Centralization and the Commerce Clause

Adrian Vermeule

The author is a Professor of Law, University of Chicago. This Dialogue is adapted from Adrian Vermeule, Does Commerce Clause Review Have Perverse Effects?, 46 VILL. L. REV. (forthcoming 2001) (presented at the Villanova Law Review symposium, New Voices on the New Federalism, Oct. 28, 2000). Thanks to Jack Goldsmith, Daryl Levinson, Eric Posner, Rachel Schowalter, David Strauss, and Cass Sunstein for helpful comments. Jamil Jaffer and Stephanie Morris provided excellent research assistance.

[31 ELR 11334]

There is a crucial, although implicit, empirical assumption in the debate about federal judicial review under the affirmative U.S. Commerce Clause. The assumption, indulged by many different camps in the debate, is that increasing the stringency of Commerce Clause review decreases the centralization of policymaking by shifting policy authority to the states.¹ I want to suggest that, on equally plausible empirical assumptions, increasing Commerce Clause review will in fact do just the opposite: it will promote the centralization of public policy at the national level by providing congressional coalitions with ex ante incentives to legislate more broadly, and to create national programs that are more comprehensive, than they would otherwise choose. So those who favor Commerce Clause review because they favor decentralization may have picked a course of action with perverse effects; they may have picked the wrong team. And those who favor Commerce Clause review because they believe the U.S. Constitution commands it should take into account that increased centralization may be a cost of their position.

This view has important consequences for environmental law and policy. Many leading environmental statutes are authorized, if at all, by Congress' power under the Commerce Clause, and many of the most important Commerce Clause cases have involved environmental statutes of one sort or another. Examples are the Supreme Court's recent decision restricting the U.S. Army Corps of Engineers' authority under the Clean Water Act (CWA) on the ground that a broader reading of the statute would have posed serious Commerce Clause questions,² and the Fourth Circuit's recent heated debate about Congress' Commerce Clause authority to regulate endangered species.³ The view advanced here suggests that proponents of national environmental regulation have less to fear from the revival of Commerce Clause review than is usually thought, while proponents of state authority have less to gain.

The Theoretical Problem

For ease of exposition, I shall stipulate that the Supreme Court Justices who support Commerce Clause review do so because: (1) they support policy decentralization; and (2) they assume that Commerce Clause review will promote that goal.⁴ This is a soft rational-choice assumption that allows Justices to hold preferences over institutional arrangements as well as over substantive policies.⁵ A different view would hold that the Justices who support Commerce Clause review do so because they believe the Constitution requires them to do so. But I need not contest that possibility, because all the points I want to make hold even if it is true.

The critical question is whether the revival of Commerce Clause review will indeed promote decentralization. The standard assumption is that it will, and that view has intuitive appeal. Suppose that at some given time courts are engaging [31 ELR 11335] in no Commerce Clause review — roughly the situation before Lopez v. United States.⁶ After that time, as the courts move incrementally from no Commerce Clause review to moderate Commerce Clause review — striking down a couple of statutes, as in Lopez and United States v. Morrison⁷ — Justices may assume that centralization decreases linearly. After all, the courts are striking down federal statutes on the ground that the statute exceeds the federal government's constitutional authority, even though the states may enact precisely the same rule (absent some other constitutional prohibition). There is also another decentralizing effect: not only are the courts wiping out national regulatory statutes, but legislative coalitions may sometimes take account of the new constitutional restrictions by filtering out proposed bills that would violate them (if the risk of judicial invalidation makes Congress less likely, rather than more likely, to enact unconstitutional laws — admittedly a murky question).

But the assumption that an incremental increase in Commerce Clause review produces a linear decrease in centralization may get things backwards. Instead, a move from no Commerce Clause review to some Commerce Clause review may produce an increase in centralization. To be sure, a further move to even more intensive review — a move from the position of the Lopez majority opinion to, say, Justice Clarence Thomas' concurring position⁸ — might begin to effect real decentralization. But that is just to say that centralization, as a function of increasingly stringent Commerce Clause review, might not be continuously decreasing, but rather might display an inverted U-shape.

