The Yale Law Journal

April 2014

Negotiating Conflict Through Federalism: Institutional and Popular Perspectives

abstract. The contours of our federal system are under constant negotiation, as governments construct the scope of one another’s interests and powers while pursuing their agendas. For our institutions to manage these dynamics productively, we must understand the value the system is capable of generating. But no single conception of this value exists, because the virtues and costs of any particular federal-state relationship, in any given federalism controversy, will appear different depending on perspective: the federal, state, and even local will each perceive their own advantages. And none of these conceptions will map perfectly onto the people’s perceptions. In this essay, I attempt to answer the question of what federalism might be good for from each of these perspectives by considering how it has structured various regulatory and social controversies in recent years on matters such as immigration, marriage equality, drug policy, and health care reform. I focus on the administrative and enforcement judgments that each of these debates has required, in order to illuminate the discretionary spaces in which much of the work of federalism occurs. I argue that the value of the system common to all participants and that should govern the negotiation of inter-governmental relations is its creation of a framework for ongoing negotiation of differences large and small. In the spirit of this Feature, I emphasize that having many institutions with lawmaking power enables overlapping political communities to work toward national integration, while preserving governing spaces for meaningful disagreement when consensus fractures or proves elusive.

author. Professor of Law, Yale Law School. In developing this essay, I benefitted enormously from conversations with Jessica Bulman-Pozen, Heather Gerken, Abbe Gluck, Alison LaCroix, and Judith Resnik. I also am very grateful for the excellent research assistance of Trinity Brown, Alex Hemmer, and Samuel Kleiner.


The question of what value federalism generates has no single answer, nor does its corollary of how the system ought to be structured to maximize its virtues. The value generated by decentralized decision-making will appear different depending on the perspective adopted when considering the matter, as will the ideal design of inter-governmental relations. If we step inside the system itself and adopt an institutionalist point of view, the answers will reflect the interests of the system’s actors and take on partisan and bureaucratic characteristics. If we try to answer the questions externally, either from a popular or scholarly vantage point, the answers will become more ideological and normative. The question of federalism’s value breaks down into several inquiries: Of what value is it to the central government to have state and local governments to contend with? Of what value is it to state and local governments to be embedded in a system with a strong central government and myriad competing governments? Of what value is it to the people to have government power split and decentralized?

A reader reasonably could conclude that the first two questions may matter to a descriptive account of our federalism, but only the third perspective—the popular one—truly matters when debating the merits of the system. But the institutional perspectives convey important information about how the system functions, which in turn helps reveal the possibilities for governance the system creates. In this essay, I adopt each of the three perspectives outlined above in order to develop a complete sense of these possibilities. In so doing, I give content to one of the central insights of the work highlighted in this Feature—that federalism does not consist of a fixed set of relationships. Instead, its parameters are subject to ongoing negotiation by the players in the system, according to the advantages each might accrue from a particular set of relations.1 Understanding the substance of these negotiations should ultimately lay the groundwork for a normative account of inter-governmental relations.

None of the actors I name is unitary, of course, which makes defining any of their perspectives on a given federalism controversy, not to mention the system itself, tricky. The federal government consists of political and bureaucratic actors arrayed into different branches and representative of different political parties. The sub-federal consists of myriad state and local governments with varying degrees of autonomy from one another and from the center. And the “popular” perspective may be the most variegated and amorphous; it consists of a large and diverse assortment of interests woven together by economic, social, and cultural affiliations and through interest groups and networks such as political parties.

And yet, if we consider each of these actors in relation to one another, a distinct perspective on federalism’s value can be attributed to each. For the federal government, the virtues of the system include having states and localities to enlist in the expansion of its capacities, as well as having other governments to take the lead in various regulatory domains, for political and practical reasons. At the same time, pulling in the opposite direction will be the federal government’s interest in maintaining control (though not necessarily dominance) over the domains to which its powers extend. The federal interest can thus be served by managed diversity, rather than uniformity. State and local officials and entities will certainly have an interest in collaborating with the federal government and taking advantage of federal largesse, whether to advance their own or their parties’ political ambitions, or to help resolve the governance problems they face. But the decisional independence the federalist system affords will often be of political and policy value, too, and will be worth fighting to preserve, even when it places them at odds with the potential federal benefactor. From the popular point of view, federalism’s value will be harder to pinpoint, because the people themselves are not a bureaucratic institution set up in relief to other institutions. Instead, they consist of a sprawling agglomeration of diverse identities and interests. Whether the federal or the sub-federal should address a particular matter may depend on how the form of regulation maps onto people’s substantive preferences. In a more high-minded sense, the virtues of the system for the people will stem from the extent to which it advances the purposes of government.

Though pursuit of their interests by each player may often lead to conflict, particularly over which institutions should control any given policy domain, I argue that the value of the system common to all of its participants is the framework it creates for the ongoing negotiation of disagreements large and small—a value that requires regular attention by all participants to the integrity of federalism’s institutions. It is in this sense that I think federalism constitutes a framework for national integration, in the spirit of this Feature. It creates a multiplicity of institutions with lawmaking power through which to develop national consensus, while establishing a system of government that allows for meaningful expressions of disagreement when consensus fractures or proves elusive—a value that transcends perspective.

In what follows, I attempt to establish these conclusions by considering how the negotiations required by federalism have structured our national debates over a number of pressing social welfare issues, including immigration, marriage equality, drug policy, education and health care reform, and law enforcement. I focus on how these debates play out in what I call the discretionary spaces of federalism, which consist of the policy conversations and bureaucratic negotiations that actors within the system must have to figure out how to interact with one another both vertically and horizontally. Indeed, within existing legal constraints, state and local actors will have considerable room to maneuver, and the federal government considerable discretion to refrain from taking preemptive action.2 I highlight questions of administration and enforcement, because it is in these domains that the system’s actors construct one another’s powers and interests on an ongoing basis, based on the value they seek to derive from the system. In these discretionary spaces, “winners” must sometimes emerge from discrete conflicts, whether through judicial resolution or political concession, and the parameters set by courts and Congress obviously define the terrain of negotiation. But the inter-governmental relationships and overlapping political communities the system creates are neither locked in zero-sum competition nor bound by fixed rules of engagement, precisely what makes federalism productive regardless of perspective.

