Le conflit normatif est une composante essentielle du fédéralisme, lequel façonne les affiliations par le droit et est constitutionnellement lié au pluralisme juridique. En dépit d’engagements préalables sur la réparation structurelle des compétences, le pouvoir est régulièrement renégocié dans les fédérations. Pour cela, des mécanismes de médiations sont requis. Dans cet essai, j’analyse deux méthodes – l’essentialisation de la répartition des compétences d’une part et, d’autre part, le fait de réduire les droits individuels ou bien de l’identité propre des entités fédérées – qu’ont développées les cours des États-Unis et d’Europe en rendant leur décision au nom du fédéralisme. La position essentialiste relie certains domaines d’activité particuliers à des niveaux de gouvernance, et justifie ses décisions à partir d’une conception naturaliste de l’autorité et qui en pose la singularité. Les réductions auxquelles donnent lieu le fédéralisme sont ici illustrés par les analyses que tient la Cour suprême des États-Unis sur la question de savoir si les juges des États fédérés ont ou non « excessivement tort » quand ils statuent sur les droits des prévenus dans le cadre d’affaires pénales, et par l’examen des décisions de la Cour européenne des droits de l’homme qui invoquent la marge d’appréciation. Ces réductions ou bien rendent moins efficient un droit spécifique par égards pour les revendications d’autorité ou d’identité d’une entité fédérée, ou soulignent au contraire le fait que la décision de l’entité fédérée outrepasse ses prérogatives pour favoriser l’efficience de ce droit.

Je soutiens que les positions essentialistes sur le pouvoir des entités fédérées ou de l’entité fédérale dissimulent l’activité et la pratique des juges par laquelle ils choisissent en réalité la manière dont sont caractérisées les activités déterminantes pour le degré de pouvoir. Au contraire, lorsqu’ils procèdent à des réductions, les juges doivent rendre compte de la manière et des raisons pour lesquelles le centre, l’unité inférieure ou les individus supportent les coûts ou récoltent les bénéfices des différences dans la protection des droits. En outre, si de telles réductions sont provisoires, elles servent potentiellement les buts du ou des fédéralisme(s), qui sont de concilier les différentes sources juridiques en reconnaissant des désaccords en cours et la nécessité d’échanges renouvelés au sujet de l’étendue des droits et des aspects qui lient l’identité des centres et des unités inférieures.

Normative conflict is a fixture of federalism, shaping affiliations by law and constitutionally committed to legal pluralism. And, despite the frequently found pre-commitments to structural allocations of authority, power is regularly renegotiated in federations. To do so, mediating mechanisms are required. In this essay, I analyze two methods – essentializing allocations of authority and discounting either individual rights or identities of subunits – that courts in the United States and Europe have developed as they render decisions in the name of federalism. Essentialist claims link particular domains of activity to levels of governance, and justify decisions by assuming the naturalism and singularity of authority. Federalism discounts are exemplified here through analyses by the United States Supreme Court about whether state judges are « unreasonably wrong » when adjudicating rights of criminal defendants, and through examining decisions of the European Court of Human Rights when invoking the margin of appreciation. Discounts either under-enforce a particular right in deference to a subunit’s claim of authority and identity or insist on overriding a subunit’s decision in favor of enforcement.

I argue that essential claims about the power of either the subunits or the larger entity obscure the agency of judges when making choices about how to characterize activities that determine the flow of power. In contrast, when providing federalism discounts, judges have to explain how and why the center, the subunit, or individuals bear the costs or reap the benefits of variations in rights protection. And, if such discounts are provisional, they have the potential to serve federalism(s)’ goals of accommodating plural legal sources by recognizing ongoing disagreements and the need for iterative exchanges about the scope of rights and the facets that bind the identity of both subunits and centers.

I. Federalism(s)’ Conflicts and Challenges: Prisoners’ Voting, Violence as Discrimination, Families as Sites of Inequality, and Incorporating Migrants*

 

In 2014, 1,015 prisoners complained to the European Court of Human Rights (ECtHR) that the United Kingdom had violated their rights through automatic disenfranchisement, precluding them from voting in elections[1]. In the winter of 2015, as it had in 2005 and several times thereafter, the ECtHR held that the U.K. statutory ban was, « by reason of its blanket character, incompatible with Article 3 of Protocol No. 1 to the Convention[2] ». As that judgment reflected, during the last decade, the U.K. has repeatedly declined to alter its disenfranchisement statute despite several ECtHR decisions that its rule deviated from its Convention obligations[3]. The U.K. rested its noncompliance on the grounds that it retained political and legal authority to decide who could participate in its elections.

The legal authority of subunits has also been central to debates about gender equality. Illustrative are conflicts over federal decisions that national equality norms govern violence against women and same-sex marriage. In the U.S., one question contested for a decade was the power of the U.S. Congress to enact the 1994 Violence Against Women Act (VAWA), announcing that gender-based violence undermined women’s equality and seeking to change practices of police, prosecutors, and judges when dealing with issues often characterized as « domestic[4] ». Critics attacked one facet, the statute’s « civil rights remedy », providing victims of violence with a cause of action to seek damages in federal courts in addition to remedies available in state courts[5]. In the 2000 decision of United States v. Morrison, the U.S. Supreme Court held that Congress lacked the power to create that remedy. The Court, by a five-to-four vote, rested its holding on an essentialist federalism account that the « Constitution requires a distinction between what is truly national and what is truly local[6] ». Stipulating that violence was not « economic in nature[7] », the Court ruled out federal power under the Commerce Clause and situated the matter as one of tort, crime, and the family, all of which it deemed « local » issues, governed by state law.

The question of relationships among states and between state and national power has also been in focus in conflicts about the regulation of marriage in the United States. During the first decades of the twenty-first century, several state courts, including Massachusetts and California, concluded that bans on gay marriage violated their constitutions. In contrast, Michigan, Kentucky, Ohio, and Tennessee amended their laws to limit marriage – and the issuance of state marriage licenses – only to « a union between one man and one woman[8] ». Although those states insisted that marriage was a matter for local government, the U.S. Supreme Court ruled that the U.S. Constitution protects the right to marry and individuals’ « personal choice regarding marriage[9] ». In its 2015 Obergefell v. Hodges decision, a five-person majority insisted that variation was not permissible; federal due process and equal protection guarantees invalidated prohibitions on same-sex marriage because such rules undermined the « equal dignity in the eyes of the law » of same-sex couples[10].

Migration is another subject that has produced decades of disputes about regulatory authority in the United States. Efforts to assert federal power can be seen in two 2015 bills, « Enforce the Law for Sanctuary Cities Act » put forth by members of the U.S. House of Representatives, and « Stop Sanctuary Policies and Protect Americans Act » proffered in the Senate[11]. Proponents want to prevent states and cities from providing employment, in-state tuition for higher education, and identity documents for non-citizens in their communities[12]. To do so, the bills propose to withhold federal law enforcement funds from states or localities that have « any law, policy, or procedure » limiting their officials from « gathering information regarding the citizenship or immigration status, lawful or unlawful, of any individual[13] ».

In contrast to « sanctuary » localities, other states want to exclude migrants through civil and criminal laws policing movement, authorizing detention, and banning migrants without the requisite documents from renting housing or doing business. In 2012, the U.S. Supreme Court asserted the exclusivity of national authority over the criminalization of migration when holding (in another five-person majority decision) that Arizona could not add crimes onto the federal immigration structure[14].

The topic of migration illustrates not only disagreements among states but also that the same states do not have static policies over time. In the mid-1990s, California was a leader in formulating restrictive policies (« Prop. 187[15] »), limiting access for both authorized and unauthorized migrants to social benefits and employment. About twenty years later, in 2014, California became the first state to permit migrants without documentation to become members of its bar[16]. Other recent initiatives in California relate to providing health care, driver’s licenses, and education; through such measures, the state is providing a kind of state-citizenship for migrants who lack national permission to enter or to remain in the United States[17]. In practical terms, California was predicted to issue 1.5 million licenses to such individuals[18].

I provide these snapshots of conflicts over prisoner voting, the relationship of equality law to violence and to marriage, and migration to make plain that no one level in a federation can be equated with a particular posture towards individual rights. Furthermore, proponents on all sides claim that their rules – whether insistent on national norms or recognizing variation – respect the ideals of federalism, which are often linked to tenets of democracy and of community. The examples thus put into sharp relief the benefits provided or the price paid by the center, its subunits, or individuals when rights are varied or when uniformity is required. Further, as the brief descriptions also make plain, judges are but one set of actors in complex, multi-iterative interactions, shaping and responding to conflicts over rights in federations.

