First of all, let me thank all the contributors to the symposium. All were, one way or another, squarely within the audience I was hoping to reach with the book, and all are scholars whose work I really admire. Of course, that also makes them a daunting crowd in front of whom to display one’s sprawling, putting-lots-of-parts-together book. There are interdisciplinary or multidisciplinary works that build a reputation by everyone thinking it’s pretty good at the areas in which they’re not experts– the constitutional lawyer thinks it’s good political theory, the political theorist thinks it’s good philosophy of law, the philosopher of law approves of the normative conclusions in constitutional law. Putting all your commentators together risks exposure: Annelien de Dijn reveals that I’m an amateur when it comes to the history of French liberalism in the 18th and 19th centuries, Russell Arben Fox corrects me on cities, Rick Garnett shows that I’ve misunderstood libertas ecclesiae, and so on. (One worries about these things.) And so my delight is matched by my relief that these experts in the various fields on which the book draws seem generally to be on board with the way I’ve put the fields together.

I’ve been having a great time reading everyone’s essays, and seeing the different directions in which people are taking ideas from Rationalism, Pluralism, and Freedom. People have been saying very gracious and generous things about the book, and I appreciate them all, but my favorite may well be David Watkins‘ “Levy’s book has the potential to be highly generative.” I do think (and I suggest briefly toward the beginning of the book) that the rationalism/ pluralism distinction is something broader than only an intra-liberal debate, and even within the liberal tradition I don’t by any means think I’ve exhausted what there is to say about it or to do with it. I’m very pleased at how much of the conversation has consisted of my colleagues taking the ideas of the book in directions of their own.

That means there’s relatively little for me to respond to that takes the form “what does Jacob think about X?” But there are some bits and pieces like that. In this first post I’ll respond to some questions Russell and Will generated on federalism and decentralization, and in the second I’ll respond to Matt on markets.

Let’s start with this exchange in comments between Will and Russell. Will said:

I have the intuition that he would say that something like Dillon’s Rule (absolutely no municipal rights as against states) is probably a mistake (and would you, too, Russell?). Does it even make sense to have states so long as we have adequately pluralist and powerful cities and counties?

The answer to the second is easy. Federalism– a system of states or provinces, not of further decentralization– has functions of its own that differ from those of local self-government in the towns that Tocqueville referred to as “civic associations.” I have argued elsewhere that we should not understand federalism as a mere partial fulfillment of some fundamental call toward decentralization or subsidiarity. (And so I was a skeptic avant la lettre of Heather Gerken’s subsequently-famous “Federalism All The Way Down.”)

One longstanding thought of the pluralist tradition holds that when a modern unified state confronts a world of isolated individuals, they will be all but powerless against it. The more atomistic a society is, the more centralized and potentially oppressive its state will be. They need ways to unite, institutions larger than an isolated person through which to engage in politics in their own defense. The United States has some 89,000 local governments— counties, cities, towns, and special districts of various kinds such as school districts. That’s a few orders of magnitude fewer than the number of persons, but still far too many for any one of them or easily-coordinated number of them to have much political weight as a counterbalance against the center. When Will made reference to “adequately pluralist and powerful cities and counties,” I take it that he meant “powerful” in terms of their ability to internally govern, so that they had a reasonable range of matters over which they wielded authority. But there’s another sense of “powerful,” too: powerful as against outsiders, power within the overall political system. I don’t think we can have “adequately powerful” cities and counties in that sense; their relationship to the central government is going to be too much like isolated individuals’ relationship to it. The French revolutionaries knew what they were doing when they smashed the old provinces and replaced them with an order of magnitude more departments. What could look like greater decentralization was, and was meant to be, just the opposite.

Moving on to the law and theory of province-city and province-town relations: I suspect Dillon’s Rule makes sense as a rule of federal constitutional interpretation, and I believe it’s the normal rule of federations other than India. Cities and towns don’t have direct entrenched standing under the federal constitution in the way that provinces or states do. A pluralistic federation may well have reason to trust each state or province to create its own internal ecology of local governments, responsive to its own internal geography, population density, and other conditions. (The structure of local government very reasonably differs dramatically in different regions of the U.S. The sparsely-populated west did not copy the institutions developed in the dense northeast.)

