The Federalism Project

American Enterprise Institute

Draft Version: Please do not cite without permission


Federalism and Insurance Regulation

Remarks Prepared for the Federalist Society’s Annual Lawyers Convention, panel on insurance regulation, November 15, 2001 

Michael S. Greve
Director, American Enterprise Institute’s Federalism Project


Since I do not actually know anything about financial markets and regulation, Peter Wallison and my distinguished fellow-panelists graciously acceded to my request for a conceptual assignment, rather than a policy-reform talk.

I will argue that there is no necessary conflict between federalism and efficient insurance regulation; it all depends on what kind of federalism. I will distinguish a bad, states’ rights federalism, which does present such a conflict, from a good, competitive federalism, which does not. I have tried to develop that distinction and argument in other contexts. With my trusted federalism hammer in hand, the forebodingly complex and confusing subject matter looks like just another nail. Having hit my nail one more time, I will make a concession to the topic at hand and argue that proposals for an optional federal insurance charter are in principle consistent with competitive federalism, though not a necessary part of it. Finally, I’ll argue that Gramm-Leach-Bliley may represent a step towards, and an opening for, competitive federalism—or the opposite.

Federalism means that states will attempt to protect and favor their own citizens and companies vis-a-vis outsiders. Bad, states rights federalism lets them get away with it. First, bad federalism allows states enact statutes that discriminate against outsiders, openly or covertly. Second, bad federalism rests on a “destination” principle: in interstate transactions, it is the law of the customer’s or buyer’s state, not the law of the producer or seller, that governs the transaction. This means that companies—regardless of their domicile—must comply with the rules and regulations in each of the fifty states where they do business.

This, by and large, is the “federalism” that we have in insurance, except that it’s worse. Obviously, it is incompatible with a common market; in fact, it is a regulatory balkans. For the dangers and disadvantages of a regulatory balkans, see the actual Balkans.

States’ rights federalism, however, is not the only possible federalism (and wholesale nationalization is not the only alternative to bad decentralized regulation). Also possible, and much closer to the Founders’ vision, is a competitive federalism. Instead of facilitating parochial protectionism, competitive federalism integrates markets and compels states to compete for citizens’ and companies’ business.

Competitive federalism’s structural principles must be supplied either by the federal Constitution or through federal law. Let me mention four such principles and contrast them with current practice:

Non-discrimination. States must be barred from enacting laws that discriminate against outside parties. To the Founders’ minds, this command constituted a principal protection against state protectionism and impositions on a common market. It is expressed in a half-dozen constitutional provisions, from the Privileges and Immunities Clause to the Export-Import Clause and the Full Faith and Credit Clause. The “dormant” Commerce Clause has also come to be understood as an injunction against (facially) discriminatory state statutes that are not specifically authorized by Congress.

The glitch with a constitutional anti-discrimination principle is that it captures only blatantly discriminatory state laws, not the states’ countless “evasive maneuvers” (Madison) to disguise discrimination. That is why the Commerce Clause authorizes Congress to legislate against state discrimination and protectionism. In the insurance markets, however, Congress has failed to exercise this option and, in fact, done exactly the opposite: under the “reverse preemption” logic of the McCarran-Ferguson Act, blatantly discriminatory and protectionist state laws that would otherwise violate the dormant Commerce Clause are rendered lawful (subject only to very weak equal protection constraints).[1]

Exit. Competitive federalism requires free exit. Citizens and companies must be able to vote with their feet and their pocketbooks—companies, by choosing their domiciles and the places where they do business; citizens, by choosing products that, under competitive federalism, come along with the legal rules of the product’s or service’s origin state. The bedrock principle of American federalism is that citizens choose their state.

The bedrock principle of insurance regulation is the reverse: no exit. A California citizen is stuck with California’s consumer protections, whether he wants them (on the books and in the product price) or not. Corporate changes of domicile are virtually unknown, both on account of regulatory obstacles and because threats of state retaliation appear to be widespread or at least widely feared by insurance companies that have contemplated redomestication. Several states even administer exit restrictions that preclude states from discontinuing product lines that have been rendered unprofitable by state regulation. We might as well legalize confiscation.[2] 

Reciprocity. Competitive federalism requires reciprocity, meaning that each state will respect and enforce the legal rights (charters, licenses, and so forth) obtained in another state. Under this “origin principle” (or “principle of mutual recognition,” as it is called in the European Union), each company need comply with only one set of legal rules—those of its domicile state. Insurance regulation operates on the opposite principle: the customer’s law governs. That is an annoyance in personal lines, where agents sell policies in one or two states; it is a much, much bigger deal in commercial lines, where the producers’ compliance obligations in multiple jurisdictions drive expense ratios (of cost to premiums) through the roof.