Sketching these disagreements introduces this essay’s claims. First, the consequences of federalism’s toleration and celebration of differences through the endowment of authority to various political sectors (be they states, provinces, Länder, cities, indigenous nations, or linguistic or other minorities) is liberal in the sense that, through building into its structure multiple internal legal regimes, federalism enables participatory opportunities that create plural sources of law, enabling layered political identities of its citizens, comparative inquiries into legal rights, and differing responses[19]. But neither the structure of federalism nor any particular level of a federation intrinsically produces liberal commitments or precludes illiberal practices[20]. National and subunits generate distinctive policies that cross the political gamut and which can endure – absent social and political efforts garnering sufficient power to undo them.

Second, federations are bound together through legal affiliations that reflect multi-level governance. The plural legal sources within a federation generate conflict. Power is regularly contested in federations. The authorization of many points of law production through layered and redundant legislative, executive, and judicial systems enable norm entrepreneurs to shop systems in efforts to persuade similarly-situated actors (such as executive officials, judges, and legislators from different levels) about the wisdom or the legality of particular points of view – for or against, for example, universal voting rights, violence-against-women as discrimination, same-sex marriage, and openness towards migrants. Because concurrency is permissible, disagreements can readily emerge about particular rights. Instead of aspiring to « the tranquility of the state » (to borrow from Thomas Aquinas[21]), federalism entails discord, as an ongoing feature and as a positive attribute of governance.

The resulting « complex concurrency » can, as Robert Cover explained long ago, provide confirmation of the strength of legal commitments, if distinct political units produce the same outcomes[22]. The persistence of disuniformity, in turn, marks the depth of disagreement about what the norms should be. Such conflicts have an expressive function, through which subunits and the larger entity mark their own identities. Illustrative is conflict over same-sex marriage in the United States. Writing for the Court, Justice Kennedy stated, the « nature of injustice is that we may not always see it in our own times[23] ». But the case arose because of conflicts among the « we » about what constitutes injustice. And the resolutions – whether longstanding or temporary – come from being part of a legal federalist « we » that continually provides a platform on which to contest what is constitutive of that « we ».

Which issues become the vortex for identity are not preordained in founding documents allocating competencies. Thirty years ago, one would not have assumed that the U.K. would insist that a blanket ban on prisoners’ voting was central to its identity, that Michigan would be at the forefront of anti-gay (as well as anti-affirmative action[24]) efforts, or that California would become a leader in the incorporation of migrants who lack status documentation.

Third, federalism analyses are often framed as if the various parts were fixed. Commentators regularly assume that the relevant participants are solo-acting subunits and that constitutional texts determine competences. Those presumptions stem from the history of federations (either « coming together » or « holding together » the subunits), as well as from constitutions, conventions, treaties, statutes, judicially-crafted doctrines, and commentary, which seem to offer fixed answers by enumerating powers or providing residual clauses specifying particular rights and assigning competencies over certain subject matters. Doing so entails reliance on essentializing rights, roles, and jurisdictional allocations and then positing that power over a given domain (such as « the family », « immigration », « equality ») or a kind of right (such as voting or marriage) belongs either to subunits or to the center, that a few arenas are concurrent, and that the relevant jurisdictional entities act unilaterally[25].

But, at both a descriptive and a normative level, this structure is askew. The import of the rights set forth in constitutions and conventions change, and the identities of both the subunits and the federated government are likewise dynamic. Conflicts about rules on voting, equality’s relevance to violence and to marriage, and about migration in Europe and the United States reflect the interaction between rights and political identity. Each issue was once – and remains – a matter committed to local governance, and each becomes freighted as new rights-holders appeal to the larger unit by reframing a behavior as one for which central protection is required by foundational commitments.

Those exchanges are not only dynamic within federations but also shaped by a host of world events, including shifting interactions with other governments, federated or not. Thus, to conceptualize states, cities, and other subunits as « lone rangers », in indigenous acts of self-rule, is to miss that their lawmaking, even when asserted to be constitutive of identity, is embedded in, as well as fueled and often funded by, translocal and transnational social movements. Authority does not move in only one direction but back and forth as well as diagonally. Law and political rights migrate, and federations in particular permit multiple sources of entry.

Thus, unit-centered federalism needs to be informed by focusing on the porousness of geographical boundaries and on legal-political configurations that do not map onto territorial borders[26]. Dozens of translocal or transnational organizations of government actors are « private » organizations of « public » actors – serving as norm entrepreneurs that alter the inputs of « local » and « national » policies. I have offered the acronym TOGAs[27] – transnational organizations of government actors – to denote diverse groups of trans-municipal, trans-regional, trans-county, trans-state organizations formed by officials or entities at a particular level of government (city mayors, for example) or organized by responsibilities (for example, sanitation, policing, courts, government buildings, prisons, utilities, the environment).

The abbreviation NGO for nongovernmental organizations reflects that NGOs are comprised of individuals or entities working in concert to influence governments from the outside. In contrast, TOGAs are sometimes organized by governments and constituted by law, as exemplified by the Committee of the Regions in Europe[28]. More often, they are private organizations drawing their authority and democratic legitimacy from the fact that their members are themselves employed by governments. TOGAs lay claim to technical expertise through the positions that their members hold. TOGAs aggregate their members’ political capital for diverse reasons, such as seeking to expand markets, to teach skills and exchange information, to buffer against intervention from other levels of government, or to get funds from other sectors.

Agreements across subunits within a given federation (horizontal federalism) have become increasingly commonplace, along with translocal-transnational accords in which subunits of more than one federation co-venture with each other across national boundaries. Instead of lawmaking authority that is « truly local » or « truly national » (to borrow again terms from U.S. case law[29]), norms travel horizontally, vertically, diagonally, and diffuse irregularly, with subunits and their officials often functioning as co-venturers rather than as solo actors.

Stereotyping groups is a mistake, for subunits and the national unit do not always line up neatly on one or another side of particular federalism fights. For example, in the United States, amici briefs may be filed by groups of states, on opposite sides of positions before the U.S. Supreme Court and disagreeing about what rule of law would be respectful of their interests. Likewise, NGOs and Contracting Parties put forth views to the ECtHR and may also diverge. And, just as subunits of national governments’ « interests » are not fixed, neither do the agendas of TOGAs consistently reflect a particular political valence. For example, the U.S.-based National League of Cities gave its name to a case decided by the U.S. Supreme Court that imposed a limit on federal regulation of state employees’ wages and hours[30], which is a position typically read as « conservative ». More recently, the National League of Cities has supported a host of resolutions read as « liberal », including calling for progressive efforts on climate change and support for immigration reform that enabled undocumented immigrants to obtain legal status and that sought to buffer local personnel from being conscripted to participate in surveillance of migrants[31].

Rather than conceptualizing the various shifts in power as « competence creep[32] », over-endowing one set of actors, competencies should always be understood as in motion – and in more than one direction, as the import of rights and the functions of government change. Once the multiplicity of intersecting and layered units is included in federalism(s)’s narratives, reappraisal is required of standard virtues claimed for federalism – autonomy, diversity, voice, and exit. Participation is made complex when policies are made trans-locally. If replicated from subunit to subunit, exit options are reduced[33]. Shopping for law is also constrained, as subunits mimic or adopt each others’ practices. Indeed, aspirations for diverse approaches may themselves be reduced. For example, the current « tagline » of the U.S. Uniform Law Commission, a body composed of expert representatives from every state, explains that the group’s goal is « diversity of thought, uniformity of law[34] ». This slogan is in tension with accounts of federalism presuming that states are focused on competition and diversity.

I have outlined three facets of federalism(s): liberalism of structures generating plural legal regimes; the regular generation of conflict over laws; and a diversity of entities defying a neat mapping of the ebbs and flows of power on vertical and horizontal grids. A fourth facet is the need for mechanisms to resolve conflicts. When doing so, the plural legal regimes such as the United States and Europe (whether termed a « federation » or not) have to respond to central and ongoing challenges: when to press for uniformity and to insist on particular obligations unmodified, how to generate shared commitments, and when to respect and to support differences. Deciding when and how to do so is made complex because subunits and TOGAs themselves regularly disagree about what serves their, and the larger, interests.