This inverted U-shape arises if and when the doctrines the courts use at the intermediate intensities of review allow Congress to enact otherwise-unconstitutional policies by broadening their scope or by bundling them together with valid policies. The proponent of Commerce Clause review assumes that if Congress enacts policy P and the courts strike it down, then the decision has increased decentralization or, equivalently, prevented a new centralization. But if the courts' rules allow or encourage Congress to enact P so long as P is broadened to include some admittedly constitutional policy Q, or bundled together with policy Q, then the result of striking down P may not be to remit the decision about P to the states. It may simply be to produce a federal statute that mandates both P and Q, either because Congress reenacts the statute in its new, more expansive form, or because Congress anticipates the effect of the judicial rule and enacts the expansive form of the statute in the first instance.

Current doctrine under the Commerce Clause has just this effect of encouraging the broadening and bundling of federal policies. Two doctrines are critical: the aggregation principle and the comprehensive scheme principle. Aggregation is familiar. The Supreme Court has said since Wickard v. Filburn⁹ that the substantial effects test should be applied not to some particular instance of a regulated intrastate activity, but to the class of all such instances taken in the aggregate.¹⁰ Morrison added that aggregation has only been allowed (and presumably will only be allowed) for intrastate activities that are themselves economic or commercial.¹¹ It is clear that the aggregation principle "condition[s] the Commerce Clause power to accomplish a certain goal on Congress legislating far more broadly than necessary," as John Nagle puts it, because the effect of the principle is that "if Congress gathers enough substantial impacts into the covered class, the trivial impacts can be regulated, too."¹²

The comprehensive scheme principle has a similar consequence. This principle holds that the regulation of some activity that Congress could not reach standing alone, because the activity occurs intrastate and lacks a substantial effect on commerce in its own right, may nonetheless be constitutionally permissible if the regulation of that activity is essential or integral to the maintenance of a larger regulatory regime governing interstate activity or commercial activity or both. This idea is at least as old as the Shreveport Rate¹³ cases, but it takes center stage in Hodel v. Indiana,¹⁴ which stated that challenged provisions not valid in themselves will be upheld if they are "an integral part of [a] regulatory program" that is valid when taken as a whole.¹⁵ And the idea surfaces in Lopez itself in a critical passage that has gone largely unnoticed by commentators. The Gun-Free School Zones Act, the Court said:

Is not an essential part of a larger regulation of economic activity, in which the regulatory scheme could be undercut unless the intrastate activity were regulated. It cannot, therefore, be sustained under our cases upholding regulations of activities that arise out of or are connected with a commercial transaction, which, viewed in the aggregate, substantially affects interstate commerce.¹⁶

The lower courts, however, did notice this passage, and they have frequently invoked the quoted passage from Lopez, both before and after Morrison.¹⁷ I'll give some examples below of cases that uphold statutes very much like the statute struck down in Lopez, merely because those statutes were packaged along with a larger national regulatory scheme. But here I'll mention a case decided after Morrison that shows that bundling provisions can ensure the validity of all of them, a case with important ramifications for environmental [31 ELR 11336] law. The decision is Gibbs v. Babbitt,¹⁸ in which the Fourth Circuit upheld federal regulations that limited the killing of endangered red wolves on private land in North Carolina.

Lopez and Morrison announced that the commerce power authorizes congressional regulation of: (1) the channels of interstate commerce; (2) instrumentalities, persons, or things moving in interstate commerce; and (3) intrastate activity that substantially affects interstate commerce.¹⁹ Gibbs upheld the red wolf regulation on the ground that it fell within the third Lopez/Morrison category.²⁰ The court invoked the aggregation principle to brush aside the objection that killing a single wolf doesn't affect interstate commerce.²¹ Killing all the red wolves would affect interstate commerce by eliminating the red wolf tourism industry and in other ways.²² But Morrison says that only intrastate "commercial" or "economic" activities can be aggregated,²³ and killing a red wolf doesn't look very much like commercial or economic activity. The court said a killing might have a commercial or economic motivation, if the farmer was trying to protect his livestock or homestead,²⁴ but so might carrying a gun around a school, if the gun owner is selling drugs to young children and wants to protect his sales territory against competitors.

Given the weakness of the aggregation argument, it is not surprising that the court also invoked the comprehensive scheme principle.²⁵ The red wolf regulation, the court said, was "sustainable as 'an essential part of a larger regulation of economic activity.'"²⁶ The plaintiffs in Gibbs hadn't challenged the facial validity of the Endangered Species Act (ESA) under the Commerce Clause, so the court could simply assume that the Act is indeed a valid national regulatory scheme that bundled red wolf protection with protections for many other endangered species in other states. The bundling appeared critical: "Given that Congress has the ability to enact a broad scheme for the conservation of endangered species," the court wrote, "it is not for the courts to invalidate individual regulations."²⁷ This can only mean that the red wolf regulation, even if unconstitutional standing alone, was constitutional because it was packaged with a broader set of valid prohibitions.