But I am sure that Dillon’s Rule should not be treated as a default in the interpretation of state constitutions in the states’ own courts, and that at a theoretical level we should not think of cities and towns as “mere creatures” of the state or province. (I recommend the essays by Loren King and my colleague Daniel Weinstock in the Nomos volume on Federalism and Subsidiarity I coedited with James Fleming.) Cities in particular– here distinguishing “cities’ from “towns” with reference to size but also diversity, connection to trade and migration networks, hosting universities, and so on– have always had a very uneasy relationship to provinces and the polities that became provinces. Those were in the first instance feudal: duchies and, in the European sense, counties later absorbed into increasingly-unified monarchies. The ideal-type province was mainly rural, heavily customary, and the institutional site for feudal-noble authority. The ideal-type city was a refuge from all of that. It was a place where custom and feudal norms did not bind so tightly, or at all. And so the uneasy political relationship between a major city and the province or state in which it’s situated is built in: they have different ancestries, different basic reasons for existing and sources of legitimacy, and very, very different outlooks. They represent different kinds of pluralism. When a heavily rural and, on some dimensions, conservative province (say, Quebec) messes around with the coherent existence of a cosmopolitan city (say, Montreal), expanding and contracting its boundaries and authority, the odds are good that what’s being done has more to do with the provincial dislike of the city than with reasonable estimates of good administrative practice. No matter how much the city and the country are economically, sociologically, and ecologically entangled, they have distinct political purposes and political cultures. Although I don’t have a proposed legal rule ready to hand, I do think it’s a mistake to treat major cities as the kinds of things that a majority of the state or provincial legislature could reasonably abolish, whether that means breaking them apart or diluting them by expanding their boundaries until the city is ruled by voting majorities of suburbs and exurbs.

When the Parti Quebecois government then in power proposed prohibiting public employees from wearing religious symbols– headscarves, yarmulkes– one question was whether public corporate bodies such as universities and hospitals would have any ability to opt out on something like associational-freedom grounds. Their employees aren’t just like civil servants, after all. As the debate developed, it became clear that towns and cities wanted that ability, too. It wasn’t going to happen politically: the symbolism of what the PQ was after just wasn’t compatible with Montreal city employees wearing headscarves. Immigrant-heavy Montreal was precisely the target of the proposed legislation. But there was something to the demand– some sense that Montreal isn’t just an agency of Quebec, and ought to have some corporate ability to say no to impositions from the province.

It’s worth acknowledging here Russell’s point that cities and towns (in different ways) are in part enterprise associations, on my model. They’re purposive and association-like in important respects, not only state-like. (To my minarchist libertarian friends: pretend we’re here talking about the very big private residential associations that you’d want to see take the place of town or city governments– though it is my view that it gets very hard, very quickly, to maintain a bright-line public-private distinction here.) People in cities and many towns have a legitimate sense of municipal distinctiveness. The person who wants to move to New York, Washington, Paris, Portland, Austin, Montreal, Seattle, Dallas, Tel Aviv, Prague, Tokyo, is choosing a place that’s unlike the others, and very unlike their various surrounding areas. They are wildly diverse, but they’re not neutral. Each is a distinctive (and distinctively evolving) hybrid world. Some small towns, of course, are likewise very distinctive, and sometimes are hardly distinguishable from associations: “around here, these 50-100 families have lived a way of life like this. People who didn’t like it moved away, and those who stay want to keep it that way.” I’ll say more in my next post about Russell’s worry about market erosion of both cities and towns. For now what I want to note is that, while big cities have distinctive identities, and some small towns do, many local governments in between don’t. I’m less concerned here with the mid-sized cities that Russell is researching than with suburbs and exurbs and large towns. These really can be very similar from one to the next, without a strong sense of local belonging. This makes them good competitors on the market for local government envisioned by political economists who write about Tiebout sorting. But it makes them not especially strong manifestations of pluralism, and poor political vehicles for the protection of local identities and interests.

This is part of why federalism is valuable: voters take a much more engaged interest in the politics of their state or province than they do in that of many kinds of localities, with which they interact as public utility providers. States, cities, and some small towns engender identification and loyalty that encourages political commitments. But many other localities aren’t experienced as worth the effort to keep locally accountable, or to protect from extralocal intrusion. (If you don’t like your bedroom exurb, you just leave.) For the many people who live in localities like those, a state or province might be as local a body as could plausibly attract committed political engagement.

I think where I’d end up is: Dillon’s rule might be a good legal rule at the federal level, but is a a bad canon for intra-state or intra-provincial law, and a very bad mindset for intra-state or intra-provincial politics. Provinces/ states and cities/ towns each have their political functions, and at federalism’s best we want a balance between them. States shouldn’t be abolished and dissolved into component localities, but neither should cities and towns be treated as just administrative conveniences of states.