A fifth facet of federalism(s) thus comes into view: the need for institutions and practices to mediate such conflicts in a manner perceived to be legitimate by the various stakeholders. Given commitments to subunit participation in the creation of the larger norms, to norm autonomy, and yet to some norms so foundational that deviation is not permitted, resolutions are needed either because individuals and entities cannot simultaneously comply with differing regimes or because the variation is seen as too far afield from the identity entailed in membership in the larger or smaller political unit.

To be successful, mediating doctrines and institutions need methods to support compromise or to facilitate structured disagreement. The U.S. Uniform Law Commission, mentioned above, is one example, aiming to reduce differences through each state’s enactment of identical laws, built from a model law on a particular topic. The Commission is what I term state-regarding or state-protective because it puts each state to the task of individually affiliating with each of the Commission’s model laws. Doing so aims – as the name suggests – at developing state-based uniformity, explained as desirable in a federation because it simplifies law and thus facilitates both economic and interpersonal activities across state lines. Parallels, albeit with different methods of bridging subunit interactions, can be found in the European Commission for Democracy through Law (known as the Venice Commission) and the 2015 creation of the Superior Courts Network, seeking to expand exchanges between the highest national courts and the ECtHR.

Dozens of such institutions exist; the focus in the remaining parts of this essay is on how courts have developed mediating methods. The two techniques to respond to federalism(s)’ conflicts in focus here are efforts to essentialize competences over specific subject matters and federalism discounts licensing deviations from legal obligations. A word of introduction about each is in order. The essentialist approach assumes the stability of the political units that comprise a federation and links subject matters to the authority of either the subunits or the larger entity. Discounts permit under-enforcement of central norms in deference to a subunit’s authority. To provide examples of essentialist premises, I analyze changing definitions of and views on authority over foreign affairs, domestic affairs, migration, and the relationship of equality to violence targeted at women and to marriage rules. To illustrate federalism discounts, I draw on U.S. rules that license systematic under-enforcement of criminal defendants’ federal constitutional rights and the European doctrine of the « margin of appreciation »; under both approaches, the subunit community identity or the right asserted is given less than the full value. My elaboration of these two practices across just two federalist entities aims to invite reflection on whether other federations have parallel doctrines aiming to accommodate differences.

Domains of authority are not natural artifacts but political decisions, renegotiated as conflicts emerge about the import of rights in relationships to the political identities of the governments involved. Therefore, I am a skeptic of the idea of an enduring principled nationalism or localism that provides guidance without a focus on specific discussions of particular rights and on the conflicts about rights generating disagreement. I argue that successful mediating doctrines require acknowledgment of the choices and the trade-offs made through self-consciously specifying the status of rulings as temporizing measures. And thus, I am a critic of jurisdictional essentialism because it cloaks the choices that judges make about federalism by asserting a naturalized complacency about jurisdictional competencies. What is needed instead is a disciplined examination of how definitions of subject-matter competencies change and why authority ought to reside over a specific issue at a particular time at one level of government or another, or ought to be shared. The resulting state of flux is a virtue (rather than a pathology) of federalism.

In contrast, federalism discounts have the potential to be generative when mediating conflicts and in making rights material for individuals. Deciding when to deploy discounts puts judges to the work of explaining the import, scope, and power of rights, so as to decide the degree to which deviations are permissible and why[35]. But not all federalism discounts have such tailoring, as is illustrated below by current U.S. law on post-conviction remedies, which gives a blanket discount to state courts’ arguably wrong decisions, if those rulings are not « unreasonable » misapplications of federal constitutional law.

Successful federalism discounts imagine – and count on – a future in which either the decision to discount norm implementation produces gradual realizations of shared commitments to particular forms of rights or the content of those rights is reduced[36]. Discounts are guesses, seeking both to protect subunits and the central authority. The burdens imposed by permitting variations in rights cannot be fully assessed at the time that decisions are made to accord them. In retrospect, a particular discount could be seen as appreciating subunits’ roles in community formation and law implementation, as an innovative moment of fuzzing rights, or as the beginning of a redefinition or deterioration of certain norms, formally posited to be central[37]. Assessing the impact of federalism discounts is itself complex, for it requires not only the short-term conflict between a single subunit and the center but also a wider lens, scanning an array of actors. For example, to look at the conflict between the U.K. and the ECtHR on prisoner voting rights as a power struggle between the two entities is to miss the contributions made by the judicial insistence of prisoners as citizen-voters to the transnational reevaluation of the treatment of prisoners now underway.

To argue the utility of federalism discounts is neither to underestimate the difficulty in deciding when to use them nor to endorse all the forms that they take. Even when courts seek to render decisions that are state-regarding or state-protective, the many voices within and across subunits, advocating opposing points of view, can complicate the task of deciding what responses are enabling of subunits. Moreover, because federalism discounts are often proffered in the name of recognizing the autonomy of a subunit and its commitments to particular views, the cacophony of positions internal to even one subunit raises questions about whether such deference is due.

Federalism discounts should therefore be understood as way-stations. When discounts are given – in that either the subunit has to conform to its opponents’ understanding of a right, or the rights-seeker gets protection of that right – costs are imposed. Courts ought to acknowledge these costs by stating the contingent status of such discounts, fashioned as temporizing accommodations that leave open the possibility of revisiting assessments of the rights in contest and the subunit’s commitments to them. If sanctioning gaps between norms and subunit activities, federalism discounts should both specify that the lines drawn in the moment are subject to reconsideration and explore how (if permitting less protection rather than more) to cushion the impact on individuals losing their rights claims. As the back-and-forths on prisoner voting make plain, doing so is also self-protective, in that what seems to have been settled – even through constitutional adjudication – can be reconfigured[38].

 

II. De-naturalizing Authority: The Ephemera of Competencies

 

It is commonplace in layered legal regimes to assign tasks by the nature of the government actors (courts, legislatures, the executive, administrative agencies of the larger and smaller units) and by certain subject matters, sometimes permitting concurrency and other times seeming to wall off joint governance. Yet, which arenas reside at what level, what issues fall within the designations, and whether all subunits have the same authority vary across the set of federations and over time. Explaining more about the use of essentialism is necessary because categories are of course regularly deployed in law. But the act of categorizing needs to be brought into focus, for the import of categories change as political and social movements reframe rights and the activities of governments. The details below are in service of demonstrating that judges, rather than federalism (as if it had fixed authority), regularly resolve conflicts by deciding what activities fall within the control of which level of government and why.

The examples discussed here come from the categories of « foreign affairs », « migration », « domestic affairs », and « family law » – all placed in quotes to remind readers that what these terms denote is neither a fixed subject matter nor an uninterrupted relationship with a particular level of governance in a federation. Moreover, which rules are state-regarding or state-protective may vary not only from the perspective of the larger entity and its subunits, but also among subunits and over time. At least four kinds of essentialisms, each making claims about the necessity and the dominion of law sourced at a particular level, are at work. Below, I show their pitfalls.

A first essentialist claim is about subject-matter exclusivity. For example, in the United States, foreign affairs and migration powers are posited as paradigmatic examples of functions of the national government[39], just as family life is assumed to be the domain of states. Yet, as I sketched in the opening examples, subject matter authority overlaps and flows back and forth. States actively make migration policies, albeit in opposite directions. During the last decade, Arizona added criminal penalties for migrants atop those of the national government, while California identified ways to regularize the status of migrants who go unrecognized by federal law. In a parallel fashion, the states of Michigan, Kentucky, Ohio, and Tennessee asserted their claim to exclude same-sex couples as part of their inherent authority over family life, and it was the national government that located rights to marriage in federal constitutional guarantees. But to describe the evolution of rules at the federal level – be it the five decades of criminalization of migration, the liberalization through legalization and deferred deportation in the 1980s and in 2013-2014, or the recent recognition of same-sex marriage rights – one has to look at state law, as so much of federal law is built from and incorporates aspects of various state rules[40].

A second kind of essentialism stipulates « state interests », as if the positions serving those interests are singular and exist outside of the conflicts about them. The records in federalism cases in the United States are replete with examples of states taking opposite sides. Illustrative of conflicting arguments about what advances state interests are filings in the 2012 Supreme Court immigration decision, Arizona v. United States, about the lawfulness of Arizona’s imposition of criminal sanctions on unauthorized migrants[41]. Michigan, joined by fifteen other states, argued that states had an interest in « ensuring that their sovereignty [was] accorded proper respect[42] ». Categorizing the question of migration control as within the « police powers of a state », these sixteen states insisted that a « state’s own sovereign authority » entailed the power to « arrest people for committing federal crimes, including violations of federal immigration laws[43] ».

