« Pacing Classes | Main | (Re)reading the riot acts? »

Sunday, December 28, 2008

Feeley's and Rubin's Argument by Lexicographical Fiat Against Constitutionally Protecting "Federalism"

Malcolm M. Feeley and Edward Rubin have finally set out their theory of federalism in a book: Federalism: Political Identity and Tragic Compromise (University of Michigan Press 2008) presents their theory that federalism is exclusively a means by which people with different "political identities" preserve their autonomy from the central government. For anyone who cares about the theory of federalism, Feeley's and Rubin's book is one of the most important books on federalism published in the last four years. (Two others are Edward Purcell, Originalism, Federalism, and the American Constitutional Enterprise: A Historical Inquiry (Yale University Press 2007) and Mikhail Filippov, Peter C. Ordeshook, Olga Vitalievna Shvetsova, Designing Federalism: A Theory of Self-Sustaining Federal Institutions (Cambridge University Press 2004)). They present the best statement of what I take to be a powerful fallacy about the justifications for federal regimes. If you care about federalism, you should buy it and read it and -- I hope -- come away unconvinced by its argument.

The essence of Feeley's and Rubin's argument can, I think, be stated in two propositions: (1) As a matter of basic definition, federalism is a regime in which the powers of subnational units are protected from the national government through some kind of constitutional entrenchment and (2) Such entrenchment of subnational power makes sense only if citizens' sense of allegiance -- their "political identity" -- to regional subgroups is stronger than their loyalty to the nation as a whole. For states with such a weak sense of national identity, it might make sense to protect subnational institutions from national power, as a way to give assurance to members of ethnic, religious, cultural, or other subgroups that they will not exploited by more powerful rivals. But, in any state composed of citizens with a firmer sense of national identity -- the United States, Germany, Australia, Argentina, among other apparently federal regimes -- federalism is a mere atavism. Indeed, such regimes cannot be genuinely "federal" in Feeley's and Rubin's sense, because "federalism" is defined exclusively as a mechanism to protect political identity (e.g., Shi'ite, Francophone, Zulu, Marionite Christian, etc.) from majoritarian exploitation. In nations where region does not coincide with such powerful loyalties, there is no need to constitutionalize decentralization of power, even if "managerial decentralization" might be a good policy.

Assuming that the preceding paragraph is a necessarily crude but reasonably accurate re-statement of their argument, then Feeley's and Rubin's book is, in my view, powerfully incorrect. The power of their error is that it is widespread: One often hears that federalism is no longer necessary in the United States because we are no longer divided by powerful regional loyalities. But this argument is erroneous, because it rests on a lexicographical fiat -- namely, the idea that, because protecting powerful regional "political identity" is one justification for constitutionalizing the allocation of power to subnational units of government, therefore it is the only justification that counts as "federalism."

Feeley and Rubin are surely correct that, if the only point of decentralization is to accomplish goals on which national citizens largely agree, then there is no particular reason why decentralization should be constitutionalized. Why not just leave such purely instrumental decentralization up to the discretion of Congress? But decentralization is not merely managerial in this mundane and uncontroversial sense. Decentralization can also be used to protect goals that Congress might not be trustworthy to safeguard without constitutional oversight -- goals like (a) protecting opportunities for political expression and political competition that incumbent politicians in Congress might wish to suppress; (b) limiting the Leviathan-like tendencies of centralized governments to engage in predatory taxation; (c) constraining the universalistic tendencies of a Congress lacking strong political parties to spread pork around in inefficient ways; (d) reducing the costs of political careers so that cub politicians can have a chance at challenging national incumbents; and etc, etc., etc.

Feeley and Rubin dismiss all of these justifications for decentralization as irrelevant to "federalism" in their narrow sense of the term, because these justifications are not related to the protection of subnational political identity. According to Feeley and Rubin, such goals are still merely carrying out the policies of the national citizenry (e.g., fiscal restraint, political competition, small electoral districts, etc.) and, therefore, do not count as "federalism," because the national goals are paramount over the goals of any subunit.

Yes, of course, federalism in this sense is a policy of a unitary people: We the National People simply like decentralizing power to units called "states," for some of the reasons offered above. But so what? What exactly hangs on this lexicographical assertion? The important issue is whether there are good reasons rooted in the unreliability of Congress for entrenching the rules of decentralization in a constitution, even in nations with a strong sense of national unity. If there are, then Feeley and Rubin have not made the case that "federalism" is irrelevant in the United States or other nations with a strong sense of national unity. Many writers -- Brennan & Buchanan, Inman & Rubinfeld, Besley & Case, Benjamin Lockwood, Albert Breton, among others -- have laid out some reasons rooted in political economy to believe that centralized legislatures cannot be trusted to decentralize power optimally. Why not use the label "federalism" to apply to this body of reasons?

