The Federalism Project


American Enterprise Institute

Thoughts on Federalism
No. 1 July 7, 2000

How to Promote Federalism and How Not to

by Michael Greve


Earlier this summer, Congressman David McIntosh (R-IN) and Sen. Fred Thompson (R-TN) introduced legislation to promote federalism and, in particular, to enhance congressional responsibility for the federal preemption of state and local laws. Both bills attracted bi-partisan sponsorship and made rapid progress in committee. In August, however, both hit a brick wall. They have now been withdrawn.

Instead of despairing, federalism's friends ought to learn a lesson. They should abandon generic, across-the-board federalism strategies and instead devote their energies to advancing more limited but more substantive, meaningful federalism initiatives in selected policy arenas.

The Political Economy of Preemption

The core of the McIntosh and the Thompson bills, and the cause of their early demise, is a rule of statutory construction: with some exceptions (prominently, anti-discrimination statutes), laws or regulations enacted after the effective date of the bill shall not be construed to preempt state or local law, unless there is a direct conflict between federal and state or local law (so that it is physically impossible to comply with both) or unless Congress has explicitly stated its intent to preempt state or local law.

This injunction against "implied preemption" prompted the Chamber of Commerce and the Business Roundtable to oppose the federalism bills. In so doing, the business community joined the regulators, as well as the environmental movement and its congressional patrons (notably, Henry Waxman [D-CA]). The only supporters are the "Big Seven" local and state government organizations that drafted and pushed the bills, along with some conservative think tanks.

Now that the bills have died, sensible people can and should consider the merits of the controversy. Whose side are we on?

The Case for McIntosh/Thompson. The basic argument for the prohibition on implied preemption is its tendency to drive up legislative transaction costs. When Congress can "imply" and mumble, it will find it easy to cut deals--among other things, because all the concerned interest groups get a second and third bite at the apple (in the regulatory agency, and in court). In contrast, when Congress must legislate clearly and responsibly (as Thompson-McIntosh would compel it to do), it will often not legislate at all--among other reasons, because the burdened interests, knowing that they have it coming, will put up a fight.

A judicial canon of construction already forbids implied federal repeals of the states' sovereign immunity, and it has had the intended effect of curbing such impositions. The McIntosh-Thompson provision against implied preemption follows that model, and would tend to produce the same effect.

The Case Against. Foremost among the business community's arguments against McIntosh/Thompson is experience. The National Labor Relations Act is a conservative-libertarian bete noir, but wholesale, often implied preemption under the Act has prevented the states from imposing European-style labor laws--which is just about the only good thing to be said for the act. So with ERISA. So with securities law. And tobacco regulation. And pesticides. And drug labeling.

Behind this experience lies logic. Federal laws and regulations that are more stringent than state or local rules rarely pose preemption problems: the federal rules govern. Preemption issues typically arise when federal rules are less demanding of regulated industries than are state or local regulations. It is then that business wants to argue, for perfectly intelligible reasons, that federal rules preclude more extensive state or local regulation.

One can find a few historical examples to the contrary, and one can imagine, prospectively, circumstances under which regulated industries might benefit from a prohibition on implied preemption. But business organizations are well advised to attend to the general run of cases--especially when their experience coheres with the basic insights of the political economy that governs states and localities.

The Case Against the States. Why are the "Big Seven" state and local organizations so keen on curbing (implied) preemption? Simple: they resent federal boundaries to their regulatory empires. As a rule, the states do not terribly mind preemption in the form of a federal "floor" of regulation (that is to say, federal rules that are more stringent than state or local rules): after all, competing jurisdictions will be locked into the same rules. What the states do mind are federal preemptions in the form of a "ceiling" on regulatory experiments they otherwise would or might choose to run. The campaign against preemption is at bottom a campaign for cost-free collectivist tinkering.