This emphasizes one important difference between the comprehensive scheme principle and the aggregation principle. The two are closely related because federal regulation of a class of activities that affects interstate commerce when taken in the aggregate will often be integral to the success of a comprehensive national regulatory scheme. But the best reading of the cases suggests that the comprehensive scheme principle, unlike the aggregation principle, may allow Congress to regulate intrastate activities that are not themselves commercial or economic, so long as the regulation is integral to the success of a larger valid scheme of (interstate or commercial) regulation. The key passage from Lopez, for example, suggests that the scheme taken as a whole must regulate economic activity, while the ancillary regulation need not itself do so, at least if the ancillary regulation "arises out of" or is "connected to" commercial activity.²⁸ Judge Edith Jones, who is occasionally unsympathetic to national regulation, understands the comprehensive scheme principle in this way.²⁹ So does Justice Stephen Breyer: in an insightful passage in his Morrison dissent, he asked whether the comprehensive scheme principle would allow Congress to "save the present law [the Violence Against Women Act (VAWA)] by including it, or much of it, in a broader 'Safe Transport' or 'Workplace Safety' act?"³⁰ We don't know, of course, whether the Court will eventually confirm this understanding, but certainly that's currently the law in the lower courts, as will be discussed later.

An Example

The aggregation and comprehensive scheme principles allow and encourage Congress to ensure the constitutionality of otherwise-suspect provisions by broadening their scope or by bundling them into a comprehensive scheme of national economic regulation. The ex ante effect of the current rules, then, may just as easily promote broader federal regulation — policy centralization — as retard it. A simple numerical example will illustrate the effect. Imagine that there are three legislators, A, B, and C. They are considering three proposals:

. Proposal 1 is a bill titled the "Gun-Free Nation Act." Section 1 of the Act prohibits the transportation, use, etc. of handguns in interstate commerce and is clearly constitutional under current doctrine. Section 2 of the Act prohibits the bare possession of a handgun, anywhere.

. Proposal 2 prohibits possession of a handgun within 1,000 feet of a school. This is the "Gun-Free School Zones Act" invalidated in Lopez.³¹

. Proposal 3 is that there be no federal handgun regulation (the status quo ante).

Reading "Z(p>q>r)" to mean that "legislator Z prefers proposal p to proposal q and prefers proposal q to proposal r," our three legislators are assumed to have the following preference ordering over these choices:

[31 ELR 11337]

. A (1">2">3)

. B (3">2">1)

. C (2">1">3)

The interesting legislator here is C. Legislators A and B are ideologues who both arrange the bills in order of their centralizing effects, although they evaluate those effects from opposing normative premises. Legislator C, however, has a complex preference structure: she is a moderate who opposes comprehensive national gun regulation, but prefers that to no national gun regulation at all. Legislator C has a puzzling world view from the standpoint of someone who is ideologically committed either to seeing any incremental centralization as good (A) or any incremental centralization as bad (B), because C's preferences aren't arranged in order of the bill's centralizing tendency. Why might C have that preference structure? Well, why not? Voting for a broad statute imposes political costs on C, in the loss of political support from regulated parties, but voting to maintain the nonregulation status quo would forfeit political support from those who desire regulation. In C's case, these forces might net out as described.

Given the preferences of legislators A, B, and C, imagine a series of pairwise votes across the three proposals. In a vote between proposals 1 and 2, 2 wins, while in a vote between proposals 2 and 3, 2 wins. Proposal 2, the Gun-Free School Zones Act, is enacted.³² After Proposal 2 is enacted, a court strikes it down as exceeding congressional authority over commerce. Proponents of localism dance in the streets. Subsequently, however, there is another vote between Proposal 3 and Proposal 1. The winner is Proposal 1, and the Gun-Free Nation Act becomes law.³³ Is Section 2 of that law — the part that prohibits the bare possession of a handgun, anywhere — constitutional? The courts of appeals think so. Consider 18 U.S.C. § 922(o), which prohibits the simple "possession" of a machine gun acquired after 1986.³⁴ The section number should ring a muffled bell. The statute struck down in Lopez was 18 U.S.C. § 922(q); both provisions derive from the Firearms Owners' Protection Act of 1986.³⁵ The eight circuits that have considered the constitutionality of the machine-gun-possession ban have all upheld it.³⁶ The opinions commonly distinguish Lopez on the ground that broader federal regulation is more constitutionally defensible. The Third Circuit, for example, reasoned that the Gun-Free School Zones Act attempted to regulate possession only within school zones — "a discrete area unlikely to have a meaningful aggregate effect on commerce" — while the machine-gun statute should be sustained because it is a general (albeit intrastate) ban.³⁷ This is the perverse effect of the aggregation principle: broaden the statute's reach so there are more applications to aggregate until the bar of the substantial effects test has been cleared.