Choice of Law and Federal Courts. Competition among governments requires that citizens and companies must be able to rely on the application of the law of the state that they have chosen, either by domiciling in that state or by contractual choice-of-law and choice-of-forum clauses. Put bluntly, market participants need protection against state attorneys general and trial lawyers who will attempt to “home cook” out-of-state insurers in front of a home state jury and judge of their own choosing. Realistically, this requires a federal choice-of-law default rule designating the domicile state’s law as governing multistate transactions; uncompromising enforcement of contractual choice-of-law and forum clauses; and access to federal courts in all cases of minimum diversity. None of these conditions exist in the insurance markets. 

Corporate chartering in the United States provides an actual example of competitive federalism in operation: companies freely choose their state charter, which will then govern its legal relations with shareholders everywhere. In principle, this Delaware model can be applied to insurance regulation; the difficulty, I think, (apart from building sufficient political support) lies in specifying the appropriate range of its application.

Solvency requirements would appear an obvious candidate for competitive state arrangements: much like big, well-informed institutional investors induce corporations to choose state charters that maximize shareholder value (in contrast to entrenching management), so profit-maximizing financial institutions would presumably choose states with optimal (rather than excessively lenient) solvency  requirements and arrangements.

Licensing and registration requirements are likewise plausible candidates, and full reciprocity in this area would eliminate the compelling complaints over the existing, protectionist system and its inordinate compliance costs. Full reciprocity for new products would solve speed-to-market issues and facilitate more efficient competition across industries.

In these areas, the competitive, mutual recognition principle threatens to undermine state consumer protection statutes and the rationales on which they are based (such as consumers’ difficulty in understanding very complicated financial instruments). But concerns over adhesion, asymmetric information, and unequal bargaining power do not apply in the area of commercial lines, where bargaining parity prevails and marginal purchasers can be presumed to be reasonably informed. Thus, the case for full reciprocity and state competition seems most plausible in precisely the area where those principles would do the most good.

A hypothetical federal insurance reform that would mandate non-discrimination, free exit, full reciprocity, and access to an impartial forum in multi-state conflicts would be fully federalist: it would not materially, substantively regulate the business of insurance at all. States, not the federal government, would continue to do so. In contrast to states-rights federalism, however, competitive federalism poses no conflict with open markets and efficiency concerns. Reciprocity ensures that no company is subject to more than one set of regulations; exit and non-discrimination ensure that companies may operate, and citizens may choose, everywhere. Differences between the state regimes aren’t politically “harmonized”; they are arbitraged through consumer and producer choice.

Arguably, such a reform might be more efficient than a uniform, substantive federal reform. We are very unlikely to find efficient regulation by drawing it on a piece of paper, and any mistake we make at the national level will by definition be a very big, nation-wide mistake. Actual state experimentation is a better bet—again, under suitably defined, competitive conditions, as distinct from a protectionist “states’ rights” regime that treats private market participants as guinea pigs. That is so especially since the federal government is less disciplined by the threat of market exit, and therefore more prone to interest group racketeering and wealth transfers in the name of “reform,” than are states that operate under competitive conditions. 

The competitive federalist will often despair of the political viability of his position. Competitive federalism substitutes private arbitrage for political harmonization and thus wrings the rents out of the system, which means that no one with a say on the matter—interests, advocates, regulators at every level—want any part of it. In the insurance context, however, the much-discussed option of an optional federal charter may provide an opening.

An optional federal charter follows the logic of competitive federalism: it offers producers and, therefore, consumers a choice between competing legal regimes (and may the best regime win)—except that the choice takes place along a vertical rather than horizontal dimension, and that only two rather than fifty choices are being offered. Thus, no serious federalist should reject the proposal as unduly “nationalist”: it is not, at least not in principle. I would only caution that the devil lies in the details. I suspect that it might be hard to mimic the principles of horizontal competition among equals—non-discrimination, free exit (both ways), full reciprocity, and a reliable, impartial forum—in competition between two inherently unequal partners. On the one hand, the federal charters must not be allowed to disintegrate into a kind of federal “floor” of minimum standards, over and above which states may still regulate; they must be a clean, unimpeded choice. Such unequivocal preemptions are extremely rare in federal law, and they may be especially hard to come by in the insurance setting. On the other hand, effective vertical competition requires some credible assurance that the senior federal partner will in fact tolerate, over time, the competition it has just established. But while these difficulties may prove considerable or even intractable, they do not constitute an objection on federalism grounds.