New York, leading ten other states, took the opposite view, filing in favor of federal preemption (a stance that the states did not take « lightly ») and likewise rested their arguments on sovereignty grounds. These eleven states reminded the Court that they too had « undocumented immigrant populations within their borders, and they have adopted diverse measures to address the impacts of that population, in a manner reflecting individual state priorities and resources[44] ». While wanting « to preserve their authority to enact and enforce » their laws, « even as applied to immigrants », federal preemption was appropriate because if Arizona’s « distinct state policy on removal » were permitted to stand, it would impose negative externalities (in the language of the brief, « would place disproportionate demands on federal resources and have other interstate effects[45] »).

In short, both sets of states argued that state interests were served, albeit by the opposite outcomes – one permitting states to layer state crimes on federal prosecutorial authority and the other opposing Arizona’s policy because it conflicted with the « unified, national removal policy[46] ». And, of course, the disagreements about what rules served « state interests » were part of political conflicts about the treatment of migrants. As a consequence, when judges want to resolve this kind of dispute based on state-protective rationales, judges have to choose among competing versions of what federalism requires.

A third aspect of essentialism relies on a parallel naturalism about « national interests », again as if they were fixed. But « the national » is likewise forged through interactions over time and in exchanges with a multitude of actors. When the Supreme Court in 2012 in Arizona v. United States held that national rules about the criminalization of migration precluded state deviations, it chose to foreground the supremacy of federal law as it stressed the exclusive authority of the federal Executive to enforce or decline to prosecute particular immigration violations.

Yet an alternative account of what « the national » required was available – that Arizona’s aggressive treatment of migrants violated American commitments to liberty and equality. The basis for such arguments comes from the 1941 decision in Hines v. Davidovitz, rejecting a Pennsylvania statute requiring persons designated « aliens » to register and display their papers on request[47]. As in Arizona v. United States, the Court concluded that federal law preempted the state initiative. But as World War II was raging, the Court foregrounded American commitment to liberty.

Justice Black wrote for the Court that « [o]pposition to laws permitting invasion of the personal liberties of law-abiding individuals, or singling out aliens as particularly dangerous and undesirable groups, is deep-seated in this country[48] ». Requiring aliens to register and to display identity cards was « at war with the fundamental principles of our free government, in that they would bring about unnecessary and irritating restrictions upon personal liberties of the individual, and would subject aliens to a system of indiscriminate questioning similar to the espionage systems existing in other lands[49] ». In Hines, the Court also located its anti-harassment rule in American hospitality towards foreigners which, through « reciprocal promises and guarantees for our citizens while in other lands » also provided protection[50]. American identity was embedded in affiliation with « treaties and international practices … aimed at preventing injurious discriminations against aliens », especially those who were « perfectly law-abiding » and who had met « rigid requirements as to health, education, integrity, character, and adaptability to our institutions[51] ».

Those aspirations were heartening, but no essentialist meaning exists of « the national », any more than it does of « the local ». The equation of U.S. identity with liberty and respect for outsiders soon gave way. As is likely familiar, in 1945, Justice Black wrote the opinion in Korematsu v. United States[52], in which the Court infamously upheld a military order, addressing « all persons of Japanese ancestry, both alien and non-alien, from the Pacific Coast[53] » and making criminal the refusal to report to an internment camp[54]. What the 1941 Hines decision and the 1945 Korematsu case illustrate are competing constructions of « the national ». The 2012 Arizona decision thus reflects a choice made by the Supreme Court when ruling the state’s additional criminal penalties impermissible. Rather than focus on their breach of American commitments to liberty and equality or international agreements about how to treat foreigners, the Court identified the violation as an impermissible intrusion on federal authority. Because Congress had provided « a full set of standards[55] », it had left « no room for States to regulate[56] ». In 2012, « the national » meant unified power, not an insistence on non-harassment and personal liberty.

In addition to subject matter, local, and national essentialism, a fourth kind of essentialist claim is the presumption of the naturalness of federal or of state exclusivity, as if certain kinds of activities were intrinsically only to be left to a particular level and as if such an either/or allocation was a logical entailment of federated forms and sustainable in fact. A brief comparative foray undermines that naturalism, as rules vary on which level of government has what authority and in practice, overlap is commonplace.

One example is migration, deemed as the discussion of the Arizona litigation illustrated, to be an exclusive national power in the United States. In contrast, Canadian provinces have some powers over immigration[57], and the constitutions of many Swiss cantons have provisions « aimed at cooperation … with foreign regions, with foreign states, and even with international organisations[58] ». Swiss municipalities control the naturalization of citizens, and the function includes not only the implementation of federal norms but also the capacity to make legal judgments about the propriety of individual applicants for citizenship[59]. German Länder have some direct relationships with the European Union and, more generally, regions are coming into more focus, as many sub-state units are participants in a wide array of « diplomatic activity[60] ».

Moreover, even as the Court in Arizona v. United States spoke about the exclusivity of the federal « voice » in « foreign affairs », states and localities regularly function as transnational actors seeking to affect international policies, importing law from abroad, and priming their own economies through cross-border exchanges. The range of activities run from states and localities aiming to change U.S. policies relating to the Vietnam War, the Gulf War, and the conflicts in Northern Ireland and the Middle East to the promotion of nuclear disarmament, the end of apartheid in South Africa, consumer product safety, environmental protection, and human rights.

Another example of trans-border activity predicated on the subunit scale comes from concerns about climate change. Domestic policies on global warming have been shaped through iterative interactions among transnational lawmakers, the federal government, and hundreds of subnational entities. After President George W. Bush withdrew American support for the Kyoto Protocol, several U.S. cities, led by Seattle and Salt Lake City, enacted ordinances to conform to the Protocol’s targets for local utility emissions. In March 2005, a group of nine mayors agreed to their own climate protection program that was approved by the United States Conference of Mayors in June 2005[61]. As of 2015, 1,060 mayors, representing towns and cities with combined populations numbering almost 89 million people, had endorsed that program, which was affiliated with the transnational principles (« ratifying Kyoto at the local level », so to speak)[62]. Canadian municipalities likewise have undertaken environmental joint venturing[63]. A parallel in Europe began in 2008 when the European Commission established a « Covenant of Mayors », inviting mayors willing to create baseline emission inventories and to comply with reporting obligations to join[64]. Within a year, one hundred mayors (whose cities had more than 167 million residents) had done so[65].

Given the range of trans-national interactions across a host of subject matters, one commentator forecast several decades ago that, « [u]nless America becomes a police state, municipal foreign policies are here to stay[66] ». Moreover, these efforts are deeply democratic, in the sense that they spring either from referenda enacted by majorities or from agendas of elected officials taking positions that resonate with their constituencies. And just as in the context of states’ relationships with migrants, conflicts abound about the merits of various kinds of state and local initiatives – some welcoming non-domestic engagement and others aiming to block it. When affected groups have resources to contest them, they argue that such initiatives breach the conventions of federalism.

Another example of the limits of subject-matter essentialism comes from state initiatives on human rights violations. In 1996, Massachusetts banned state purchases of products from what was then called Burma, where forced labor had become commonplace[67]. Within a decade, as awareness of the genocide in Darfur grew, several jurisdictions prohibited investments in the Sudan (« not on our dime » was the phrase used in congressional hearings on the issue[68]). Those decisions had potentially significant economic impact. According to a report by the U.S. Government Accountability Office, between 2006 and 2010, thirty-five states had « enacted legislation or adopted policies affecting their investments related to Sudan, primarily in response to the Darfur crisis, as well as in response to Sudan’s designation by the U.S. government as a state sponsor of terrorism[69] ». As a result, state treasurers and public pension fund managers had « divested or frozen about $3.5 billion in assets primarily related to Sudan in response to their states’ laws and policies[70] ».

