There are many challenges in designing a federalist system of government. Perhaps the most daunting is how to create incentives for government officials to preserve a regime of state-by-state decisionmaking—especially when constituent pressures, partisan allegiance, or ideological beliefs tug in other directions.

The U.S. Constitution tries to preserve state prerogatives by enumerating the powers of the federal government and announcing that all non-enumerated powers are left to the states.[1] But “parchment barriers” of this sort are a weak and ineffective means of preserving state authority. Words are always subject to interpretation, even when government officials are acting in good faith. And phrases like “necessary and proper,” “appropriate legislation,” and “commerce among the several States” are plastic enough to allow federal lawmakers to credibly claim constitutional support for almost all of their preferred policies—and to find judicial appointees who will uphold those policies.

The Constitution also tries to protect state decisionmaking by making it exceedingly difficult to enact federal statutes, treaties, or constitutional amendments. Article I’s bicameralism-and-presentment requirements prevent any bill from becoming law unless it obtains approval from three separate institutions—the House, the Senate, and the President—or secures a two-thirds supermajority in the House and the Senate to override a presidential veto. Article II requires the President and two-thirds of the Senate to concur in a treaty before it can be ratified. And Article V requires new constitutional amendments to win the approval of two-thirds of the members of each of house of Congress and ratification by three-fourths of the states.

But these state-protecting mechanisms are also subject to evasion. The rise of the administrative state allows Presidents and agencies to impose nationwide policies unilaterally, without running the bicameralism-and-presentment gauntlet. Sometimes the courts will block these efforts if they find that they lack sufficient statutory authorization. But Presidents get to appoint federal judges, and they seek to appoint judges who will defer to their administrative edicts.[2]

The prevalence of living-constitution interpretive methodologies has also empowered judges to impose all sorts of nationwide policies under the aegis of constitutional interpretation—even when these court-imposed policies have no basis in constitutional text and would never have obtained the supermajorities needed to enact a federal statute or constitutional amendment.

All of this shows that one cannot count on government officials to promote federalism simply for its own sake. And that remains true even though there will often be systemic benefits that arise from leaving matters to the states. The benefits of state-by-state decisionmaking have been well-rehearsed: it encourages innovation and experimentation; it produces information by enabling policymakers and analysts to study the effects of different policy choices; and it increases political-preference satisfaction by allowing states to tailor policies to the wishes of their electorates.[3] But even politicians and judges who understand the benefits of decentralized decisionmaking will occasionally face countervailing incentives to impose nationwide, one-size-fits-all solutions to policy disputes. This can arise from a desire to please important constituencies, to advance the agenda of one’s political party, or from a dogmatic belief in the rightness of one’s cause.

Professor Nolette’s Liberty Forum essay shows how even state officials face incentives to subvert the cause of federalism—and he is surely right to reject the “outmoded myth” that state institutions will instinctively push back against the expansion of federal power and centralized decisionmaking.[4] Many different factors influence the behavior of state officials, especially state attorneys general who are often elected by the voters and aspire to higher office. And sometimes those factors will induce state AGs to subordinate the values of federalism to other goals, such as punishing or deterring wrongdoers, enacting new regulatory regimes that they sincerely believe will enhance the public welfare, and, of course, boosting one’s public profile and electability to future office.

But one should be careful not to overstate the scope of the problem. Yes, state attorneys general will occasionally bring lawsuits and file amicus briefs that seek nationalized rather than localized solutions to policy disputes. But the claim that state attorneys general are “ruining” federalism with their activist behavior seems to me exaggerated.

First, there are serious limits to the AGs’ ability to impose a nationwide regulatory regime by suing and settling with private corporations. The first is that an AG-initiated lawsuit cannot get off the ground unless a company has engaged in illegal or tortious conduct. Companies that have done nothing wrong cannot be sued, so there is no roving authority for state attorneys general to regulate by lawsuit and settlement.

The second limit is that a settlement agreement binds only the parties to the case. It cannot regulate the behavior of new entrants that pop up after the settlement, nor does it govern the conduct of pre-existing companies that were not subject to the litigation.

The multistate tobacco settlement, for example, does not apply to companies that were not parties to the agreement. Those companies need not abide by the agreement’s restrictions on tobacco marketing, and they need not contribute to the settlement fund that the agreement established. As a result, new entrants can sell cigarettes at lower prices than the manufacturers that signed off on the agreement, and they can engage in marketing practices that are off-limits to their competitors.

Many states have responded to this by enacting “equity taxes” targeted at small tobacco companies,[5] and in 2009 Congress enacted legislation authorizing the U.S. Food and Drug Administration to regulate the marketing of all tobacco products, including products from companies not subject to the master settlement agreement.[6] But all of those laws were needed to fill the gaps in the settlement agreement—and agreements of that sort cannot establish an effective regulatory regime unless Congress and the legislatures of those states have the political will to back the settlement and extend its restrictions to all industry participants.

So there are serious drawbacks to pursuing regulatory goals through the multistate-settlement mechanism, and it is hardly an effective substitute for regulations enacted by legislative bodies.

Second, there are occasions in which state-by-state governance should give way to nationalized decisionmaking. No one believes state-by-state decisionmaking is always preferable to nationwide, one-size-fits-all solutions; the task is to find the optimal balance between nationalized and localized policymaking. The mere observation that state attorneys general have occasionally sought to impose nationwide regulatory regimes through litigation does not show that their behavior is “ruining” federalism. It’s quite possible that the AGs have pursued multistate settlements in cases where nationwide solutions were entirely appropriate.