An objection to this analysis is that after Morrison, only economic activities can be aggregated, and possession of a machine gun does not look economic in any simple sense. So an even more popular rationale for upholding § 922(o) has been the comprehensive scheme principle. The Second Circuit, following and summarizing case law from all over the nation, distinguished Lopez on the ground that the machine-gun statute is "integral to a larger federal scheme for the regulation of trafficking in firearms."³⁸ The standard claim is that the federal regulatory scheme both dampens supply, by prohibiting trafficking and sale of firearms in interstate commerce, and also dampens demand, by criminalizing intrastate possession. The demand-side regulation, then, is a necessary auxiliary to the supply-side regulation, and the supply-side regulation is clearly valid, as an interstate regulation of commercial traffic. The flaw in the Gun-Free School Zones Act, on this theory, is that it didn't ban enough intrastate possession to squelch the demand side of the firearms market.³⁹

Both the aggregation theory and the comprehensive scheme theory support our hypothetical Gun-Free Nation Act. Section 1 is clearly constitutional. Section 2 can be upheld on the aggregation ground if we describe the prohibition as "economic," and by the Second Circuit's reasoning it can certainly be upheld as an essential part of a comprehensive regulatory scheme. The ban on handgun possession is a necessary auxiliary measure for dampening the demand side of the market for illegal handguns, a market whose supply side Congress has attacked by an interstate regulation of unquestioned validity. By either expanding the reach of the prohibition to aggregate more conduct, or by bundling the intrastate prohibition with a comprehensive interstate regulatory scheme, Congress can ensure the constitutionality of a provision that would be unconstitutional in its narrower, unbundled state. The upshot is that the Gun-Free Nation Act will probably be upheld by the courts. The final result is an increase in centralization, relative not only to the no-regulation baseline but relative to the law held unconstitutional on federalism grounds. That result is perverse from the decentralizers' point of view.

Put another way, people like Justice Thomas have overlooked the fact that if you can't get your first choice, you're not necessarily better off the closer you get to your first choice. From Thomas' point of view, the first choice would [31 ELR 11338] be an extremely restrictive view of congressional authority under the Commerce Clause. But he can't get that, so Thomas has joined opinions in Lopez and Morrison while issuing brief concurrences that tell us he is merely voting for the rule closest to his own preferences within the feasible set.⁴⁰ But the effect of the Court's intermediate position may be to move outcomes away from, not toward, Thomas' preferences, even relative to the pre-Lopez baseline. If Thomas' first-choice position derives from a preference for localism, he should consider joining Justice Breyer and the other dissenters.

From the point of view of moderately pro-federalism Justices such as Sandra Day O'Connor and Anthony M. Kennedy, the error is also perfectly natural. Here is one plausible picture of the moderates' thinking. Starting from a baseline of no Commerce Clause review at all, these Justices desire to increase the intensity of review just a bit, eliminating a few statutes they see as largely symbolic outliers, such as the Gun-Free School Zones Act and the VAWA,⁴¹ but without destabilizing politically entrenched legislation such as the New Deal entitlement statutes or the ESA.⁴² So the natural position is to say that the outlier statutes don't have a sufficient effect on commerce, or the right kind of effect, while developing doctrinal exceptions that protect comprehensive federal programs. The problem is that so long as the reigning legislative coalition understands the rules and prefers some program to no program at all, the newly enacted programs will tend to become more comprehensive.