Now What? Both McIntosh/Thompson and the business community have half of the federalism equation right--the former, in their insistence on legislative responsibility; the latter, in demanding a stopping point to statist experiments. These principles and protagonists ought to work in tandem. Both sides have good arguments. And yet, they are at loggerheads. Meanwhile, Henry Waxman, the Sierra Club, and the rest of the statist cabal--well, if they had to divide and conquer, that would actually be progress. But all they do, and all they have to do, is to sit back and laugh.

Both regulated industries and conservative constitutionalists have too much at stake to let this go on. They should figure out what went wrong here--and look for federalism initiatives on common ground.

Two Views of Federalism

I have elsewhere distinguished a faux (phony/ersatz/statist) federalism from the real (constitutional/competitive) thing. (See Greve, Real Federalism, Ch. 1 et pass.) A brief summary and extension to the preemption issue:

Faux Federalism says something like the following: three different levels of government offer their regulatory services to the highest bidder. The feds' power is plenary, meaning that Congress can satisfy demands for regulatory services in any arena, on any issue. Private conduct will be governed by the most stringent rule extant (federal, state, or local)--which will then serve as the floor for the next round of regulation. Thus, the states serve as "laboratories of democracy" in the sense of experimenting with collectivist schemes. As noted, a federal preemption "ceiling" tends to limit such experimentation. Hence, the Big Seven's push for McIntosh/ Thompson, which makes it harder for the feds to construct a ceiling.

Real Federalism, in contrast, begins and ends with the constitutional premise of enumerated powers. Within its constitutionally enumerated powers, Congress can preempt absolutely everything--for instance, and prominently, things and persons that move in interstate commerce. Outside that realm, however, Congress can preempt nothing--not even with the clearest statement of intent.

Under this constitutional arrangement, the states will still experiment; in fact, they will have to. But the experimentation with all the stuff that's beyond the purview of Congress will take place under competitive constraints. Assume (correctly, to my mind) that Congress may not regulate purely intrastate matters--say, the use of mud puddles and other local "wetlands." States that regulate unreasonably will at the margin repel investors. Interstate competition for the citizens' talents, assets, and business will act as a powerful constraint on collectivist experiments and interest group transfers--far more powerful than some rule about federal preemption.

Strategic Implications

Despite the foregoing, it would not be wise to oppose "faux federalism" in every instance and to push "real federalism" full-bore, in every instance. The better bet is to (1) examine--and support or reject--"faux federalism" initiatives for the tendency of their effect; (2) push real federalism initiatives in selected policy areas; and (3) hold out real federalism as an attractive alternative to the second-best (and worst) policy regimes.

Faux federalism can be a step towards the real thing. E.g., judicial decisions to the effect that the Americans With Disabilities Act (or the ADEA, or the Family and Medical Leave Act) can't be enforced through private damage suits against the states merit two cheers. While such rulings put the states (as employers) at an undeserved and unwarranted advantage vis-a-vis private employers, they also signal that a solicitude for special-interest constituencies can't override constitutional constraints. That is a first step towards sanity.

Arguably, one can say the same about McIntosh/Thompson: saying that Congress can't preempt implicitly is a step to saying it can't preempt explicitly, either. But that inference is in fact highly questionable. Moreover, and in any event, the price of splitting constitutionalists from business--when business has adopted an anti-regulatory position--is too high.

Real federalism, on the other hand, cannot be promoted wholesale, generically, and as a general proposition in the legislative arena. One, there are no votes for any challenge to the premise of congressional omnipotence. (Conversely, any generic law to restrict the powers of Congress will, by definition, not be serious; otherwise, it would not have been enacted in the first place.) Two, there is the still-widespread belief that interstate competition amounts to a "race to the bottom," meaning that a limit to congressional authority will turn the entire country into one vast Mississippi.

In this light, we should concentrate on federalism initiatives that take the feds out of some policy arena (or preclude them from entering it) and, in the process, force the states to compete. In the legislative arena, the emphasis shouldn't be on pushing real federalism as a constitutional doctrine but rather to push--and to illustrate the benefits of--policies that (by the way!) happen to be consistent with federalism (whereas the unconstitutional, centralized regimes would wreak real havoc). Such initiatives will (a) tend to unite constitutionalists with (most of) business and pit both against regulators and redistributionists and (b) undermine the "race to the bottom" rationale.

