essence of the discontinuity strategy, and it gets its best showing in contractarian theories like that of Rawls and Jan Narveson (although I know from personal experience the latter would not much like to be grouped in with Rawls.)

If justice is a module, distinct from any single conception of the good, where does its content come from? How do we know what justice demands? Here there are different answers. Rawls thinks the content of justice can be generated by working out the implications of certain basic ideas he thinks are already floating around in the culture of liberal democratic societies. These include the idea that citizens are free and equal, that society is a fair system of cooperation, and so on.

Whether he is right or wrong about this, one thing is clear: if this is where justice comes from, its scope is very limited. We cannot say our conception of justice is true, only that it is reasonable. And, even then, it is only reasonable for people like us, who already accept the basic ideas. Rawls explicitly disavows any attempt to link those basic ideas to a more comprehensive moral theory, like those found in the work of Immanual Kant and John Stuart Mill.

Another way to generate a "discontinuous" conception of justice is Narveson's way. For him, the content of interpersonal morality is the system of rules people would agree to, if they started from a Hobbesian state-of-nature that was devoid of moral constraints. What's interesting about Narveson's proposal is that he thinks the rules people would agree to are libertarian in character: don't worsen others; otherwise, do what you want. What makes his argument an example of discontinuity is that this basic moral rule is not an extension of any particular conception of the good. It is the right rule -- he thinks -- because anyone would accept it, almost regardless of his/her conception of the good.

Narveson's view is an important jumping off point for us. I would say he is pretty much the "thin" libertarian par excellence. Consider Seavey's claim:

Libertarianism’s chief strength, then, has always been in recognizing the vast gulf between, on one hand, myriad, never-ending social complaints (along with the conflicting social philosophies built around them) and, on the other hand, the minuscule and tightly constrained range of things that rise (or, if you prefer, fall) to the level of political/legal complaints.
Continuity, to recall, doesn't really see a vast gulf between what ought/ought not to be promoted (simplicter) and what the state ought to promote/not promote. However, discontinuity -- whether in Rawls or Narveson -- does see the gulf. People reasonably disagree about what ought to be promoted. Hence, from the mere fact that X ought to be promoted (even if it is a fact), it does not follow that the state ought to promote X (Jaworski calls this the "ought/state" gap; I like that.)

Libertarianism, understood in the thin sense, is an example of discontinuity. Libertarianism must rise above conflicting "social philosophies." Its prohibitions must be based on different (firmer?) ground. Libertarian property rights, one must argue, can and must be detached from any particular conception of the good, and defended in terms acceptable to many different reasonable views about the good.

Seavey sells his argument a little short by claiming that the Taliban (to take an example) merely have a preference that people live one way rather than another. Rather, I would say that the Taliban have a deep, very comprehensive view about the way people ought to live, about what ought to be promoted. The fact that other people strongly disagree with their conception of the good is a reason to seek a political conception that depends neither on the Taliban's conception of the good, nor on a Millian conception of the good that emphasizes experiments in living.

My own view is that, taken to the logical extreme, discontinuity will almost always end up at something libertarian-like. On this, I think Narveson is correct: if what's important is identifying the norms all would agree to in a suitably circumscribed baseline situation (finding the appropriate baseline is a problem in itself, let me tell you), then the set of norms meeting this description will be small, and mostly "negative" in character: a list of "thou shalt nots", rather than "thou shalts."

One can be a libertarian without being a thin libertarian; but it certainly seems easier to be a thin libertarian.

So what's the problem with thin libertarianism? The issues Kuznicki identifies are practical in nature (and not in a derogatory way.)
The problem I see here is that thin libertarianism...wouldn’t last very long. A thin libertarian state inhabited by sufficiently prejudiced people would all but instantaneously transform itself into an extension of their prejudices. Prejudiced people with full negative rights won’t just sit around muttering curses on their chosen outgroups. They will agitate for instantiating prejudice in law. If they are determined enough, they will eventually succeed. Thus, thin libertarianism is possibly very thin indeed.
I think Kuznicki is right about this. Rawls' own theory has a similar problem: once you've detached political norms from most of morality, what incentive do people have to uphold the political norms? To some extent, Rawls can handle this problem, because he thinks that, in private, different groups will be able to reconcile their various conceptions of the good to his favored conception of justice. The only groups that can't or won't be able to do this will be those with unreasonable conceptions of the good, to whom we don't have to justify ourselves anyway.

I still don't see why this should be, unless one defines the reasonable in a very question-begging way. Thin libertarians are in even bigger trouble: what motivation would a racist majority have to respect the absolute right of self-ownership of minorities? Why would they not work to chip away at it, in whatever way possible?