In both cases the proponents of localism have overlooked the fact that once the intermediate option is eliminated, there is no particular reason to assume a priori that the second choice of federal legislators will be no federal regulation as opposed to far-reaching federal regulation. Precisely analogous perverse effects turn up in many areas of law in which courts or regulators knock out an intermediate option without anticipating that the regulated parties may then be driven to an extreme that is even less desirable. Consider Nollan v. California Coastal Commission,⁴³ in which the Court held that localities couldn't condition a development permit on a nongermane concession, such as beachfront access. Some property rights fans expected that after Nollan local governments would start granting permits without conditions. William Fischel has argued that, in fact, localities might just refuse to grant permits at all,⁴⁴ and although the issues are complicated, he is clearly right that that effect is empirically possible. For another example, consider the claim — whether valid or not — that the Court has erred by requiring states to provide equal welfare benefits to long-time residents and to recent arrivals.⁴⁵ The states' reaction may not be to level up by giving recent arrivals as much as established residents are getting; the states may well level down by cutting benefits for both groups. These are just two examples among many. The first example illustrates a general problem in unconstitutional conditions cases, where the question is whether invalidating the condition will produce benefits without conditions or no benefits at all. The second illustrates a general problem in equal protection cases, where the question is whether barring a discriminatory program will cause political actors to provide a nondiscriminatory program or to abolish the program altogether. In all of these cases, whether the result can be described as "perverse" turns upon what we are trying to achieve; from an egalitarian standpoint, for example, it doesn't matter whether an inequality is cured by leveling down or leveling up. The important point is that the effect of knocking out intermediate options is unpredictable across contexts.

The Empirical Problem

All I have demonstrated so far is that a perverse effect is possible; there is absolutely no reason to assume a priori that increasing the intensity of Commerce Clause review, from a baseline of no review, causes a linear decrease in centralization. It may well cause an increase in centralization, as congressional coalitions who prefer a broad federal regulatory scheme to no federal regulation at all broaden and bundle provisions to ensure their constitutionality. That response would persist until the Supreme Court moved to a far more intensive form of review a la Justice Thomas. But the Court has no stomach for that course of action; it is not politically feasible. As a result, the principal consequence of the recent revival of Commerce Clause review may be to increase the centralization of national policymaking.

The next question is whether the possibility of perverse effects is empirically serious or not. There is no "burden of proof" on this question; proponents of Commerce Clause review who assume a decentralizing effect have no foundation for their assumption, so we have to approach the empirical question without presuppositions in either direction. Here I shall first sketch some of the variables that an empirical analysis would have to consider, and then I'll ask what courts should do about the Commerce Clause if (or during the period that) they have no definitive answer to the empirical question about the magnitude of perverse effects.

One natural starting point leads down a blind alley. This is the observation that striking down statutes is likely to reduce the total volume of federal law because political inertia — the costs of legislating produced by bicameralism, presentment, and so on — make it difficult for Congress to respond by reenacting a broader or bundled statute to save a statute that has recently been invalidated. It's not so clear that the conventional image of the inertia-ridden Congress is sensible, but the real flaw in the observation is analytical. It's true that when narrow statutes like the Gun-Free School Zones Act are invalidated, inertia may prevent Congress from enacting a correcting statute. The contrary effect, however, is that legislative coalitions, anticipating judicial behavior, will in some cases broaden the scope of statutes when they are first enacted, packaging suspect provisions with unimpeachable provisions in order to ensure that the suspect provisions are held constitutional. The volume of federal lawmaking is a function not only of the number of extant statutes but also of their scope, and the perverse effect of intermediate intensities of review is to reduce the number [31 ELR 11339] of statutes while broadening the scope of new enactments. The latter effect may dominate the former.

There is also a more subtle argument: courts should not worry about the perverse effect of current doctrine because the harms of the perverse effect, if any, have already been felt. The aggregation doctrine has been around at least since Wickard⁴⁶; the comprehensive scheme principle has roots in the Shreveport Rate⁴⁷ cases from 1914; and it flowers no later than the 1977 case Hodel.⁴⁸ Perhaps Congress has already broadened and bundled its enactments to account for these doctrines. If, on the other hand, Congress is not sensitive to Commerce Clause doctrine, then the Court's recent decision to initiate more aggressive Commerce Clause review can't make things worse (from the decentralizers' standpoint). But this point fails to distinguish two possible roles that doctrines such as the aggregation and comprehensive scheme principles might play: (1) as safe harbors for congressional exercise of the commerce power; or (2) as restrictions on the exercise of that power. In Wickard and the other cases, the effect of emphasizing aggregation or comprehensiveness is to make safely constitutional a law that pressed the limits of contemporaneous Commerce Clause doctrine.⁴⁹ After that sort of opinion, it's quite easy to imagine congressional coalitions drawing the lesson that such doctrines are safe harbors but not restrictions; the Court would invoke them to sustain legislation that regulated intrastate activities, but might not point to their absence to invalidate intrastate legislation. In that case there would be no incentive to broaden or bundle enactments. The incentive arises only if the doctrines cut both ways so that meeting their conditions validates an intrastate regulation, while failing their conditions makes the regulation invalid.