But some of the affected corporations argued that states had no power to regulate the use of their own tax dollars because the matters were issues of « foreign affairs » over which the federal government had exclusive authority[71]. The critics succeeded in 2000 when the U.S. Supreme Court, five-to-four, concluded that the Massachusetts law intruded on federal power[72]. In a 2003 decision, the Court addressed a California statute, requiring insurers doing business in the state to disclose information about whether they had sold insurance during the Holocaust era; the Court held that the state rule was preempted by federal executive decisions[73]. In 2007, a lower court relied on these rulings to stop state-level Sudan divestment[74].

In response, Congress enacted the Sudan Accountability and Divestment Act, authorizing states to decide about investments in four arenas (military equipment, power, oil, and minerals), if sufficient notice and opportunities for comment were given to affected corporations[75]. Thus, the Act sought to recognize both the federal government’s foreign affairs authority and the states’ interests in controlling their own funds. The Act reflected that a set of rigid delineations (i.e., « domestic » versus « foreign » affairs) does not capture standard allocations across federations or withstand the changing content of the « domestic », the « foreign », and their intersections. The mix of transnational economic interests affected by new demands, produce conflicts about which « we » (to paraphrase Justice Kennedy in Obergefell) has the power to redress newly-understood « injustices ».

A parallel instability undermining essentialist claims resides in U.S. discussions of family, marriage, and « domestic violence », as reflected in debates about the constitutionality of the 1994 Violence Against Women Act (VAWA) and of same-sex marriage. Claims about necessary relationships between sovereignty, equality, gender, race, and households were made on all sides, by both state and federal actors.

A bit of history is in order, for the backdrop to federal legislation on violence against women has been state licensure and then toleration of such violence through doctrines such as a « marital rape » exception precluding men from being convicted of raping their wives[76]. During the second half of the twentieth century, critics documented the failures of local police, prosecutors, and judges to respond to victims of violence. As concerns about discrimination within the justice system mounted, the chief justices of many state courts chartered task forces to investigate « gender bias in the courts[77] ». The results were « voluminous » evidence of « pervasive bias in various state justice systems against victims of gender-motivated violence », and « unacceptably lenient punishments » for those convicted of « gender-motivated violence[78] ».

Yet the federal legislative response codified in VAWA was neither to regulate such officials nor to authorize lawsuits to obtain court orders mandating local and state officials to change their practices. Instead, the legislation, enacted in 1994, gave significant federal dollars to support improvements at the local level, created new federal penalties for interstate violence predicated on gender bias, and permitted victims of gender-animated violence to file lawsuits against their attackers[79]. The statute also authorized public enforcement through criminal provisions and private enforcement. Congress created new federal crimes – of « interstate domestic violence » and of « interstate violation of protection order » – to prosecute a « person who travels across a State line or enters or leaves Indian country with the intent » to « injure, harass, or intimidate » a « spouse or intimate partner » and « intentionally commits a crime of violence » causing « bodily harm » or who violates a protective order against « credible threats of violence[80] ».

On the civil side, Congress crafted a « civil rights remedy », permitting individuals alleging that private actors had targeted them to file in federal, as well as in state, courts. That remedy reflects just how state-protective the statute was. Rather than displace state law or authorize suits against state failures to police or prosecute, Congress only permitted private parties to sue assailants, if plaintiffs could prove that individuals had committed gender-motivated crimes.

But even that tepid civil rights remedy met with resistance. Before it was enacted, protests came from an organization of state court judges – the Conference of Chief Justices, another TOGA – a private group of individuals holding the position of chief justice in each state. Their claim was that if permitted to do so, women would take state-based claims relating to marriage, divorce, and property allocations and flip them into federal cases, thereby divesting state courts of their primacy over those issues[81]. Federal judges spoke through the United States Judicial Conference, a statutory body comprised of the chief judges of the federal appellate courts, a few district court judges, and chaired by the Chief Justice of the United States. The Conference argued that, if enacted, VAWA’s civil rights remedy would inappropriately relocate « family » disputes in the federal courts, which would be flooded by tens of thousands of cases[82].

While the formal objection was to volume, the subtext was about a risk to the identity of the federal courts, dealing with « national » problems rather than « familial » disputes. The conflict was tinged with concerns about maintaining the prestige and hierarchical authority of the federal courts, seen as at-risk by an infusion of large numbers of low-value claims founded on inter-personal disputes. State judges, in turn, seemed to want to protect their own identity as uniquely competent authorities governing family life. And, as this explanation reflects, the arguments of the judges from both state and federal courts conflated violence animated by gender with family life, rather than locating the injuries as harms to equality and economic agency.

Legislative debates about VAWA again exemplify disagreements among state actors about what serves their interests. Another TOGA, the National Association of Attorneys General (NAAG), joined with individual attorneys general from thirty-eight states to register support[83]. Many of these attorneys general were elected, and they were proponents of VAWA not only because it provided significant funds for state law enforcement and did not target states as potential defendants in lawsuits but also because it was politically popular[84]. A worldwide social movement was underway to reframe what once had been seen as interpersonal disputes or ad hoc crimes and had demonstrated that violence was a mechanism of subordination, cutting women off from full participation in economic and civic life.

That political predicate explains why VAWA was enacted. National legislators, like state legislators, were keen to announce that they understood that equality was at stake, given the harms that widespread violence imposed on women[85]. In terms of the legal authority to enact VAWA, Congress relied on both its powers to implement the Fourteenth Amendment’s guarantees of equal protection and its power over interstate commerce. One reason to mix the two was that Supreme Court case law had long imposed a « state action » requirement for Fourteenth Amendment Equal Protection remedies, and the focus of VAWA’s coercive measures were private actors. While VAWA proponents argued that the statute met that predicate due to state inaction in face of the knowledge of harms, the Commerce Clause appeared to provide even more secure footings, for Congress had relied upon it in the 1960s when enacting civil rights legislation banning race discrimination in public accommodations[86].

Flashing back to debates about the legality of the 1964 Civil Rights Act serves as a reminder about the many struggles over the boundaries of state and national interests. The 1964 Act aimed to prohibit racial segregation in « any inn, hotel, motel, or other establishment which provides lodging to transient guests[87] », as well as in restaurants, gas stations, and entertainment venues[88]. The ability of Congress to do so stemmed in part from its view that the « operations » of such places « affect[ed] commerce » by offering services to « interstate travelers » or because « a substantial portion » of the products used or sold have « moved in commerce ». In terms of remedies, Congress authorized individuals alleging violations of these rights to bring civil lawsuits for injunctions; further, the federal government could file actions if entities or individuals engaged in « a pattern or practice of resistance[89] ».

Federalism was central to attacks on the 1964 statute’s constitutionality. Opponents read the Commerce Clause as a limit on federal power that, in conjunction with the Ninth and Tenth Amendments, reserved authority to states. The owner of the aptly-named « Heart of Atlanta Motel[90] », which had 214 rooms, brought a lawsuit arguing that he « had followed a practice of refusing to rent rooms to Negroes, and … intended to continue to do so[91] ». Joining that private proprietor were the states of Virginia and Florida, which filed amici briefs arguing that to the extent the private businesses of hotels, restaurants, and gas stations could be licensed or regulated, state law was the source[92]. As Florida’s brief put it: « Interstate commerce having terminated prior to the sale and service of the food to the customer, there is nothing to support the application of the Civil Rights Act to the service of food in restaurants and other places of accommodation not bearing a direct and specific connection with interstate commerce[93] ».

In contrast, proponents – including the states of California, Massachusetts, and New York – interpreted the Commerce Clause as enabling the federation to function through authorizing congressional regulation of commerce, broadly construed[94]. As Massachusetts explained, the Commerce Clause represented « a major principle of federalism … essential to bind a loose confederacy into a strong union[95] ». California likewise argued that Congress had authority because the Civil Rights Act addressed « a matter of national concern ». Because some states could « choose not to act or to act in a manner inconsistent with the national welfare, the interests of both state and nation demand that our federal government remain free to act[96] ».

The Supreme Court chose the category of « the national » by framing a doctrine that defined the Commerce Clause as authorizing regulation not only of activities having a direct effect on commerce but also those having substantial effects on or burdening interstate commerce[97]. Given « the disruptive effect that racial discrimination has had on commercial intercourse[98] », Congress had the authority to respond[99]. The Court also recognized that Congress was dealing with « a moral and social wrong[100] » which, in U.S. legal conventions, is often associated with state « police » or general « welfare » powers rather than with federal authority. That concurrency did not, in 1964, undercut federal authority to act but did provide examples of what to do; the federal law had drawn on models from thirty-two states, which had parallel provisions, some going back decades[101].