The question then becomes, what follows from this conception of federalism, according to which the value generated by federalism varies depending on who’s speaking. The talk of negotiation and the corresponding positionalist perspective on federalism work against the development of totalizing theories about how best to allocate power. It becomes difficult to move from describing the functions federalism performs in different domains to an overarching normative structural theory. As Heather Gerken has argued, many different federalisms exist.3 Even if we can identify a proper conception of federalism as a matter of original understanding, or of Supreme Court doctrine, that conception still will depend on the regulatory domain in question—the Court’s immigration federalism looks quite different from its economic federalism.4 More importantly, when we step into the discretionary spaces I describe, the “look” of the system will depend on the particular choices made by the players in question—choices that will depend on the advantages both the federal and the local perceive from either asserting or declining to use their power, which in turn will be motivated by political and partisan commitments.5

I ultimately believe we can still express proceduralist preferences for decentralized decision-making, regardless of the perspective adopted, based on observations about the value over time of such a structure. Elsewhere, I explore how decentralization can help simultaneously shape political consensus and channel ideological diversity.6 Here, I focus on the dynamics of negotiation in order to understand better the possibilities federalism creates for the players in the system. This conclusion does not preclude acknowledging that national institutions should be strong and sometimes cut off decentralized debate in the interest of protecting a national norm or the public good, or assert centralized authority to overcome regulatory dysfunctions.7 But, again, it does point in the direction of developing rules of engagement that keep federalism’s institutions robust.

This positionalist inquiry draws on my federalism-related work to date and dovetails with the approaches to federalism highlighted by the contributors to this Feature in at least two ways. First, in my work on immigration federalism, my central preoccupation has been to figure out what purposes state and local activity might serve in a traditionally federal domain, as well as to understand why state and local officials have claimed authority to act—whether those reasons differ from the concerns that motivate the federal government’s regulation.8 I have been concerned less with the identification and delineation of the scope of federal versus state power and more with understanding immigration federalism as an example of how the overlapping political communities in our body politic negotiate with one another to address matters of national concern.9 My efforts to understand the motivations behind and significance of sub-federal activity has entailed challenging broad conceptual assumptions of federal exclusivity with an appreciation of how deeply integrated the regulatory regime has become across levels of government, not to mention how fractured and decentralized the political conversation has been.

Second, in this same work, I have sought to highlight how decentralized ferment helps constitute our national debate on the subject of immigration. I have de-centered the national from the federal to explore how “national” issues—those whose salience cuts across state lines and constituencies—are not always or necessarily best served by a federal monopoly,10 and how state and local debate and regulation can serve national interests. In more recent work on federalism generally, I highlight these same dynamics of decentralized ferment and inter-institutional negotiation in other contexts of highly charged social policy—same-sex marriage, drug policy, and gun regulation—and argue that among the most significant values of our federal system is its utility in managing social conflict across policy domains.11 In line with the overarching frame of this Feature, I show how decentralized decision-making can promote national integration and national problem solving in a world of deep demographic and ideological diversity. But I also emphasize that integration does not depend on the achievement of a clear national consensus, though decentralized debate can certainly result in the consolidation of national norms or policies. Instead, integration can emerge through the achievement of an equilibrium that contains within it the possibility of ongoing debate—a picture of federalism I hope to illuminate here.

I. the federal government and the mixed interest in control

National debates can happen trans-locally with or without the federal government in the lead. Under this view, the federal government becomes one actor in the system. But because it has the authority to monopolize regulatory domains and displace states, localities, and even private actors, its conceptions of federalism will inform the character of the system more than any other single government. Considering how the federal government might perceive the value of federalism and therefore use its power—whether to consolidate its own positions, tamp down decentralized ferment, create room for diversity in regulation, or learn by observing how other actors address social problems—becomes crucial to understanding federalism as a governing structure.

It might seem futile to attempt to identify a federal perspective on federalism, given the complexity of the federal government.12 Congress and the Executive Branch will often have divergent interests, particularly when controlled by different parties. The myriad agencies of the federal government itself will also interact with one another according to logics of cooperation, as well as competition.13 The political layer of the bureaucracy may have different priorities from the civil service.14 This multiplicity is part of what gives federalism its negotiated character. As I explore in Part II, it creates opportunities for state and local actors. But it also enables the federal government to take advantage of federalism in numerous ways. And thus, even if no single perspective can be attributed to the federal government, we can still think in general terms about how a variegated center might approach the governments contained within it.

The federal government (both Congress and the Executive Branch) constructs its relationships with state and local actors in myriad ways: through delegation; by incorporating state and local officials into federal bodies, commissions, and regulatory regimes; or by crafting legislation or enforcement policies to address tensions that might arise when state and local actors exercise parallel but overlapping regulatory authority. Abbe Gluck has explored this dynamic in the healthcare context, revealing the varied ways in which the federal and state governments are intertwined through federal design.15 The Immigration and Nationality Act also balances delegation to and constraint on state authority and is full of different inter-governmental relationships designed by Congress to both employ and shape the federal structure to suit its policy objectives.16 In each sort of relationship, the Executive Branch must regularly assess how to calibrate its involvement and assertions of authority.17

Numerous factors will inform the federal government’s choices about how to interact with sub-federal legal and political institutions and bureaucracies. The existence of the latter can expand the federal government’s capacity to govern and enforce the law.18 The federalist structure also can amplify the influence of political parties and national politicians. Turning to its institutions can help federal actors advance their substantive agendas through lawmaking, either by locating a substitute for it at the state or local level, or by laying groundwork for future federal action.19 The federal system also enables federal actors to shift the burden of regulation and accountability for the handling of difficult issues to other officials and politicians.20 Each of these interests will likely be at work in the federal government’s “use” of the federal system.