Feeley and Rubin appear to argue that a constitutional policy cannot be an instance of federalism unless it is a norm selected by a subnational unit. But this is strikes me as a non sequitur: One might as well say that "individualism" is not constitutionally ensconced because the norms protected by individual rights are not determined by individuals. Yes, We the National People have decided not to trust Congress with complete power over decentralization policy: We constitutionally require decentralization of power to states, juries, newspapers, churches, and a host of other subnational institutions. "Federalism" in this sense is a national policy of a national people that this national people could change at will through an Article V amendment or by changing the composition of SCOTUS or some other formal or informal constitutional means.

So what? Federalism in this sense stands on exactly the same basis as individual rights -- as a constitutional policy of distrust towards centralized institutions. It is an empirical question, of course, whether such distrust is justified; It is a normative question whether We the National People actually care about the sort of decentralization that federalism protects. But these are questions that Feeley and Rubin do not fully engage, because they are caught up with the lexicographers' argument that "federalism" is only concerned with political identity.

Posted by Rick Hills on December 28, 2008 at 11:03 AM in Constitutional thoughts | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Feeley's and Rubin's Argument by Lexicographical Fiat Against Constitutionally Protecting "Federalism":


Thanks, Ed, for these thoughtful remarks. Just to clarify where we agree and disagree: Here are three propositions germane to the issue, and I think that we agree on two of them.

(1) States are not affective communities like (for instance) our families, the Presbyterian Church USA, or even the Quebecois of Canada: Their members are not tied together by strong bonds of affection, culture, belief, or values. (Indeed, I've argued that it would be unconstitutional for a state to attempt to be an affective community in this sense in Poverty, Residency, and Federalism: States' Duty of Impartiality Toward Newcomers, 1999 SUP. CT. REV. 277).

(2) Decentralization of power to regional and local governments is valuable as a matter of policy. You tend to regard it as merely instrumentally valuable to increase governmental efficiency; I think that it is valuable as a protection for our individual capacity to act and reason collectively.

(3) Decentralization sometimes needs constitutional protection, because Congress inadequately protects such long-term, second-order values in the rush of short-term policy-making.

We agree, I think, on propositions (1) and (2); we disagree on (3).

Part of our disagreement, I suspect, rests on our differing conceptions of the function performed by constitutional rights. I maintain that there are lots of constitutional rights that serve approximately the same function as federalism -- i.e., to protect our individual capacity for collective decision-making. Among these rights include (1) The right of private schools to manage their own curriculum (Meyer v Nebraska (1923)), (2) The right of charitable trusts to immunity from state abrogation of their charters (Dartmouth College v. Woodward (1819)), (3) The right of large-scale organizations to enjoy some immunity from regulation of their membership (California Democratic Party v Jones (2000), Boy Scouts v Dale (2000)), and (4) the right of an individual to serve on a jury (Peters v Kiff).

None of these private institutions are close-knit affective communities. None have any entitlement to adopt a moral norm contrary to those fundamental norms adopted by We the People. But all provide important venues for collective decision-making that Congress and/or the state legislatures are likely to slight. The Court, therefore, has protected them for instrumental reasons through constitutional doctrine.

I suggest that federalism doctrine is not different in kind from these protections for private governmental organizations. I've argued at length that the notion that constitutional rights often protect private government for instrumental reasons in "The Constitutional Rights of Private Government," 78 NYU L Rev 144 (2003)), and I won't repeat that argument here. But I think that the burden of our disagreement rests on the central question of whether state governments are all that different from (say) non-profit corporations, juries, trade unions, political parties, private universities, and many other organizations that enjoy constitutional rights for what I've claimed are instrumental reasons.

You could, I think, respond in at least two different ways. First, you could deny that states are analogous to, say, NYU -- that somehow the values protected by gigantic private universities or other non-affective private governments are fundamentally different, more deontological, more "moral" than the values protected by federalism.

I think that this first view is really, really hard to defend. (At least, I've argued at length that it is indefensible in that NYU piece). The Court protects the autonomy of political parties, etc, because these devices are instrumentally useful for the goals of political participation, pluralism, diffusion of power, etc. I just do not see how states are different in kind from these private governments.

Or you might argue that doctrines protecting the constitutional rights of private governments should be eliminated, because they overly ossify policy-making. The second view has a lot to recommend it: If you are willing to bite the bullet and de-constitutionalize a vast swathe of protections for private governments, then I'd agree that states have no greater claim to constitutional protection.

So which position do you adopt? Or is there a another reason for our disagreement?

Posted by: Rick Hills | Dec 31, 2008 6:27:42 AM

Hi Rick,

Thanks for your very thoughtful comments on our book Rick. Your summary is excellent, and almost entirely correct. But you did miss one point, and that’s the issue that forms the basis of your criticism.