Consider the tobacco situation. Before 1994, the tobacco companies had successfully rebuffed private lawsuits brought by smokers—but the tide was turning in the mid-1990s as newly uncovered documents showed that the industry had concealed the dangers of smoking and targeted its products to teenagers. Had the state attorneys general sat around and done nothing, the tobacco companies would have been left to fight off individual litigants on a case-by-case basis. It is far from clear that solution would have been good for federalism. It would have produced decades of litigation under the tort laws of 50 different states, state courts imposing different legal standards on companies who must market their products nationwide, massive amounts of forum shopping, and a race to the courthouse as litigants try to bring their claims before the companies file for bankruptcy.

None of this is to say that the tobacco master settlement is perfect; it has been criticized on many grounds, including the attacks made by Professor Nolette. But the alternative might have been even worse—and it likely would have demonstrated the pathologies that arise from leaving mass-tort litigation to be resolved on a state-by-state basis.[7]

Climate change is another issue on which state-by-state policymaking is likely to produce perverse outcomes. Greenhouse-gas emissions travel freely over state (and national) borders. And controls on those emissions are a public good that benefits everyone—regardless of whether a state has undertaken pollution controls of its own. That leaves the states with every incentive to free ride on the emission controls imposed by other states (and by other nations), and with little incentive to undertake individual action to curb greenhouse-gas emissions when each state’s marginal contribution to global warming is negligible.

The upshot is a massive collective-action problem that cannot be solved without federally mandated pollution controls. So the decision of state attorneys general to seek a federally imposed solution in Massachusetts v. EPA seems perfectly sensible and consistent with values of federalism—so long as one believes (as those state officials did) that the Clean Air Act is properly interpreted to require the regulation of greenhouse-gas emissions.

Finally, much of the litigation brought by state attorneys general has worked to protect state prerogatives against an overreaching federal government. A few weeks ago, a coalition of states obtained a nationwide injunction against the Obama administration’s transgender-restroom edicts, which sought to require school districts to allow any student who asserts a gender identity that departs from his or her biological sex into the restrooms, locker rooms, and showers of the opposite biological sex.[8] Lawsuits brought by state attorneys general have also blocked EPA regulations that rest on dubious constructions of statutory language.[9] And state-AG lawsuits have limited the states’ Medicaid obligations under the Affordable Care Act.[10]

Professor Nolette is surely correct to note that these lawsuits were motivated, at least in part, by partisan and ideological goals, rather than an abstract desire to advance the cause of competitive federalism—one need only to look at the captions to see that nearly every state official who signed on as a plaintiff was a Republican. But why is that a problem if the ultimate outcome of these lawsuits is to thwart the President’s efforts to rule through administrative decree? Surely the Democratic attorneys general will be equally vigilant when Republican Presidents seek to displace state prerogatives with rule by administrative agencies,[11] and these partisan motivations will ensure that someone will always be there to challenge an overreaching executive in court.

State attorney-general offices have the resources and motivation to challenge these federal initiatives in court, precisely because the state attorney general is a political creature who responds to political incentives. In many of these situations, the burdens imposed on private citizens will be too diffuse or speculative to justify the retention of counsel—or even to establish Article III injury.

So one has to look at the overall package before declaring the state attorneys general guilty of ruining federalism. Much of their litigation has sought to strengthen state prerogatives and local decisionmaking—and in the rare situations in which the state AGs have used litigation to impose nationalized policies, they have done so in areas where nationwide solutions seem clearly preferable to localized ones.

Partisan and electoral motivations of course remain, and they will continue to supply the impetus for much of a state attorney general’s behavior. And, as Professor Nolette candidly admits, there is nothing we can do about that. Politicians will always be responsive to constituent pressures, partisan allegiances, and their own ideological beliefs. But there is every reason to think that these motivations can be channeled to advance rather than undermine the cause of federalism. No political party is likely to win the White House while sweeping all 50 of the state attorney general positions. And so as long as there are opposition-party attorneys general in some of the states, the federal government will have every incentive to ensure that its initiatives have firm constitutional and statutory foundations before imposing those policies through legislation or administrative decree. This seems to me something that advocates of federalism should applaud and not bemoan.

[1] See Article I, Section 8 of the United States Constitution, and the Tenth Amendment.

[2] See G.G. v. Gloucester County School Board, 822 F.3d 709 (2016), in which the U.S. Fourth Circuit Court of Appeals held, in a 2 to 1 vote, that the federal courts must defer to an interpretation of Title IX’s implementing regulations that appears in an unpublished letter and that declares that school districts must allow transgender students to use the restrooms, locker rooms, and showers that match their “gender identity,” even if that gender identity departs from their biological sex; both judges in the majority were appointed by President Obama.

[3] See Michael W. McConnell, “Federalism: Evaluating the Founders’ Design,” University of Chicago Law Review 54 (1987), 1483.

[4] See, for example, Daryl J. Levinson, “Empire-Building Government in Constitutional Law,” Harvard Law Review 118 (2005), 916, and Daryl J. Levinson and Richard H. Pildes, “Separation of Parties, Not Powers,” Harvard Law Review 119 (2006), 2311.

[5] See, for example, Hegar v. Texas Small Tobacco Coalition, — S.W.3d. —, 2016 WL 1267843 (Tex. 2016).

[6] See the Family Smoking Prevention and Tobacco Control Act, Public Law 111–31, 123 Stat. 1776–1858 (2009).

[7] See Ortiz v. Fibreboard Corp., 527 U.S. 815, 821 (1999), observing that the asbestos litigation “defies customary judicial administration and calls for national legislation.”

[8] See Texas v. United States, — F. Supp. 3d —, 2016 WL 4426495 (N.D. Tex.).

[9] See Utility Air Regulatory Group v. Environmental Protection Agency, 134 S. Ct. 2427 (2014).

[10] See National Federation of Independent Business v. Sebelius, 132 S. Ct. 2566 (2012).

[11] See, for example, Gonzales v. Oregon, 546 U.S. 243 (2006).