Put differently: just as school choice need not be (and has not been) advanced all at once or as a constitutional mandate to make real headway, so federalism need not be advanced as a constitutional mandate. In each case, successful experiments pave the way. Just as welfare reform was a partial federalism experiment that has increased public confidence in federalism, so real federalism experiments in other arenas need not be perfect. And, just as genuine federalism advances have happened on "social" issues no federal legislator wants to handle, so one might wish to concentrate on economic/regulatory issues that have proven an embarrassment and inconvenience to all sides.

First-Best Solutions. Constitutionalists and business have gotten into a tiff over preemption because the option of real federalism--firm limits to congressional power--isn't even on the table. The incident confirms a general pattern that we should attempt to break--again, not all at once, but on selected issues.

Consider tort reform--not a far-fetched example, since it was in fact a major cause of the disagreement on McIntosh/ Thompson. Business argues for facilitating federal intervention (by means of easy preemption) in the interest of uniformity: manufacturers or service providers can't be expected to comply with fifty different product standards. Constitutionalists are nervous about preempting the entire field.

In tort law, plaintiffs' lawyers trawl the country in search of a hospitable jurisdiction, and the craziest jury in Beaumont or Texarkana gets to determine the legal regime for the rest of the country. (Talk about a race to the bottom.) So business lobbies for "uniformity" and federal preemption.

However, uniformity is preferable to diversity and choice only so long as the choices are being made by one's enemies. We have fifty different regimes governing corporate chartering. Yet one does not hear the business community yammering about a lack of uniformity--au contraire. The reason is that the firms, not their customers, get to choose the jurisdiction whose laws will govern their transactions with customers. Firms sort themselves into the jurisdictions that are most to their and their shareholders' liking. Diversity--lack of uniformity--is an advantage for business (and shareholders), not a liability. It also turns out to be the socially optimal regime, as Ralph Winter and Roberta Romano have shown.

Thus, the best, real federalism tort regime would let producers, rather than plaintiffs' lawyers, choose the jurisdiction. America once had such choice-of-law and jurisdictional rules, and it is plainly within Congress's purview to resurrect them. (Congress would be regulating stuff that crosses state lines). Proposals to this effect, alas, aren't even on the table. Well: put them on the table--if not as immediate reform proposals, then at least as items for discussion.

A Final Note on States

Among the biggest obstacles to a promising federalism agenda is a pronounced conservative sentimentality about "the states"--exemplified by an obsession with "federalism impact statements," the Unfunded Mandates Reform Act, and (most recently) McIntosh/ Thompson.

In some respects, on some issues, there is something to the intuition that a united front of state and local governments will help to promote federalism. As already suggested, however, the interest that principally unites state and local entities is a desire to remove federal barriers to organized looting. (The systematic shake-downs of tobacco and gun manufacturers illustrate the general dynamics.) Unity among these characters isn't a chance for federalism but a clear sign of mischief-in-the-making.

Decades ago, the conservative fantasy of a "business community" gave way to the more sophisticated understanding that just about every issue confronts real competitors with losers who clamor for protection. By that same token, proposals for real, competitive federalism will divide states who think they can compete from those who believe that they are better off under a centralized, federal regime. Real federalism will always attract some competitive states. We should seek out the competitors among the states in the same way in which the Left seeks out the rent-seekers among the corporations.


If federalism is to make headway in Congress, its friends must learn to be much less sentimental about state and local prerogatives--and much more attuned to federalism's purposes and politics. They must also understand that, on economic issues, there is no strategic force or constituency for federalism per se; there are only tactical alliances on this or that issue. These considerations counsel for skepticism vis-a-vis generic federalism initiatives. Careful issue selection and case-by-case coalition building seem a more promising approach.