Of course, decentralization can have many benefits, including the ones you mention. In fact, it’s our dominant mode of public administration in this country, just as I’m sure it’s the dominant mode of management in many businesses with nation-wide operations. So why not constitutionalize it and secure these benefits? Well, for the same reason why the Framers decided not to constitutionalize the departments of the federal government, why we don’t treat the specification of the army and navy in Art I, § 8(12-14) and Art II, § 2(1) as of constitutional significance, and why the current Court, and almost all observers, think the Lochner-era Court was wrong to constitutionalize laissez-faire economics, or the contours of the common law, or Mr. Herbert Spencer’s Social Statics, or whatever they were doing. Things change, and a good constitution will give the government it creates the ability to adapt to those changes. To wall off an area, like speech or religious observance, from the control of the government that we created and whose basic policies we determine, requires a strong norm, a norm that sufficiently strong that we want it to prevail under any circumstances, including ones we cannot currently envision. Preserving the nation provides such a norm for federalism; I don't think the pragmatic considerations that you mention (controlling pork barrel expenditures, nurturing young politicians) don’t reach that level.

If constitutionalizing federalism didn’t have any major disadvantages, then you might still think that doing so is a good idea, if only to allow the national government to make credible commitments to the good managerial approach of decentralization. But it has a very serious disadvantage – it allows a geographical subdivision of the nation to adopt policies that the majority of people in the nation find immoral. In our own history, it protected slavery, then apartheid, in the South after the rest – the majority – of the country found those practices to be anathema. Maybe the same thing could be said now about the legalization of marijuana, gay marriage or assisted suicide (lefty that I am, I think norms are changing in the direction of these policies, but for the sake of the argument, let’s assume they don’t). In a nation where the people see themselves as members of a single political entity, and are willing to accede to the normative judgments of the majority (or the national elite in a non-democratic regime), there is no reason to tolerate actions by the regionally-based minority that the majority regards as morally wrong. When the people don’t have that view, however -- where their political identity is linked to the region instead of the nation, or when it's sufficiently divided between the two that they will side with their region in many cases – such toleration may be necessary to hold the nation together. That's not a lexicographer’s argument, but a basic feature of political organization.

I’m actually a fan of decentralization; as an administrator, I’ve decentralized our law school. And I get as annoyed as anyone else when national policies that I don’t like prevail over state policies that I prefer. But federalism is really strong medicine, and using it in a nation whose people have a unified political identity is like using chemotherapy to cure a cold. The recognition of individual rights, and thereby prohibiting the government from regulating speech or religion is strong medicine too, but we constitutionalize these individual rights because they are central to our concept of political morality. Decentralization may be good policy, but it doesn’t rise to that level, so it doesn’t merit constitutional protection in a nation whose people have a unified political identity.

Posted by: Ed Rubin | Dec 30, 2008 1:30:17 PM

Thanks for the responses, guys. Dan's question touches on a particular obsession of mine, so I responded in a separate post.

As to Rick's question, I think that Malcolm and Ed take the view that subnational political identity actually exists, then there really cannot be much of a national political identity: One sort of identity crowds out the other. Either citizens views themselves primarily as Southerners, Zulus, Quebecois, or they view themselves primarily as Americans, South Africans, Canadians. Political allegiance is like monogamous marriage: One can have only one mate and one political loyalty.

This does not mean that there could not be a national policy protecting loyalties less intense than "subnational political identity." One might have a national policy protecting autonomy for enclaves of people with beliefs distinct from the rest of the nation, even if those people ultimately acknowledge that they are Americans first. But I infer that Malcolm and Ed would argue that this sort of devolution can be accommodated without constitutional protection for federalism -- that it is just ordinary "managerial decentralization."

It is this last point that I do not quite understand: Why do they think that Congress will always optimally decentralize without constitutional constraints? But I may have missed the gist of their argument, so read their book for yourself.

Posted by: Rick Hills | Dec 29, 2008 10:53:39 AM

Great post, Rick. Thanks. I need to read the book, I know, but, going just on your review, I don't see why F & R would rule out the possibility that it could be a "national policy" to "protect subnational political identity" -- assuming that there is national consensus that this identity is worth protecting and, in the absence of federalism, vulnerable to the effects of centralization -- by a particular means, i.e., federalism.

Posted by: Rick Garnett | Dec 28, 2008 11:46:33 PM

Hi Rick,
thanks for this helpful review. I'm not sure if this is the right post to lodge this query, but the bulk of your federalism posts (all very interesting!) have got me wondering: could you explain which sub-national entities you find normatively more authoritative when there are conflicts between cities and states, or cities and counties, etc. (When I was in practice, I worked on a SCt case that involved a conflict between the telecom preferences of the city vs the telecom preferences of the state.)

Assuming one wants to become a fan of "certified federalism nuttiness," what's the best strain of thinking on resolving this kind of conflict? What are the options?

Posted by: Dan Markel | Dec 28, 2008 5:34:36 PM

The comments to this entry are closed.