This view explains the seeming puzzle that Lopez has been the most dramatic Commerce Clause decision in decades even though it effected little surface change in Commerce Clause doctrine. The answer to the puzzle is that before Lopez, the nominal limits on the commerce power weren't enforced; legislators had every reason to believe that the Court's repeated warnings about limits on the commerce power were cheap talk and that any new statute regulating intrastate activity would be upheld on some ground or other. That answer, however, exacerbates our uncertainty about the possibility of perverse effects in the future. In the prior period, legislators might have thought that any intrastate regulation could be enacted under the commerce power, but in important instances — such as the Gun-Free School Zones Act — they chose to enact regulations of very confined scope. Lopez and Morrison now give legislators reason to believe that the aggregation and comprehensive scheme principles serve as restrictions and, thus, create the incentive to bundle and broaden provisions into larger and more centralized packages. That incentive will operate, to some degree, in the short and medium term, but it is still far too early to tell how strong the perverse effect will be.

Conclusion

So what should judges do in the face of this uncertainty? The judge who favors centralization shouldn't worry too much about the revival of Commerce Clause review; in this arena the decentralizers may just be spiking the ball in their own end zone. True, the centralizing judge who is certain that the perverse effect will hold and that no other judge knows this might slyly vote for expanded Commerce Clause review, but in the face of uncertainty that strategy is too clever by half; it would be better for the centralizing judge to turn to other, more pressing battles.

The harder question is what the judge who favors decentralization should do. The sensible answer, it seems to me, is to drop Commerce Clause review altogether. The possibility of perverse effects could probably be avoided by moving to extremely aggressive Commerce Clause review, but that is politically infeasible. Given that the only real choice — between moderate review that might be perverse and no review at all — is afflicted by severe uncertainty; the tiebreaker ought to be the consideration that if decentralization is the aim, it can be more efficiently pursued by other means.

One option for decentralizers would be further development of the commandeering case law or the state sovereign immunity case law. I mentioned at the outset that the story about perverse effects seems much cleaner and easier to tell for the commerce power than for the other areas of the Court's recent federalism jurisprudence, so there should be little fear that these substitutes will prove equally perverse. True, in those areas the decentralizer gets less bang for the buck because rules that prohibit commandeering and invasions of state sovereign immunity merely preclude certain means of enforcing federal policies; they do not preclude the policy altogether. But if Commerce Clause review really does have perverse effects, then these substitute doctrines, however feeble, cannot help but make a superior contribution to the decentralizer's cause.

For environmental law in particular, the consequences of this argument cut against the received wisdom. Decentralizers sometimes claim that Congress lacks Commerce Clause authority to enact statutes such as the ESA and the CWA, and believe that judicial scrutiny of those statutes will shift the locus of environmental policymaking to the states. Defenders of national legislative authority defend Congress' Commerce Clause authority, in part because they fear the decentralizing effect of heightened Commerce Clause review. On this account, however, both camps should rethink their positions. Environmentalists who favor national-level regulation need have no great fear of Commerce Clause scrutiny, while decentralizers should fear that the revival of the Commerce Clause will hurt rather than help their project.

1. See Kathleen F. Brickey, Crime Control and the Commerce Clause: Life After Lopez, 46 CASE W. RES. L. REV. 801, 842-43 (1996) (explaining impact of Commerce Clause review under United States v. Lopez on congressional action to federalize state or local crimes). In Brickey's view, Lopez counters the congressional trend toward federalizing traditionally state-controlled acts by serving as a "reminder that, contrary to contemporary thought, congressional power under the Commerce Clause is not unlimited, that states have primary authority to define and enforce criminal laws, and that much of what Congress has enacted needlessly alters the balance between federal and state jurisdiction." See id. at 843-44. See also Robert F. Nagel, The Future of Federalism, 46 CASE W. RES. L. REV. 643, 655-58 (1996) (noting one assumption underlying Lopez is that the U.S. Supreme Court favors "significant decentralization").