This construction of federal power provided the structure for subsequent civil rights enactments, including VAWA, whose legislative history included analogies between the harm that violence did to women’s economic agency and the harm that segregated public accommodations did to blacks in the 1960s[102]. The result, in 1994, was the first congressional acknowledgment that equality, dignity, liberty, and economic agency were jeopardized by targeted, gender-biased violence. But 1994 was also the year in which the Supreme Court, in a 5-4 ruling, pruned back its Commerce Clause analyses by announcing that Congress had no power to regulate guns within a certain distance of schools[103]. Thus even as VAWA opened federal courthouse doors for victims of violence (if they could prove that their assailant acted with animus based on gender-bias), it became the target for attacks that doing so intruded on state authority.

In practice, the symbolism of VAWA’s national remedies was broader than its on-the-ground impact. Although federal judges had predicated opposition to VAWA’s civil rights remedy in part on floodgate arguments (that thousands of cases would be filed, overwhelming the docket), fewer than fifty federal cases relying on VAWA’s civil rights remedy resulted in published decisions between 1994 and 2000, when the Supreme Court heard the constitutional challenge[104]. On the criminal side, federal prosecutions were likewise small in number; from 1995 to 2000, fewer than 150 cases had been filed[105].

The case, Morrison v. United States, that the Supreme Court elected to hear involved a college student, who alleged that two athletes had raped her and then bragged about their sexual aggressions against women as women – demonstrating, she argued, that their violence was predicated on gender-biased animus under VAWA’s civil rights remedy[106]. The individual defendants argued that neither the Fourteenth Amendment nor the Commerce Clause licensed Congress to authorize access to federal courts for victims of violence in Morrison.

As in Arizona v. United States, many state officials supported the use of federal authority. Thirty-six state attorneys general filed an amicus brief, urging the Court to uphold VAWA’s civil rights remedy as a constitutional exercise of Commerce Clause authority[107]. Recall that in the 2012 Arizona litigation, sixteen states opposed and eleven supported federal power. In Morrison, only one state – Alabama – argued that VAWA’s civil remedy trampled on state prerogatives, which, Alabama averred, protected individual liberty[108].The brief explained that, despite « the nearly-unanimous showing of governmental support » for VAWA’s civil rights remedy, the Court had an « essential role » to play: « Federalism … is no less a counter-majoritarian right than the Constitution’s other structural guarantees of liberty[109] ».

And Alabama won. Unlike the 1960s Heart of Atlanta Hotel decision permitting the layering of federal anti-discrimination laws on top of state licensing and safety codes regulating hotels, restaurants, and gas stations, the Morrison Court held unconstitutional VAWA’s addition of a federal civil rights claim to state remedies[110]. To do so, the five-person majority decision by Chief Justice Rehnquist relied on a categorical essentialist approach that not only slighted Congress as a political repository of state interests[111] but also the thirty-six attorneys general who had argued that VAWA was federalism-friendly. Yet, to avoid overturning Heart of Atlanta Motel required describing the harms of violence to women as lacking the hallmarks of commerce that the Court had identified as the impediments that blacks had faced when wanting to eat at restaurants, buy gas, or sleep at hotels around the United States. The Morrison Court therefore insisted that violence was not « economic in nature[112] » – ignoring not only women’s agency as wage workers, but also their contributions as household workers to national prosperity and the economic impact of violence on the health care industry.

The Court opinion aimed to limit federal power more generally (as did its 2012 ruling on health care legislation[113]) by announcing a narrow reading of the Commerce Clause. The Morrison Court reasoned that, were Congress constitutionally permitted via the Commerce Clause to regulate violence against women, Congress could « equally as well » regulate « family law and other areas of traditional state regulation » (tort and crime included) that likewise could be said to have effects on the national economy[114]. Because « marriage, divorce, and childrearing » could, like violence against women, have an « aggregate effect » on the national economy[115], sustaining the civil rights remedy of VAWA could mean that « Congress might use the Commerce Clause to completely obliterate the Constitution’s distinction between national and local authority[116] ». The opinion thus embraces what the Court had rejected (again by a bare majority) in 1985 – that it could specify certain activities as « essential » state functions, immune from direct federal regulation[117].

Given the identification of the federal government as the font of marriage rights in the 2015 Obergefell ruling on same-sex marriage (resulting in high-profile conflicts with local authorities about the obligation to issue marriage licenses), the 2000 opinion quickly appeared anachronistic. But even when the Morrison decision was rendered, the Court’s jurisdictional constructions were mis-descriptions of the layers of federal constitutional and statutory law that then regularly interacted with state regimes related to families, as well as to criminal law.

The Obergefell decision, while novel in articulating same-sex marriage rights, is part of a long line of federal cases regulating families. Constitutional federal family law has prohibited racial barriers to marriage[118], regulated procedures by which states determine parental rights[119], authorized parents to direct educational choices of their children[120], and insulated custodial parents from state rules conferring rights on grandparents[121]. Statutory federal family law has likewise provided an overlay – as well as the reminder that national authority does not intrinsically provide progressive rulings. In 1996, Congress insisted in the « Defense of Marriage Act » (DOMA) that, for purposes of federal law, « marriage » can only occur between a man and a woman[122], a proposition eroded at the state level through a series of decisions by state judges and legislatures that made such restrictions unlawful. By 2013, the state cases helped to produce federal law, with a ruling in Windsor[123] that refused to apply DOMA to federal tax law and then in 2015 in Obergefell.

Moreover, federal family law not only is found under the title of « marriage », but also is embedded in statutory regimes governing bankruptcy, pensions (the Employee Retirement Income Security Act, or ERISA), tax, and immigration, all of which oversee aspects of family life by allocating benefits and burdens depending on status relationships[124]. Indeed, as Kristin Collins has documented, from military pensions provided in the 1790s through social security and veterans’ benefits, federal law entails decisions about what relationships merit recognition as marriages[125]; as a result, « sovereignty-based accounts » of the law of the family are « misguided and misleading[126] ». Despite the Morrison Court’s claim of jurisdictional essentialism, family law in the United States is both state and federal, laced as well with international precepts, such as the Hague Convention on the Civil Aspects of International Child Abduction[127].

Just as the jurisdictional divide unravels, so too does the categorization of the kinds of injuries that violence entails. The Morrison Court assumed (correctly) that violence can be both tortious and criminal. But violence can also be an attack on equality rights. The historical analogue proffered by VAWA proponents was the lynching of African-Americans which, under slavery and thereafter, was rarely subjected to criminal or tort law sanctions, let alone found to be a violation of civil rights. (Indeed, in the 1930s, a county courthouse in Idaho displayed a mural of an « Indian being lynched » as if the depiction of such an act provided an appropriate decorative touch[128].) Not until after the Civil War and well into the Second Reconstruction of the 1960s did lynching and other race-targeted violence come to be understood as torts, crimes, and civil rights violations.

Thus, the Morrison Court’s decision is steeped in delineating bounded categories (state/federal, family and criminal law/civil rights law) as if those distinctions enhanced accountability through articulating specific lines of authority. Its essentialism obscured that judges – from Morrison to Obergefell – play critical roles in deciding what falls within jurisdictional boundaries and when to embrace the multiple forms that rights can take. When refusing to acknowledge the local and national consensus that gender-animated violence was simultaneously a violation of local law and of national constitutional rights, the Morrison majority used federalism to obscure its own agency in sapping violence of its equality and its economic valence. As Justice Souter’s dissent put it, « the States will be forced to enjoy the new federalism whether they want it or not[129] ».

But jurisdictional essentialism cannot stop rights claims from crossing boundaries. During the latter part of the twentieth century, the disproportionate degree of violence visited upon women became the subject of local, state, national, and transnational law; indeed, the 1994 enactment of VAWA was part of activities crossing jurisdictions. Two years earlier, in 1992, the U.N. Committee empowered by the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) issued a General Recommendation, « Violence Against Women », outlining the degree to which violence contributed to women’s subordination[130]. Thereafter, international criminal tribunals recognized rape as a war crime, and the treaty launching the International Criminal Court included gender-based violence within its definition of crimes against humanity[131].