But because the Constitution makes federal law supreme, working against these incentives to utilize the federal structure will be a strong impulse to maintain control over the domains to which federal power extends. When interacting with the federal system, particularly when attempting to use it to its advantage, the federal government frequently will face the questions of whether and how to assert its supremacy, particularly when dealing with questions of enforcement and administration. In the end, such control will be elusive and may even be counterproductive, as it is in tension with some of the strategies discussed above, but the desire to assert it will be an ever-present part of the federal government’s negotiation of the federalism dynamic.

The federal government’s decision to file a lawsuit challenging Arizona’s Senate Bill 1070, which contained various provisions designed to crack down on illegal immigration, reflects a particularly robust effort to assert control over a decentralized debate. Despite a longstanding practice of significant state and local involvement in various aspects of immigration enforcement, the federal government sought to assert its primacy within that system through litigation.21 From the Department of Homeland Security (DHS)’s law-enforcement point of view, the Arizona law abandoned cooperation with the federal government in favor of disruptive confrontation over how stringently to enforce the law.22 For the Civil Rights Division, and perhaps even the Attorney General, S.B. 1070 may have become emblematic of the risk of racial profiling23 inherent in police enforcement of immigration laws, thus requiring a federal response. Perhaps most important, Arizona’s challenge to the federal enforcement agenda, particularly when juxtaposed with the numerous other states that followed suit, threatened federal leadership of the immigration debate by offering a vision of immigration policy that relied on attrition through enforcement, directly challenging President Obama’s preferred legalization strategy.24

The federal government thus felt compelled to reassert its primacy over the politics and policy of immigration through the highly unusual strategy of a government-sponsored preemption lawsuit. In litigation, the administration characterized Arizona as having made itself a “rival decisionmaker[] based on disagreement with the focus and scope of federal enforcement,”25 reinforcing the premise that state and local participation in this domain could only be at the invitation of the federal government.26 The government’s rhetoric of control and supervision in the Arizona case could have just reflected its particularly strong interest in maintaining primacy in a traditional federal sphere. Despite the erosion of the spheres conception of federalism in doctrine and practice, the federal government might still cling to its traditional exclusivities. Or the government’s positions could have been a function of the especially oppositional behavior of Arizona and like-minded states. Whatever its motivation, the lawsuit suggests the federal government will often want its federalism both ways.

Developments in the drug policy arena suggest a slightly different, more cooperative conception of control—that federal primacy can be maintained and even served in the absence of monopoly. Numerous states have adopted regulatory regimes meaningfully distinct from the federal Controlled Substances Act (CSA), first by adopting exceptions to prohibitions on marijuana use for medicinal purposes,27 and more recently by decriminalizing use of marijuana even for recreational purposes while also authorizing its production and sale, as in Colorado28 and Washington State.29 Both sets of developments, but particularly the latter, could be said to disrupt the field of federal enforcement. The state laws sanction marijuana use, thus giving rise to an expanded market, and they deprive the federal government of the state and local enforcement resources on which it historically has relied to serve the purposes of the CSA.30

In response to these developments, the Obama DOJ has issued three memoranda articulating how the federal government intends to enforce federal law in the wake of state divergence. But what exactly those memoranda were meant to accomplish remains unclear, perhaps intentionally so. On the one hand, the memos have served to provide notice to state and local actors of federal intentions31—that consistent with its priorities the Department would continue to enforce the law. The Department identified certain federal “lines” that it would continue to police, namely preventing distribution of marijuana to minors; the diversion of revenues to cartels or criminal enterprises; and the use of violence in the cultivation and distribution of marijuana.32 But because its goals arguably have never included prosecuting minor possession offenses, the federal government could also announce its intention to look the other way with respect to much of the activity authorized by state law. This mixed commitment to enforcement likely reflects the imperatives facing a law enforcement agency, reconciled with political calculations concerning how best to respond to shifts in public opinion regarding marijuana use and possession.

But while asserting control essentially by reserving it, the federal government also struck a collaborative posture with states such as Colorado and Washington different in kind from its definition of cooperation in the immigration setting. Though similar in a certain spirit to the DHS guidance issued during the S.B. 1070 litigation laying out the forms of state and local assistance that constituted cooperation, the Department’s drug policy also went a step further by expressing a willingness to adjust federal policy in light of how the state regulatory regimes play out in practice.33 Indeed, the fact that Attorney General Holder rejected the aggressive preemption strategy adopted in the immigration setting suggests that, in some settings, federal purposes and even primacy can be maintained amidst cacophony on ultimate policy objectives.

This tolerance for divergence could simply reflect the government’s calculation that state laws will not disrupt the federal enforcement status quo, or it could stem from the federal government’s own long-term interest in de-escalating the drug war. The fact that a Democratic administration has taken this position also underscores the role politics can play in the federal government’s construction of its power and its relationships with the states. Whereas Arizona’s immigration law, adopted in a highly partisan Republican environment, was anathema to key political constituencies of the Obama Administration, the movement to relax marijuana prohibitions, particularly for medicinal purposes, does not threaten any such constituencies and may be in line with the substantive preferences of the average voting Democrat (or even American). Regardless of the reason for its position, the Department acknowledged that its priorities could be maintained in the absence of uniformity of policy.The very fact that the Cole memorandum suggests that the federal government will think about prosecutions in states with effective regulatory regimes differently from prosecutions in states that maintain prohibitions suggests that local development can prompt variegated federal action that nonetheless constitutes an integrated national enforcement strategy.