2. See Solid Waste Agency of N. Cook County v. Corps of Eng'rs, 531 U.S. 159, 31 ELR 20382 (2001).

3. See Gibbs v. Babbitt, 214 F.3d 483, 30 ELR 20602 (4th Cir. 2000).

4. These are descriptive premises. I mean to take no position on the claim that the normative arguments for federalism are merely arguments for administrative decentralization. See Edward L. Rubin & Malcolm Feeley, Federalism: Some Notes on a National Neurosis, 41 UCLA L. REV. 903, 914-26 (1994).

5. But see Frank B. Cross, Realism About Federalism, 74 N.Y.U. L. REV. 1305, 1307-13 (1999). Cross' view seems to be that nobody cares about which level of government makes which decision; instead people care only about substantive outcomes, and all federalism rhetoric is opportunistic. See id. at 1307 (arguing that "federalism is consistently … employed only derivatively, as a tool to achieve some other ideological end, rather than as a principled end in and of itself"). But that claim faces too many counter-examples. Any view that requires attributing to Chief Justice William H. Rehnquist a substantive preference for allowing guns to be carried near schools never gets off the ground. See United States v. Lopez, 514 U.S. 549 (1995) (invalidating Gun-Free Schools Zone Act of 1990). To be sure, the proponent of decentralization may hope that it will promote some more remote substantive end. The libertarian defender of property rights, for example, may hope that interjurisdictional competition will protect property from redistributive legislation. See Richard A. Epstein, The Proper Scope of Commerce Power, 73 VA. L. REV. 1387, 1454 (1987). But the proximate preference is still institutional, not substantive.

6. 514 U.S. 549 (1995). For examples of the near-total deference to Congress before Lopez, see Hodel v. Virginia Surface Mining & Reclamation Ass'n, 452 U.S. 264, 276-80, 11 ELR 20569, 20571-73 (1981) (applying rational basis test, which requires courts to defer to congressional findings that regulated activity affects interstate commerce); Perez v. United States, 402 U.S. 146, 155-56 (1971) (same); Heart of Atlanta Motel, Inc. v. United States, 379 U.S. 241, 252-53 (1964) (same); Katzenbach v. McClung, 379 U.S. 294, 299-301 (1964).

7. 529 U.S. 598 (2000).

8. See Lopez, 514 U.S. at 597-602 (Thomas, J., concurring) (advocating judicial enforcement of a distinction between "commerce" and "manufacturing").

9. 317 U.S. 111 (1942).

10. See id. at 127-28; John Copeland Nagle, The Commerce Clause Meets the Delhi Sands Flower-Loving Fly, 97 MICH. L. REV. 174, 193 (1998) (positing that Wickard "has come to stand for" aggregation).

11. See Morrison, 529 U.S. at 673 ("While we need not adopt a categorical rule against aggregating the effects of any noneconomic activity in order to decide these cases, thus far in our Nation's history our cases have upheld Commerce Clause regulation of intrastate activity only where that activity is economic in nature.").

12. Nagle, supra note 10, at 200-01.

13. 234 U.S. 342 (1914).

14. 452 U.S. 314, 11 ELR 20581 (1981).

15. Id. at 329 n.17, 11 ELR at 20585 n.17.

16. Lopez, 514 U.S. at 561.

17. See, e.g., Gibbs v. Babbitt, 214 F.3d 483, 491, 30 ELR 20602, 20604 (4th Cir. 2000) (quoting the passage after Morrison); United States v. Kirk, 105 F.3d 997, 1016 (5th Cir. 1997) (quoting the passage before Morrison).

18. 214 F.3d 483, 30 ELR 20602 (4th Cir. 2000).

19. See Morrison, 529 U.S. at 671 (quoting Lopez, 514 U.S. at 558-59).

20. See Gibbs, 214 F.3d at 491-99, 11 ELR at 20604-07.

21. See id. at 493, 11 ELR at 20605 ("While the taking of one red wolf on private land may not be 'substantial,' the takings of red wolves in the aggregate have a sufficient impact on interstate commerce to uphold this regulation.").