In 2011, the Inter-American Commission on Human Rights categorized gender-based violence as « one of the most extreme and pervasive forms of discrimination, severely impairing and nullifying the enforcement of women’s rights[132] ». The context was a case from the United States in which a woman had obtained a protective order against her estranged husband. Despite her efforts to enlist the local police employed by the Town of Castle Rock, Colorado to implement the order, her husband killed their daughters. She lost her suit against the town in the United States Supreme Court, which held that federal constitutional law gave her no protection against police inaction[133]. In 2011, the Inter-American Commission disagreed; it ruled that the United States had violated « the State’s obligation not to discriminate and to provide for equal protection before the law under Article II of the American Declaration[134] ».

In the same year, the Council of Europe promulgated a new Convention on Preventing and Combating Violence Against Women and Domestic Violence, to « [p]rotect women against all forms of violence, and prevent, prosecute and eliminate violence against women and domestic violence[135] ». In sharp contrast to current United States doctrine, recognizing no constitutional duties of protection even when state statutes mandate that police enforce protective orders, the ECtHR has concluded that states are obliged to respond to violence imposed by private as well as public actors[136]. In 2013, in Valiuliené v. Lithuania, the ECtHR held that Lithuania’s failure to provide remedies for a woman victim of violence breached Article 3 of the European Convention on Human Rights because the state had failed its positive obligation « to ensure that individuals within their jurisdiction are not subjected to torture or inhuman or degrading treatment or punishment, including such ill-treatment administered by private individuals[137] ». In other cases, the ECtHR has concluded that, in addition to inhuman and degrading treatment, violence to which the state fails to respond can violate rights to life, to respect for private and family life, to fair trials, and to equality[138], atop its valence as contravening domestic tort and criminal law.

That resulting positive obligation, announced in several cases, calls for countries to put adjudicative structures in place to respond to violence visited on « vulnerable » individuals, confronted with household violence[139]. Required are « effective criminal-law provisions … backed up by law-enforcement machinery[140] ». The ECtHR describes itself as deferring to the judgment of national authorities and only ensuring that « a State’s obligation to protect the rights of those under its jurisdiction is adequately discharged[141] ». Doing so is its own task; repeated applications from particular jurisdictions make plain that these obligations to provide remedies do not easily translate into action[142]. Whether a focus on criminalization is a wise or sufficient response is of course appropriate to debate, but the insistence on protection of and remedial rights for « vulnerable » persons has become a facet of European law[143].

The 2000 Morrison judgment limited federal judges from joining this inter-jurisdictional exchange about the scope of government obligations about violence, once that practice is understood to be a method of subordination. But Morrison has not shielded the federal courts from being identified as central to the structure of family life. As noted, over objections that marriage was beyond the scope of the Supreme Court’s own powers, five justices – building on dozens of state decisions recognizing same-sex marriage, held in 2015 that the Constitution required states to license it.

As in Arizona v. United States, Heart of Atlanta Motel, and Morrison, states in the same-sex marriage argument diverged on what rules were in states’ interests. Fifteen states, led by Louisiana, argued that their decisions on marriage were « the most elementary form of self-government guaranteed by our Constitution », and that only the « democratic process » could constitutionally alter such laws through the development of a consensus « at the state level[144] ». In its independent filing arguing for a presumption of constitutionality for state rules, Alabama explained: « this case is about more than marriage. It is also about the proper role of the federal courts in scrutinizing state policy decisions[145] ».

Yet the 2015 Obergefell majority focused instead on the foundational importance of individual rights to the American constitutional order that rendered the denial of marriage licenses an « injustice[146] ». Indeed, the term « federalism » is mentioned by name in only one of the dissents[147]. The Obergefell majority’s decision to ignore federalism was not the first time it gave little attention to arguments from localism in marriage discrimination. In the 1960s, when ruling in Loving v. Virginia that criminalization of biracial marriages breached the Constitution, the Court likewise spent little time on state authority over family life, and only noted that Virginia had not argued it had « unlimited » authority « to regulate marriage[148] ». Virginia had staked its claims on the history of the Fourteenth Amendment and argued that in the nineteenth century, it had not been intended to dislodge statutes on biracial marriages. Virginia asserted that because questions about whether interracial marriage was harmful required analysis of scientific evidence and social policy, the issue was one for state legislatures to decide, rather than by either state or federal courts[149].

This dip into marriage discrimination debates offers another example of the plasticity of state interests. By 2015, Virginia had developed a very different understanding of its own state interests. Virginia submitted an amicus brief in support of same-sex marriage. While extolling the state’s « contributions to America’s exceptional form of democracy », Virginia also described its own repeated failures to understand the meaning of equality[150]. « Virginia’s government fell short of fidelity » to « equality-of-right » when the state defended public school segregation in the 1950s, bans on inter-racial marriage in the 1960s, and the exclusion of women from its military institute in the 1990s[151]. Each time, Virginia had « invoked federalism » to argue that « education policy and marriage regulation » were « quintessentially State prerogatives that federal courts should leave alone[152] ». In 2015, Virginia abandoned those claims and argued for recognition of a national equality right to same-sex marriage.

Because essentialism seems easy and comfortable, I have provided details of shifts in understanding. Various arenas of social ordering – foreign affairs, migration, violence, and marriage – that have been linked to one level of government gain new valences through political and social movements. In each, rules emanating from different levels of government overlap, and conflicts emerge[153]. The judicial resort to essentialism ought not obscure that judges – and not federalism, pre-ordained – resolve the conflicts as they choose where constitutional identities lie and when to insist that diversity is intolerable or not.

 

III. Federalism Discounts and Court Discounts: « Reasonably Wrong » Convictions and « Margins of Appreciation » on Prisoners’ voting rights

 

Another mediating mechanism is what I term a federalism(s)’ discount, resulting in a given case either in the differential enforcement of specified legal precepts or in the under-appreciation of the identitarian claims of subunit authority. My examples come, again, from the United States and Europe, where bodies of law direct judges at the national level to think about the identity of the subunit when ruling on issues of individual rights.

In the U.S., the law of post-conviction remedies offers an illustration. The right of judicial review is predicated on federal constitutional guarantees that the writ of habeas corpus not be « suspended[154] ». Federal statutes translate this precept into opportunities for prisoners, held in state or federal custody, to bring to federal courts claims that their convictions or sentences violate federal rights. Akin to the European requirement that claimants before the ECtHR seek redress before national courts, in the U.S., prisoners in state custody must « exhaust » available state remedies. In recent decades, both Congress and the Supreme Court have sharply narrowed the grounds for review by requiring federal judges to decide not only that state court judgments wrongly applied clearly established federal precedents but also did so « unreasonably ».

That inquiry resembles the European doctrine of the « margin of appreciation », which requires judges sitting on the European Court of Human Rights (ECtHR) to assess whether legal positions taken by Contracting Parties are « outside » a range of reasonable variations. A voluminous literature addressed the European doctrine of the margin of appreciation, begun in the courts and now being enshrined in Protocol No. 15 to the European Convention on Human Rights and becoming linked to the European Union’s doctrine of subsidiarity[155]. When the ECtHR relies on the margin to accede to a national judgment, it could – in U.S. terms – be understood as permitting a fair-minded but potentially wrong decision to stand. And when the Court declines to do so, it may be read to rule that the subunit’s stance is outside the boundaries of reasonableness. In both the U.S. and Europe, the degree to which the subunit’s position is an outlier has relevance, as judges consider the clarity of a right (in U.S. parlance, « clearly established ») and, at times, explore whether a « consensus » (to borrow from Europe) exists about a right’s import.

Further, and again paralleling the U.S. federalism discounts, the margin can be read as an effort to calibrate the institutional challenges that implementation could pose[156]. The ECtHR relies – like all courts – primarily on voluntary compliance for implementation of its orders, complemented by help from the executive branch, in this instance the Committee of Ministers[157]. Given that the ECtHR is committed to reading the Convention as a « living » document[158], federalism discounts presume the desirability of spaces for debate and revision both within and across Contracting Parties. The margin could also be understood to be « living » in that it can and has shifted, as demonstrated by a series of cases challenging the refusal to change birth records to reassign genders. In 1986, a challenge to a U.K. provision prohibiting revisions was denied, but in 2002, the ECtHR decided that the U.K. rule « no longer » fell within the margin[159].