Both approaches to cooperation ultimately suggest that, no matter how strong the desire to assert it, control will be elusive, and sometimes even counterproductive, for at least three reasons. First, as I noted at the outset, the federal government historically has relied on state and local actors as agents in the development of federal policy, and the law is replete with delegation strategies of numerous sorts.34 Regardless of the particular form it takes,35 as a method of regulation, delegation carries with it the possibility of divergence from what the federal government, working through its own agents, might choose to do. The federal government’s embracing of the advantages of delegation also reflects a degree of acceptance of the divergences it generates. In some settings, the federal government will even have an affirmative interest in the divergence itself. In implementing the No Child Left Behind Act,36 for example, the Obama Administration has granted waivers of the law’s requirements to certain states. This strategy arguably helps facilitate states’ development of new strategies tailored to their own needs to meet the federal law’s requirements under federal supervision, thus using the tools of administration to simultaneously calibrate regulatory burdens in the states’ interests and advance federal policy objectives.37 We may bemoan this widespread delegation, following Michael Greve,38 as inconsistent with the original design whereby states were to provide a check on the federal government. Or, we may regard delegation strategies as reflecting lamentable path-dependencies that stem from the original constitutional design but might be less effective than comprehensive federal regulatory regimes.39 But the federal interest in making use of the structures of federalism requires constant adaptation and adjustment and therefore some limited loss of control.40

Second, because the federal government itself is not unitary in its views about how the law should be enforced, its responses to developments in the states can be complicated. Federal positions on federalism are often the product of internal compromise. The formulation of the DOJ marijuana memos was likely complicated by divergent interests that stemmed from institutional “role.” Whereas Attorney General Holder and other policy-makers might have seen the long-term benefits of de-escalation, the bureaucratic culture and professional interests of the officials of the DEA likely placed a heavy thumb on the scale in favor of enforcement vigilance, thus leading to the open, non-committal quality of the memos. Even the DHS memo on immigration cooperation, designed to point up the conflict between Arizona’s law and federal policy, reflected an institutional commitment to some state and local participation in immigration enforcement, another position likely reflective of compromise between law enforcement and political and civil rights interests. Further, certain actors within the federal system might prefer to work with local partners to advance their own particular preferences, which might differ from their superiors’. Devolution in immigration law enforcement highlights this dynamic. Federal field agents might share more in common with state and local officials interested in widespread enforcement, both for institutional and partisan reasons. This tension between center and periphery does not justify the latter in resisting the former, including through collaboration with more like-minded state and local officials, but it does underscore the difficulty of achieving a truly uniform enforcement policy.41

Finally, the very existence of state regulatory authority, as well as the concurrent nature of much of federal and state authority, means state and local lawmaking will challenge federal positions, requiring the federal government to react. The preemption remedy will only infrequently be available, not least because litigation is slow and costly. In fact, preemption may also not be desirable. The federal government might have an interest in disharmony, because state and local forays into fields that touch on federal interests can provide the federal government with opportunities. Consider again the marijuana debate. The federal government may well have an interest in de-escalating the war on drugs. The Democratic politicians currently in charge might favor decriminalization as a matter of policy, perhaps in order to reduce incarceration rates and diminish drug-related violence abroad. Law enforcement officials might share the interest in reducing violence and prefer a world in which resources can be devoted to more dangerous public health threats than those marijuana poses. But because of a variety of political and institutional pressures, it cannot be the prime mover in that process. Much as I have argued in other work that the federal government benefits from states and localities fighting out the gory details of how to conceptualize illegal immigration,42 the federal government may need a decentralized debate over legalization.

The debates concerning marriage equality and the Defense of Marriage Act also highlight the opportunities state-level divergence can create for the federal government. If we take DOMA’s rise and fall as the lens through which to view the federal government’s interest in the marriage equality debate, we can see that the federal position has been ideological and partisan, as well as reactive, and that the federal government has benefitted from state and local governments taking the lead in promoting marriage equality. Both Congress’s enactment of DOMA in 1996 in response to developments in the states, and the administration’s decision not to defend DOMA in court in 2011, were precipitated by decentralized developments throughout the country. In the case of the latter, the President began his administration publicly opposed to same-sex marriage. It therefore seems plausible that his and his administration’s evolution on the question was made possible by the growing momentum in state and local governments in the direction that otherwise aligned with the preferences of Democratic voters, or at least core principles of the Democratic Party.

More concretely, in response to these decentralized trends, the administration began a concerted effort in June 2010 to require the bureaucracy to do what it could with administrative tools, within the confines of DOMA, to extend federal benefits to same-sex partners.43 After the Supreme Court’s invalidation of DOMA in 2013—a decision grounded largely in equal protection principles but informed by developments at the state and local level44—the administration innovated still further. For example, DHS, backed up by DOJ, now permits U.S. persons to sponsor their same-sex spouses for admission to the country.45 As this essay went to press, the Attorney General took perhaps the boldest stance yet by the Department in relation to federal law. He declared that the federal government would recognize same-sex marriages performed in Utah pursuant to a district court order, extending benefits to same-sex spouses of federal employees married before the Supreme Court stayed the district court decision.46 The Attorney General aligned this move with the Department’s substantive equal protection agenda, despite ambiguous signals coming from the governor’s office about the legal status of those marriages under state law.47 He thus untethered, if only slightly, the federal position from the state one. In its domain, then, the federal government has been able to advance civil rights principles now held (and perhaps long held) by its leaders that redound to the benefit of its employees.48 These positions arguably have been made possible by developments in the states and have grown bolder as the debate has unfolded through dialogue between and among state and federal institutions.

As the Obama Administration’s actions have shown in the immigration and drug policy contexts, not to mention DOMA, full enforcement of the law and federal dominance are often not the salient federal interests within the federal system. Federalism can ultimately provide the federal government with opportunities to extend its influence and capacities. Decentralized conflict can work to its advantage.49 To be sure, as the case of S.B. 1070 suggests, federalism can also create obstacles to the realization of federal goals. But the framework for negotiation the system creates makes these obstacles surmountable and the need to assert primacy real but not totalizing.