22. See id. at 492, 11 ELR at 20605.

23. See Morrison, 529 U.S. at 676. Justice Stephen H. Breyer's dissent reads the majority as permitting two exceptions to this rule: (1) Congress may aggregate noneconomic activity that takes place at economic establishments; and (2) Congress may regulate intrastate noneconomic activities where the regulation is an essential part of a comprehensive national regulatory scheme. See id. at 700 (Breyer, J., dissenting). I discuss the second exception below. See infra notes 35-40 and accompanying text.

24. See Gibbs, 214 F.3d at 492, 11 ELR at 20605.

25. See id. at 497-99, 11 ELR at 20607.

26. Id. at 497, 11 ELR at 20607 (quoting United States v. Lopez, 514 U.S. 549, 561 (1995)).

27. Id. at 498, 11 ELR at 20607.

28. For the text of the passage from Lopez, see supra note 16 and accompanying text.

29. See United States v. Kirk, 105 F.3d 997, 1013-14 (5th Cir. 1997) (en banc) (Jones, J., dissenting) (distinguishing regulations of economic activity that substantially affect interstate commerce from regulations of simple "activity" essential to maintaining a larger regime of interstate economic regulation).

30. Morrison, 529 U.S. at 700 (Breyer, J., dissenting).

31. 18 U.S.C. § 922(q)(1)(A) (1988 & Supp. V).

32. Note that the order in which the pairwise votes were taken is completely immaterial. Proposal 2 is a "Condorcet winner" — it defeats all other proposals in pairwise competition. There is no voting cycle here and no aggregation problem. The preferences of the individual legislators and the group-level preferences are all perfectly well behaved.

33. The analysis would be entirely different, of course, if we replace legislator C with a legislator D, who has the preferences (2">3">1). In that case the ultimate winner would be Proposal 3, no federal regulation, rather than Proposal 1. But that is not inconsistent with my claim, which is just that Commerce Clause review can in principle have perverse effects, so long as legislators' preferences are arranged in the way illustrated in the text. Which arrangement of legislators' preferences actually obtains in any given case is an empirical question; I discuss the empirical issues below.

34. See 18 U.S.C. § 922(o) (1994). The statute actually prohibits both "possession" and "transfer," but in the cases I discuss the latter prohibition was not at issue.

35. Pub. L. No. 99-308 (enacting prohibitions against gun possession and usage).

36. See United States v. Franklyn, 157 F.3d 90, 96 & n.3 (2d Cir. 1998) (upholding § 922(o) of the Firearms Owners' Protection Act under the Commerce Clause); United States v. Wright, 117 F.3d 1265, 1270 (11th Cir. 1997) (same), vacated on other grounds, 133 F.3d 1412 (1998); United States v. Knutson, 113 F.3d 27, 30-31 (5th Cir. 1997) (per curium) (same); United States v. Rybar, 103 F.3d 273, 285 (3d Cir. 1996) (same); United States v. Beuckelaere, 91 F.3d 781, 784-87 (6th Cir. 1996) (same); United States v. Kenney, 91 F.3d 884, 891 (7th Cir. 1996) (same); United States v. Rambo, 74 F.3d 948, 952 (9th Cir. 1995) (same); United States v. Wilks, 58 F.3d 1518, 1521-22 (10th Cir. 1995) (same).

37. Rybar, 103 F.3d at 282.

38. Franklyn, 157 F.3d at 94.

39. Thus the Gun-Free School Zones Act could not have been upheld on the analysis discussed here.

40. See generally Morrison, 529 U.S. at 682 (Thomas, J., concurring); Lopez, 514 U.S. at 549 (Thomas, J., concurring).

41. 42 U.S.C. § 13981 (1994) (enacted as part of the VAWA); 18 U.S.C. § 922(q)(1)(A) (1988 & Supp. V) (enacted as part of Gun-Free School Zones Act of 1990).

42. 16 U.S.C. § 1538(a), ELR STAT. ESA § 9(a).

43. 483 U.S. 825, 17 ELR 20918 (1987).

44. See William A. Fischel, Introduction: Utilitarian Balancing and Formalism in Takings, 88 COLUM. L. REV. 1581, 1583 (1988) (suggesting prohibitions on conditions are barriers to land improvement).

45. See Saenz v. Roe, 526 U.S. 489 (1999).

46. 317 U.S. at 111.

47. 234 U.S. at 342.

48. 452 U.S. at 314, 11 ELR at 20581.

49. See Wickard, 317 U.S. at 128.