Since optional federal chartering follows competitive federalism’s general logic, it is fully consistent with that system. Under a fully competitive state regime, the additional option of a federal charter would simply transform the federal government into the 51st state. Obviously, that option is not strictly necessary: state competition, theoretically, does all the work that an optional federal charter is meant to do. There may be good reasons for providing that option even under a fully competitive regime, and there may be good arguments against it. But again, these considerations have to do with policy, not with federalism.

To give credit where some small measure of credit is due, the potential attraction of reciprocity has not entirely escaped federal legislatures. Title III of the Gramm-Leach-Bliley Act prods the states towards mutual recognition of insurance licenses.[3] Should a sufficient number of states fail to achieve that objective by November 2002, GLB threatens the establishment of a supra-state agency, under the supervision of the NAIC and, failing that, directly under the federal government, to bring about the statutory objectives. That is a step towards a more competitive environment—but it is a small step and, moreover, the sort of small step that may make matters worse.

The problem is not so much that the statute exempts some of the most odious, protectionist restrictions (notably, countersignature requirements), nor that it does not encompass all states, nor even that states have appeared to make little progress. The central problem, rather, is that federal legislators, state regulators, and certainly the NAIC as a collective body appear to view reciprocity as a first step towards “harmonization.” As suggested, however, the two principles are polar opposites, and mutually exclusive means of market integration. Reciprocity works through private arbitrage; “harmonization,” through political negotiation. Reciprocity means decentralization; harmonization means a single, uniform, central rule. Reciprocity implies jurisdictional competition; harmonization, a policy cartel that suppresses competition.

GLB entrusts the administration of that cartel to the NAIC and perhaps a future NARAB, rather than an actual federal agency. That concession to states’ rights, however, exacts a heavy price in terms of political accountability and transparency. The NAIC’s principal harmonization vehicle, for example, is a “Producer Licensing Model Act.” It can be obtained from the NAIC—at the modest price of $650. Not exactly a bargain for the ordinary citizens for whose protection these markets are ostensibly regulated.

GLB thus threatens to exemplify a depressing scenario that is being played out in a growing number of policy areas, from e-taxation to nuclear waste management. Some states can move on their own towards reciprocity, but that doesn’t help very much so long as California insists on being California. In other words, the states can’t on their own overcome free-rider and hold-out problems, meaning that balkanization persists. Congress recognizes as much. But instead of trumping the states’-rights regime with federal rules, it pays ostensible deference to state autonomy and “federalism” and nudges, bullies, or bludgeons the states into “harmonizing” cartels and compacts, under the authority of a supra-state authority that is off the constitutional charts, that is accountable to no one, and that no one outside the regulated industries has ever heard of. NARAB was certainly news to me[4]

The cleaner, competitive solution would be to build on the GLB reciprocity provisions—to make them mandatory rather than hortatory; to cover all states and all licensing requirements; to cover at least commercial product lines; to add federal prohibitions against discriminatory and “no-exit” provisions at the state level; to relieve the NAIC of its harmonizing obligations; and to put the envisioned NARAB out of its misery before a federal court strikes it down as an extra-constitutional abomination.

Against my better instincts and initial disclaimers, I seem to have ended up with what looks like an actual reform proposal. Instead of actually defending it, let me end on a more modest, but hopefully helpful note: there is a real, competitive federalism that is fully consistent with the Founders’ dual commitment to decentralization and to a common market. But the federalism we actually have in insurance isn’t federalism; it is its parochial, evil twin. There are lots of reasons to be cautious with federal interventions in financial markets, and lots of ways to get confused about it. But the last thing that should deter or confuse reformers are false appeals to a false federalism.

[1] See Prudential Insurance Co. v. Benjamin, 328 U.S. 408 (1946); Metropolitan Life Ins. Co. v. Ward, 470 U.S. 869 (1985) (equal protection constraint). 

[2] See generally Peter Wallison, ed., Optional Federal Chartering. And Regulation of Insurance Companies, (AEI, 2000); and, on exit restrictions, Richard A. Epstein, “Exit Rights Under Federalism,” Law & Contemp. Probs (1992).

[3] The redomestication provisions of the statute also embody an element of competitive federalism. The provisions, however, appear not to have been used, and I will ignore them for the purposes at hand.

[4] The experts in this room are of course aware of the NAIC and its activities. Many of you, however, are also governed by the Ozone Transport Commission and soon, if the states’ rights apostles in Congress have their way, by a Simplified Sales Tax Project. Familiarity with one supra-state agency in nowhere land, or even an ability to do deals with it, provides no reason for complacency about the broader trend.