II. state and local interests in opportunity and influence

Articulating the value of federalism from the standpoint of sub-federal governments might seem like a strange enterprise, given that those entities owe their very existence to the Framers’ choice to create a federal system in the first place. But much as the previous Part explored what use the federal government might have for state and local actors, and when it might seek to keep those governments at bay, here we can ask how states and localities approach their relationships not only with the center, but also with each other. Three basic features of the federalist system will be relevant to exploring this sub-federal perspective on federalism: the existence of a strong central government with the power to displace most state and local regulation; the decisional and jurisdictional independence of states and localities; and the horizontal relationships and effects that drive a great deal of politics and policy in a federal system.

Even still, the question of federalism’s value will be difficult to answer, because of the density of the system: the sub-federal encompasses multiple sovereigns with varying degrees of power. Though in theory the fifty states stand on something resembling equal footing,50 the vast disparities in their size affect the extent of the their influence in national debates and in relation to the federal government during tugs-of-war over enforcement. Think of the capacity of the state of Texas to determine the content of school textbooks for national audiences given its share of the market and of California’s ability to drive environmental regulation.51 As a result, the ideal vision of federalism, both as a system of dual sovereignty and as a system of inter-governmental relations, is likely to appear different across states. In addition, if we decentralize past the state level to the regional and the local,52 the possibility of conflicting interests and multiplicity of relationships among governments proliferates. New York City and Los Angeles may be able to attract national audiences and resist encroachments on their prerogatives by their creators and supervisors in the state legislature, but among the most interesting and important federalism dynamics scholars have begun to explore of late is the divergence between state and local interests and the use of state power to flatten the latter.53 Finally, as I have emphasized in the immigration context, public opinion at the sub-federal level is better characterized by diversity than homogeneity, our political shorthand of blue states and red states notwithstanding.54 The debates that rage across the country also occur within states themselves, no matter how ideologically coherent the state might seem on the electoral map. And so whereas it is possible to speak semi-coherently about the interests of the federal government, it will be far more difficult to speak in the abstract about the interests of states and localities within the federal system.

Despite this complexity, it should still be useful to evaluate federalism’s virtues from the sub-federal point of view, because states and localities must all negotiate their relationship with a higher level of government. I therefore assess the potential values of the system to sub-federal entities in the discretionary spaces where these negotiations occur: when the federal government has invited states or localities into a regulatory scheme; and when states or localities seek to take positions using their regulatory powers that might put them in the crosshairs of other governments, not just in a manner susceptible to preemption, but also in a way that might force the federal government’s enforcement hand. Without wholly discounting the ways Supreme Court doctrine has cabined federal power since the Rehnquist era,55 I also assume that our central government is remarkably powerful. It has a broad regulatory reach that states themselves might embrace, to the extent they want the federal government to solve collective action problems or take charge of policy problems on their plates. That said, states themselves retain considerable decisional independence in the exercise of their regulatory authority, because of the discretion the federal government exercises when wielding its own power, as well as the practical and political constraints on the federal government’s reach.

When it comes to federal invitations to cooperate, state and local interests align to a substantial degree with the federal. Rather than think of the federal government as a rival, states and localities will often (if not usually) think of it as the bearer of benefits. Joint federal-state operations and delegation schemes enable the lower levels of government to develop close ties to federal regulators, which in turn can help states expand their capacity to solve problems. The integrated scheme of disaster relief offers a good example of this benefit of federalism. Though recognizing the states’ frontline police powers, federal law enables governors to declare states of emergency and thus trigger federal assistance, leaving the power of initiation in the hands of states but providing invaluable opportunities to expand states’ capacities to respond to disasters.56 Through mechanisms such as joint law enforcement task forces and federal committees that include state and local officials, sub-federal actors gain access to useful information either generated by the federal government or pooled from disparate local sources by federal entities. Perhaps most valuably, as Abbe Gluck shows in her work on the Affordable Care Act, state actors can help shape federal regimes and priorities through their participation in joint regulatory exercises57—a phenomenon not lost on state and local bureaucrats who participate in immigration enforcement either.58

These possibilities generated by federalism will seem advantageous to state and local actors for high and low reasons. As a source of funding and technical support, the federal government ultimately enables states to expand their regulatory power and capacity to address the needs of their citizenry and amplify local policy preferences. Intertwinement with the federal government can also serve state politicians’ interest in re-election (federalism provides opportunities for advancing personal ambition, after all) and satisfy politicians’ and bureaucrats’ sense of professional obligation and reputational pride. The handful of instances in which states reject federal funds reflects the value of this relationship to state and local officials (though some might decry it as dependency).

To be sure, whether state and local actors will see the federal government as a source of opportunity will depend at times on partisan alignments. The states that have rejected the Medicaid expansion under the Affordable Care Act are all Republican-run. And yet, not all Republican governors have rejected the Medicaid funds, and we have reason to believe there will be less non-cooperation than feared after the Supreme Court’s unexpected decision enabling states to reject the Medicaid expansion without losing all program funds, though perhaps greater levels of non-cooperation than we have seen historically.59 But in many instances, the party in control of the federal administration will matter less to the choice to enter cooperative schemes or accept federal funds than the substantive domain in question and how regulation in that field generally aligns with ideological or partisan preferences. Think, for example, of the fact that states and localities that skew Democratic have resisted participation in immigration enforcement even within the Obama Administration, and that those that skew Republican have maintained their enthusiastic involvement in enforcement to the extent permitted by the current Administration.60

For state and local officials, there also will be value to a system of federalism that safeguards their decisional independence and capacities for regulation, and here the state interest in federalism has the potential to diverge from the federal, though not as significantly as might seem initially intuitive.61 On a simultaneously quotidian and high-minded level, independent lawmaking authority creates an institutional framework to address local problems that might not otherwise register with a centralized bureaucracy, particularly in a vast nation-state, in ways that more closely reflect the preferences of the local constituency being served.62 But the federal system also creates a valuable antagonist for state and local officials. Asserting independence from the center, even in ways that conflict with federal policy, can enhance the reputation and professional interests discussed above.63

