So far I have found this month's Cato Unbound extremely stimulating. It sure helps when you get to invite the discussants, but the problem of how exactly limited-government types think government can realistically be limited really is of the first importance.
I think Anthony de Jasay is right that incentive-compatibility problems plague attempts to keep government lean and limited. That said, I think a certain kind of anarchist, like de Jasay, tend to somewhat oversell the impossibility of limited government. As Gordon Tullock likes to emphasize, given the vast amount that could be extracted by political predation, the puzzle for the political scientist is to explain why so little is invested in rent-seeking. Part of the answer lies in the structural constraints de Jasay mentions in his essay. The prospect that financial and human capital may flee a grabbing hand, or the fear that the electorate will rise up in anger and panic when the thicket of opportunistic regulation has begun to strangle prosperity, may rein in government. But these are constraints implicit in the nature of things, not ones imposed by law as limits on lawmakers. So it is interesting that he also mentions the campaign-finance rule as a constraint on the size of government, since that seems open to choice, to design, in a way that the other constraints are not. This seems like an admission that certain rules can successfully bind.
I think I'm almost entirely in agreement with the main thrust of Jerry Gaus's reply. The problem isn't so much the weakness of formal, paper constraints, but the weakness of formal constraints that are not reinforced by our moral sentiments. If a formal rule is seen as merely conventional, and therefore revisable by the relevant authority, and not as moral, there may be little resistance to overriding it in order to meet the demands of weightier moral rules. I found this passage especially illuminating:
[I]f the basic normative commitments of classical liberals were widely conceived of as moral rules, then there would be much deeper resistance to government-made rules that seek to cancel or override them. The problem is that the opposite seems nearer the truth: for many citizens, their understanding of the moral norms related to fairness endorses government-made rules overriding the conventional rules of property. The welfare state reigns supreme not because the state and it allies have tricked the rest of us in a power grab; it reigns supreme because in the eyes of most citizens it conforms to the egalitarian fairness norms that have evolved with humans (Fong, Bowles, and Gintis, 2005). Classical liberals who convince themselves that the New Deal is best explained as a power grab by Roosevelt and his allies are manifestly deluded: it was (and still is) very widely seen as demanded by our sense of fairness.
I think this is on the right track. But I think it's worth emphasizing that the power grab explanation is not at all inconsistent with the “mandate of fairness” explanation. Power-seeking politicians can create the perception that their role and their power is legitimate by appealing to deep-seated moral sentiments. Second, I'm not so sure that our egalitarian sentiments are all that close to a pure expression of egalitarian sharing norms. First, there is the artificiality of nationalism, and the modern welfare state is nothing if not an expression of economic and moral nationalism. To see co-nationals in a vast pluralistic territory as part of a common tribe in which even an attenuated form of ancient sharing norms apply requires an incredible, imaginative, “unnatural” expansion of the circle of affinity.
But I think the general point stands. Moral rules are processed differently than conventional rules. If limited government is going to have a chance, it must be in sync with our moral sentiments and dispositions to moral judgments. I don't think this is impossible. I'm pretty well sold on something like Jonathan Haidt's multidimensional conception of the moral sense. And there may be something like a classical liberal calibration of the moral sentiments, such that certain rules limiting the domain of political power and collective choice may come to be experienced as distinctively moral, and therefore non-optional.
Now, I don't know that there is such a thing, but there might be. I do think there is a broadly liberal calibration of the moral sense, I think that it is prevalent in liberal societies, and that is what makes them stably liberal. That means, in no small part, that the government is effectively limited in what it may do to people. Limited government is evidently possible because it is actual.
The idea that the there are various dimensions of the moral sense each with its own parameters implies that morality is a fill-in-the-blanks slate. The moral sense then isn't an exogenous variable acting as a hard constraint on feasible social coordination. Nor is it infinitely malleable. There are only so many combinatorial possibilities, and the feasible cultural/developmental paths from one combination of settings to another may be quite limited.
But this kind of view does I think put ideas about pluralism and liberal neutrality that both Jerry and I are very fond of in a tight spot. The multidimensional moral sense view makes it pretty clear that liberal society requires that a certain kind of moral personality become common in the population. A specifically classical liberal society, in which the certain further limits on the scope of politics are felt strongly to be moral, may require an even more tightly-focused and even more-broadly shared, fine-tuning of the moral sense. But I'm not certain I'd even want that.