The worries about thin libertarianism are quite similar to the worries that drove me to a pro-14th Amendment libertarian view (a la Randy Barnett) and away from the "states's rights" libertarianism espoused by Ron Paul and people at Lew Rockwell. Racist factions at a local level, possessing plenary police power, would easily be able to pass laws stripping minorities of rights and dignity.

Rights will not be continually protected unless people agree that rights matter. But one cannot explain why rights matter without delving into the deeper moral issues discontinuity bids us avoid. This is why thin libertarianism fails.

Madison's Federalist 10 is very instructive here.

Sandefur's objections to thin libertarianism are a bit more theoretical in nature. He argues that thin libertarianism -- and likely any "merely" political conception of justice -- is disingenuous. "Every attempt to create a "value free" politics fails, usually by showing that at bottom there is some normative conception about how people ought to live."

I think he's right about this. In any event, I've read a lot of stuff from liberals who defend the view in a discontinuous way, and always, always there's a deeper, more controversial moral view at work (typically, a kantian one.) Kant emphasized the overriding importance of autonomy, arguing (basically) that to fail to respect a person's autonomy was to fail to respect him as a human being.

While this kind of kantian reasoning does seem to prohibit the state from making strongly paternalistic laws, it does not -- or, at least, many believe it does not -- prohibit the state from subsidizing education, to take an example. But the point, as Michael Sandel would say, is that the kantian view is an extremely controversial one. If one's political view is an application of kantian principles, this does not show that the view is wrong; it does, however, belie the suggestion that one is getting beyond moral disagreement.

On the contrary: when one advocates kant-inspired liberalism, one is immediately mired in philosophical debate, much of it as resistant to rational resolution as the disagreements about the good these liberals claim to want to avoid.

Sandefur is also correct that one cannot treat rights as "primary moral principles." At times, Nozick seems to do this (at other times, he just takes a kantian position.) This is not just because the motivation to adhere to rights has to come from somewhere. It's also because -- and this is my point, not Sandefur's, though maybe he'd agree with it -- the application, scope, and shape of rights is always somewhat indeterminate.

Must every state protect free speech at least to the extent guaranteed by the 1st Amendment? In fact, that's a trick question: the 1st Amendment is, in some sense, indeterminate. The Founders could not have foreseen the Internet; yet "freedom of speech and the press" clearly extends to the Internet. Why?

Well, here, I think Ronald Dworkin is on to something. "Freedom of speech" is an abstract principle. Its content can be filled in once one understands that it is linked to a particular conception of the good. If "freedom", broadly construed, is important because the good life is the autonomous life, then we know just how wide the scope of that principle is.

Suppression of speech inhibits participation in the good life, because it threatens autonomy. Suppression of speech on the Internet also (and for similar reasons) inhibits participation in the good life. By the way, this isn't inconsistent with original intent jurisprudence, because, as Dworkin suggests, the Founders could and probably did intend to lay down abstract principles, at least in places.

5. Conclusion:
"Thin" libertarianism has one attractive feature thick, comprehensive libertarianism cannot match: it makes the promise of transcending reasonable disagreement about the good. But it may not be able to fulfill that promise.

At the same time, I'm reluctant to say thicker libertarianism, based on a view of the good that emphasizes autonomy, can do it, either. Autonomy is a slippery idea. Joseph Raz's version of liberalism is firmly based on a plausible interpretation of the concept, but it's not exactly libertarian in its details.

One might say: that's because Raz's conception of autonomy is too robust, too thick. We need a thinner version, like the one Rasmussen and Den Uyl present (they call it self-directedness.) For example, a thin version of autonomy might identify infringments of autonomy with acts of violence, such that the only way one can interfere with another's autonomy is if one uses violence against him or her.

That conception of autonomy would be very thin, and very congenial to libertarianism. But now the argument can now be turned back on the thick libertarian: why should we accept the thin version of autonomy? The argument simply can't be that a thin version of autonomy is more widely accepted than a thicker version. After all, that wasn't enough of a reason to accept thin libertarianism!

My concern is that a thick version of autonomy, like the one Raz espouses, will almost always lead away from a libertarian politics. Thicker versions of the concept of autonomy might imply that a very poor black person who can't get a job in a racist community is lacking autonomy.

And I'm not sure this is wrong; but if we've already accepted that the state ought to treat what will/won't promote autonomy as a reason to act, then we will also be quickly led to accept that maybe the state ought to help this person out in various ways (incidentally, Rasmussen et al. do not respond all that well to this objection in their book.)

Again, I'm not sure it is a bad idea for the state to help out the truly desperate. But it isn't a very libertarian one. To borrow a clever saying from Kuznicki, there may be no such thing as thick libertarianism, at least once one has augmented it with a plausible, thicker view of autonomy.