The governing figureheads of Arizona’s anti-immigration movement, such as Governor Janice Brewer64 and Sheriff Joe Arpaio of Maricopa County, or California Governor Jerry Brown, who recently signed a slew of state laws designed to make life easier in California for unauthorized immigrants,65 or New York’s crusaders against Wall Street, such as former Attorney General and Governor Eliot Spitzer, traffic in the rhetoric of federal failure to frame their own regulatory projects.66 Their use of this trope highlights how the federal system creates political opportunities, even when accusations of federal failure or indifference might be unfounded. These moments can turn into wins for local constituents, to the extent the contrast helps propel policy innovation in line with their preferences, and they can advance the objectives of the larger social movements that are often behind state and local assertions of independence,67 which might in turn enhance the profiles and political horizons of state and local actors.68

While these uses of independent state power might be connected to advancing party interests on a larger scale, or to establishing a national profile for the local official, they also often will reflect genuine policy disagreements with the center. Having independent decision-making authority thus enables state and local officials to address local problems in the manner they see fit. Take for example immigration enforcement. For decades, local officials across the country have attempted to constrain federal enforcement of immigration law to protect local communities using forms of local power. New York City mayors such as Rudolph Giuliani and Michael Bloomberg adopted variations on policies that limited communication of immigration status information by local officials to the federal government.69 Today, state and local actors such as Governor Jerry Brown of California and Mayor Rahm Emmanuel of Chicago have signed state and local measures that decline to honor federal detainers, or requests that local police maintain custody of a person the federal government has identified for arrest under its immigration laws.70 Though most supporters of these sorts of measures have been Democrats (but see Giuliani and Bloomberg), their adoption appears indifferent to which party controls the federal bureaucracy and reflects the particular interests of local qua local officials.

Of course, the very fact of sub-federal multiplicity and complexity can also stymie the interests of local actors. Local government law scholars have contributed a vital perspective to our understanding of the value of federalism by highlighting the localization of preferences and the value of more localized decision-making,71 emphasizing that state bureaucracies and legislatures can be remote from electorates and operate in ways that flatten out points of popular disagreement.72 But as I have written in relation to immigration federalism, just as federal officials might seek either to tamp down or unleash sub-federal activism, depending on how it serves federal objectives, state officials will face similar incentives. The phenomenon of state-level preemption of local laws not only demonstrates how state-level political communities are themselves internally diverse, but also highlights how giving effect to decentralization “all the way down” will be challenging. Though this dilemma merits close consideration, the most important element of this dynamic for my purposes here is that it demonstrates the importance of institutional position to thinking through how decentralization should take institutional form.

Finally, the value of federalism to state and local actors does not stem exclusively from the vertical dynamic—the perpetuation of localized decision-making and the preservation of local preferences on the one hand and the forging of relationships with the federal government on the other. Among the most important elements of the federal system that legal scholars have only recently begun to explore is the horizontal one. The forms of influence, cooperation, and competition that flow across the local often do so without direct engagement of the federal government or federal interests.73 Indeed, the policy goals of state and local actors might often be better served through the horizontal forms of information-sharing and influence that a federal system makes possible and robust. Officials can be connected to like-minded politicians and bureaucrats in different states through the kinds of inter-governmental networks Judith Resnik and co-authors have written about.74 In addition, lobbying groups, public interest organizations, and policy entrepreneurs have been central in putting certain issues (marijuana legalization, immigration restriction, gun rights, and gay marriage, for example) on a national agenda by working through state and local governments with receptive electorates and institutional frameworks that make law reform possible,75 in turn enabling local officials to advance their policy preferences and political profiles.76

These mechanisms of policy diffusion have been controversial. They are often made up of highly ideologically motivated groups. They can seem opportunistic in that they grow up not from localized interests but from moneyed networks that transcend any given local setting. They seek to extend their influence by shopping around template laws to potentially receptive state legislatures.77 In other words, the influence of such groups challenges the idea that local actors have interests peculiar to their context, undermining democratic justifications for federalism based on the system’s capacity to enable local expressions of popular will. Local interests can be constructed and even co-opted by larger networks. But the fact of our increasing inter-connection, enabled by technology and the evolution of trans-local associations, also highlights a different function of federalism—that it creates an inter-governmental connective tissue of benefit not only to local actors who derive support and even status from these inter-jurisdictional dynamics, but also of benefit to people who share the views amplified through these networks—a possibility whose value I explore in more depth in Part III.

For state and local actors, then, federalism generates opportunities. But those chances to serve both constituents’ needs and their own ambitions depend on a high degree of integration, both vertically and horizontally. That is not to say that the desire to assert independence from the center, including when it does not serve federal interests, won’t be strong, particularly when it does serve partisan objectives. But the very tensions the system creates will themselves be productive, and ultimately the variable nature of the relationship will produce value.

III. the popular interest in multi-level politics

The final perspective on federalism I explore in this essay requires asking what the value of the federal system might be to the people whom it represents and regulates. For a variety of reasons, the official interests in federalism will not necessarily map onto the popular ones. To the extent that officials approach the negotiations required by a federal system with a view to preserving either institutional prerogatives or advancing reputational or partisan objectives, the particular arrangements they strike might not serve popular interests directly at all. But attempting to define popular interests could be futile, as they are enormously varied and emanate from the multiple and overlapping political communities that exist in any large and diverse republic.

It also might be hard to escape the banal observation that popular constituencies’ objectives are best served by national regulation some of the time, such as when those constituencies make up a national majority, or when their state and local governments lack capacity or will, and local regulation at other times. More cynically, it might be tempting to think of federalism as nothing more than a procedural framework for opportunistic ideological struggle.78 Perhaps federalism provides nothing more than a set of platforms for interest groups and the politically mobilized to shop policies and preferences around to receptive governments in hopes of advancing a policy agenda and perhaps eventually capturing a national majority, whether through statehouses or Congress.79 A turn to federal power may be on balance preferable if its institutions are receptive to one’s policy goals. But localism becomes appealing when federal action remains elusive, either for partisan political reasons or because of the relative difficulty of securing congressional action or even influencing executive decision-making.

Though these observations suggest that there may be no principled basis on which to defend a particular version of federalism, they do at least underscore that federalism can be useful from the popular point of view. To determine whether this utility generates consistent value, I would judge whether it serves the ends of government. At the risk of oversimplification, those would entail solving social problems and enabling the realization of popular values and preferences.80

The first question—does federalism maximize welfare and effectively enable solutions to those problems the electorate believes need to be addressed—is of a piece with one of the traditional justifications for federalism, namely that it creates laboratories of democracy. The value of federalism to this activist conception of government is not clear.81 The experimentation defense has not survived academic scrutiny in robust form,82 though examples of welfare-enhancing experimentation made possible by federalism certainly abound, as the drug policy narrative explored in the previous Parts underscores. Whether federalism serves problem-solving ends by generating new programs or ideas for reform that could be borrowed by other states or scaled up to the national level likely will depend in part on the sort of problem at issue.83 Federalism might have its greatest problem-solving value when the issue either doesn’t rate on the national agenda or the veto gates built into the national policymaking apparatus mean the only government action possible is at the state or local level, and popular pressure is strong enough to spur a response. An anti-preemption norm, at least within federal practice, if not doctrinally, thus may be appealing from the popular perspective as a means of ensuring that states and localities are left with the capacity to solve particularized problems. What is more, because the federal system already has entrenched certain mechanisms of governance that integrate the state and local with the federal, effective problem solving will necessarily entail negotiating the terrain I outlined in Parts I and II.

The clearer value of federalism from the popular point of view stems precisely from its creation of multiple electorates—a design feature that channels the complexity of public opinion by creating varied political communities with institutional features that can serve as vehicles for the realization of multiple and contradictory preferences.84 These communities may be overlapping and connected, but they do not blend into an undifferentiated mass.85 By expanding the capacity for politics, our federal system amplifies opportunities for the expression of popular preferences through law.86 This value of the system will exist even when states and localities fail to function as the proverbial laboratories of democracy serving broader systemic interests in creative problem solving.

The marijuana legalization referenda in Colorado and Washington nicely illustrate how federalism as a system of governance can serve the interests of discrete groups of voters in this way. State-level decision-making has generated concrete benefits for citizens of the two states, in an environment in which the same benefits are unachievable at the national level. Despite the continuing potential for federal enforcement, the referenda have eliminated state prohibitions that mirrored federal ones and thus substantially alleviated the risk of prosecution for marijuana possession, especially given the remote nature of the federal government and the comparative difficulty of federal prosecution. In addition to advancing certain individuals’ liberties, this development could have salutary public policy consequences by shifting law enforcement resources to more serious dangers and chipping away at high levels of incarceration for non-violent offenses. In other words, sub-federal decision-making is not merely expressive (in this case of the irrationality and perverse consequences of the drug war)—it enables political communities to translate expressive interests into policy benefits not achievable at the federal level. To be sure, it remains to be seen whether these experiments will give rise to some of the federal government’s fears, such as diversion to juveniles and neighboring states that would prefer not to have a marijuana market in their midst. One person’s popular benefit is another’s externality. But my strong intuition is that these experiments will promote long-term policy gains beyond the states where they began, even if in fits and starts.

The value of federalism to the realization of popular preferences stems not only from its creation of multiple sites for the people to exercise power, but also from the fact that the system enables polities to structure their governing processes in different ways that might change the scope and intensity of democratic decision-making. The referendum process that yielded the marijuana legalization laws in Colorado and Washington might not have been immune from the effects of money in politics, but it arguably expanded the people’s capacity for politics by providing them a way around divided, cautious, or slow-moving legislatures and then forcing those legislatures to act to implement their will.87 In the case of same-sex marriage, different forms of direct democracy also have enabled voters to respond to court decisions with which they disagree, either amending state constitutions to prohibit same-sex marriage or recalling judges who have struck down such prohibitions.

The horizontal dynamics created by federalism will re-enforce the expansion of the people’s capacity for politics through these alternative forms of decision-making by allowing voters to derive external support for their efforts through the networks and fundraising of like-minded people. These expanded possibilities for deliberation will also help local interests influence national debates by giving their preferences profile and thus the opportunity to influence others.88 As Heather Gerken puts it, the existence of “state and local platforms . . . connect[s] dissenters to the large and powerful networks that fuel policymaking in the United States.”89 In other words, lawmaking power at the state and local level can translate into influence at the national level, thus giving both minorities and dispersed majorities greater purchase on public debate and policy.

But the system also will create opportunities for national majorities by enabling national politics to take shape through sub-federal politics and decision-making—a lesson apparent in the discussion of horizontal federalism in Part II.90 Particularly when channeled through the mechanisms of direct democracy, local lawmaking can provide a platform for outside groups to influence policy developments in other states and localities through contributions to political campaigns and issue drives.91 To be sure, these horizontal forms of influence and information sharing can occur in unitary regimes as well. But the existence of sometimes rival but often peer governing institutions in other jurisdictions to learn from can make any policy gains made as a result of the sharing deeper and more lasting. As a result, contrary to assumptions made by some nationalists, the multiplicity fostered by federalism need not be balkanizing but can instead be productive and integrative, particularly if we assume the existence of a diverse or polarized polity.

Of course, in a world of expanded politics, voices that might otherwise be dominant may become muted or ineffective because of the multiplication of chances for contradiction—a particular dilemma for groups that might seek the entrenchment of a national governing norm to replace a multiplicity of regimes. For dispersed national majorities—for example, Democrats in Republican states when Democrats control Washington, or for the unorganized elements of the national electorate that nonetheless share common views—an active federal government might be preferable. Majorities that cut across jurisdictions might sometimes seek to foreclose diversity and impose commonality on popular constituencies in disagreement, sometimes simply to advance policy preferences but often to protect fundamental rights. The belief in the need for the federal government to impose a consensus can be fierce. The highly charged debate over abortion, for example, pits a strong commitment to the protection of a constitutionally guaranteed baseline right to terminate a pregnancy—the ultimate expression of a national norm—against deep ideological and moral disagreement that has resulted in pitched state-level politics and erosion of the right for the better part of thirty years.

The debate over abortion in fact highlights the most vexing puzzle, in my mind, that must be addressed in any effort to judge whether and how federalism serves the popular interest. Even if we can see the clear benefits of diversity in decision-making, when ought the central government, whether through the courts or the political branches, attempt to consolidate a principle or policy and thus shift the balance of the federal system from diversity to uniformity?92 The trajectory of the marriage equality debate also highlights the challenge of determining when consolidation serves the public good. In 2003, when the Supreme Judicial Court of Massachusetts issued its landmark decision declaring prohibitions on same-sex marriage to be unconstitutional under the state constitution,93 pursuit of a similar holding at the federal level seemed dangerous: to advocates of same-sex marriage because the Court could have entrenched the constitutionality of such prohibitions and generated political backlash, and to opponents who resisted national resolution of a moral issue over which broad national consensus had hardly materialized.94 Over ten years later, the swift progress through the states of a marriage equality norm and the dramatic shifts in public opinion seem to have vindicated the value of federalism to the debate, but they also have brought us closer to the brink of a national-level reckoning. It seems just a matter of time before the Supreme Court will have to face the constitutional question on the merits of whether same-sex marriage can be prohibited. If it finds ways to avoid the issue, as it did in Hollingsworth v. Perry,95 or finds prohibitions constitutionally permissible, then minorities in states that refuse to recognize marriage equality will be stuck in second-class status. But if it broaches the subject and declares prohibitions on same-sex marriage invalid, it will have foreclosed ongoing debate and opportunities for dissenters within the system to govern, in Heather Gerken’s formulation.

To decide which road is in the popular interest, we must be able to answer the question: at what point do ideological or other forms of diversity degenerate from being democracy-reinforcing and constructive into producing rights violations, or balkanizing the polity in a way that undermines the integrative project of the nation-state, or creating mundane policy chaos? No doctrinal test will be adequate to answer this question, but it is the key question for a theory of federalism that characterizes the system as one that advances the national interest. We can at least begin by trying to spell out the relative values of conflict and uniformity—an inquiry that does not receive enough attention in federalism debates. I grapple with this tension elsewhere, and so I only note here that “it can be difficult even after nationalizing moments to entrench a clear national norm, because such norms are elusive, or even ephemeral, and a commitment that at one point might have been a matter of consensus often gives way to disagreement about its meaning, as public opinion evolves and political fortunes change.”96 In other words, the project of bringing outliers within a federal system (whether at the national or the state level) into some sort of uniform line will be politically perilous and hardly straightforward,97 in large part because consensus often exists only when principles are stated at a high level of generality. It is in the implementation of consensus that things begin to fracture into competing visions—precisely when having institutional means available to channel disagreement will be vital to the popular interest.98 I thus return to an observation with which I began—that the popular interest depends on having a system of government that makes ongoing negotiation possible, which requires robust local institutions and fairly generous tolerance of disharmony.


The preceding discussion accepts federalism as a hardwired feature of our Constitution and our political culture but emphasizes that, within its structures, a great deal of space remains for arguing about how inter-governmental relationships ought to be constructed and competing interests reconciled when they arise. I began with an internal point of view and considered how the actors within the system understand their relationships to one another, and the corresponding advantages that might be harnessed from decentralized decision-making as opposed to consolidation. This sort of inquiry, which defines the work of federalism as a system, highlights an important feature of the regime—that its contours are always under construction, determined in large part by the advantages the different actors might accrue through their interactions with one another. Though this point of view may make it difficult to articulate totalizing theories about the system’s value and even its purposes, implicit throughout my discussion is an appreciation of the value of decentralization to all of the actors in the system, including the federal government. The discussion also highlights the importance of facilitating interactions and trade-offs among governments, not only by resisting overly rigid and hierarchical rules to govern relations, but also by identifying opportunities for institutional integration that enable either joint or concurrent decision-making.

The fact that federalism need not take on a fixed form does not mean that the processes of negotiation should not be informed by certain principles that transcend institutional interests. Ideally, the actors that shape its parameters would think of their roles in broader systemic perspective, or through the lens of how best to advance the popular interest and achieve the purposes of government.99 I have only begun this sort of external evaluation here. But if we assume that the purposes of government are to advance the popular will and help solve social problems that affect the people’s health and wellbeing, I would defend the proposition that the federal system works well, at least to promote the former, by expanding the capacity for politics and therefore for popular expression. The primary systemic benefit of this expansion may ultimately be that it enables national conversations that embrace contradictions and that ultimately lead to better national integration over time. A decentralized system makes it possible for contradictory policies to coexist and keeps open the capacity for change—an especially important feature in a system of government in light of the elusive quality of consensus and the diverse nature of our polity. As I have attempted to show, these features of federalism are ones also appreciated by its institutional players, and they therefore inform federalism in practice as well as in theory.

I take one of the central contributions of this whole Feature to be its challenge to the nationalist’s suspicion of the sub-federal through its demonstration of how decentralization can serve national and integrative ends by leaving open opportunities for negotiation. My own view is that federalism has been and will likely continue to be crucial to maintaining a functioning polity amidst a deeply diverse electorate. Throughout my work I have expressed strong intuitions that decentralization is well suited to the project of achieving equilibrium in a diverse setting, not only because it permits a sharp focus on the institutional locations of integration, but also because of its expansion of our capacity for politics and therefore for ongoing negotiation about our differences.