The ECtHR’s 1976 Handyside judgment is the oft-cited origin of the doctrine; the Court described itself as according a « margin » and a « power » of appreciation by permitting the United Kingdom to destroy a Danish book, claimed to violate U.K.’s Obscene Publications Act, over objections from the publisher that doing so violated the European Convention’s free expression rights[160]. The ECtHR’s formal explanation of its deference was that the national government, acting pursuant to law, was better situated to assess the need to protect youths than was the European Court. As in the context of habeas corpus in the United States, the result could be characterized as changing the underlying right by altering the test of free expression or as acquiescing to a deviation from that Convention right. A key feature of discounts – as detailed below – is that the breadth of a deviation is not readily assessed in the individual instance. The contours depend on subsequent rulings.

Since the 1970s, the ECtHR has regularly resorted to « the margin of appreciation », often in conjunction with discussion of whether a « consensus »[161] exists among the Contracting Parties (now numbering forty-seven members) about a particular issue[162]. (Consensus may imply a degree of unanimity that has prompted one commentator to argue that looking for a « trend » would be a better description of the nature of the inquiry[163]). The margin/consensus analytics intersect with another technique of interpretation: proportionality, which launches a series of evaluative decisions about whether a right has been violated and, if so, whether the state did so in accordance with law, had permissible reasons to do so, and the intrusion was proportionate to the state’s interests[164]. Just as the U.S. Supreme Court is regularly criticized for inconsistent or uneven application of jurisprudential federalism doctrines, the ECtHR’s margin case law offers many variations – prompting commentators to argue the wisdom of applying this form of a discount.

Federalism discounts deepen or relax commitments to particular rights by marking some to be central to the identity of the subunits or to the larger configuration. All of the decisions entail a series of judgments about the existence and scope of rights, the degree of variation, and how to balance competing claims. When subunits are given discounts, asymmetries can result. Those who reside within that subunit may have more or less protection than individuals in other subunits; uncertainty about the expansion (the « width ») of margins likewise follows. When rights claims prevail, questions of compliance and of backlash can emerge. These federalism discounts are generally explained as protective of subunit decision-making but they can also serve the central authority by enabling it to avoid confrontations about the implementation of rights in unreceptive venues.

Terms such as « incrementalism » and « soft nudges » or « hard shoves » are sometimes deployed to talk about aspects of the interactions I discuss here. Yet such nomenclature overstates the relevant actors’ capacity to calibrate their decisions (often negotiated in a multi-judge court) and to predict the impact and subsequent course of events. Whether an action turns out to be a gentle nudge or a hard shove can be ascertained only later – after particular rulings gain (or not) saliency and metrics are chosen to assess the impacts. My choice of the word discount (with deliberate echoes of economic valuation) aims to underscore that authorizing variation may be loyal to federalisms’ commitments to diversity but also entails tolerating deviation from a previously-set norm. At the time that the discount is given, one cannot know whether (to keep the commodification metaphor) the price of the contested behaviors will go up or down thereafter, whether the subunit will conform, and whether the larger unit will be besieged or praised for its decision.

Below, through detailing aspects of the U.S. law of habeas corpus and conflicts in Europe about prisoner voting rules, I explain why discounts are normatively preferable to essentialist approaches, even as I also detail the difficulties of assessing what interests are served and which harms imposed by the mediating possibilities of discounts. Further, not all discounts are the same and when discounts incorporate jurisdictional essentialism (as does the form of blanket exemptions from federal oversight currently provided by the U.S. habeas jurisprudence), discounts lose their potential for creative mediation.

 

A. Giving Discounts: « Fair-minded » Disagreements about the Meaning of the Constitutional Rights of Criminal Defendants in the United States

To understand the context prompting federalism discounts in criminal constitutional law in the United States requires a quick review of the different demands placed on state and federal courts. State courts not only predate the creation of the federal system but also have far outstripped federal courts in terms of the growth and breadth of their dockets. The lower federal judiciary was (and is) small in part because, in the contemporary language of American federalism, state courts were (and are) « commandeered » to enforce federal law[165]. The Supremacy Clause of the Constitution obliges « Judges in every State » to enforce federal law, « the supreme Law of the Land[166] ».

But in the wake of the Civil War, Congress saw the rejection by state judges and juries of federal equality obligations through the persistent subjugation of former slaves. Legislative responses included expanding federal court jurisdiction to enable public and private enforcement of the developing federal norms. In 1867, Congress first gave federal judges the authority to review claims by state prisoners arguing that their custody was in violation of federal rights[167]. That legislation, one of many providing new grounds for federal court jurisdiction, was matched by new investments in federal construction, creating a visible « federal presence » through investing large amounts of federal dollars in buildings in localities around the United States[168].

In the decades thereafter, federal laws multiplied, as did the girth of federal institutions. Yet the federal judiciary remains a tiny workforce, with some 850 authorized life-tenured judgeships[169], assisted by senior life-tenured judges and a comparable set of statutory judges (magistrate and bankruptcy) serving term appointments. All told, about 2,000 judges work in the more than 550 federal courthouses around the United States. The docket is likewise small; federal courts receive between 350,000-400,000 civil and criminal filings a year, plus more than one million bankruptcy petitions[170].

In contrast, more than 30,000 judges sit on state courts and do more than 95 percent of the adjudicatory work in the United States. They deal with 90-110 million cases filed annually that range from traffic citations to family, contract, tort, statutory, and constitutional disputes[171]. Thus, state judiciaries continue to bear the primary responsibility for the implementation of both state and federal law. Specifically, about 60,000 criminal filings come into the federal courts annually; state courts deal with more than 20 million criminal cases yearly.

The quality of that enforcement and the relationships between the state and federal legal systems have prompted a host of accommodations, developed through statutes, the common law, and interpretation of the Constitution. Although Supreme Court justices debated allocations of authority, they did so in terms such as « states’ rights » rather than by reference to « federalism ». That word is not found in published opinions by the federal judiciary until the 1930s. Initially, « federalism » was used to describe the political configuration of federations in the United States and elsewhere.

Reliance on « federalism » as a legal justification for deferring to states emerged in the later part of the twentieth century, when « states’ rights » were linked to opposition to the Second Reconstruction of the 1960s civil rights era. In the 1970s, Justice Black invoked « Our Federalism » to explain that, even in the face of allegations that state prosecutions violated federal constitutional rights, federal courts were not to enjoin pending criminal prosecutions[172]. In the decades thereafter, an expanding law of immunities of states and their employees from lawsuits alleging violations of federal statutory and constitutional rights came to be explained in federalism terms[173], as has been the general freedom of state judges to interpret state law free from federal judges’ oversight[174].

Federalism is a doctrine of deference that, as discussed above, often entails essentialism and, as explicated below, can provide a discount. But it is not a blank check. Thus, even as the U.S. Supreme Court has expanded a variety of doctrines so as to insulate state judges and executive officers from review, the Court has also in recent decades regularly displaced state law through the doctrine of preemption (as the decision in Arizona v. United States illustrates) and through constitutional interpretation (as in Obergefell v. Hodges). These rulings eclipse areas of state tort, criminal, and family law, even though the Morrison majority, ruling on the Violence Against Women Act, had posited these categories as iconic examples of state court exclusivity.

Turning to the interaction of federal and state systems in post-conviction litigation, a role for state courts is built into the statutory scheme. Congress requires that, before a federal judge can decide whether a state prisoner is in custody « in violation of the Constitution or laws or treaties of the United States[175] », a prisoner has to make that claim in state court. Federal judges are thus put into the position of reviewing two state court processes – the initial proceeding (a trial or guilty plea) producing the conviction and then a subsequent state habeas proceeding.

One might assume that the federal judge – presumptively expert on federal constitutional rights – is then empowered to decide anew. But beginning in the 1970s, the Supreme Court added layers of constraints on federal court habeas jurisdiction. Fourth Amendment claims were barred if a defendant had had a « full and fair opportunity » to raise the claims in state court[176]. Failure to comply with state procedural rules could likewise result in preclusion, absent a showing of « cause and prejudice » – interpreted to mean something more than inept lawyering[177]. Criminal defendants had to establish a reason external to their case (such as a hidden document) for the failure to raise a claim and further, that if the information had been known and the claim raised, the outcome could have been different. Moreover, direct allegations of unconstitutionally inadequate counsel services are reviewed, after conviction, with deep deference. In the 1980s, the Court held that to show ineffective assistance of counsel, a convicted prisoner had the burden of demonstrating both that the lawyering was below what a minimally adequate lawyer would have done and again that prejudice resulted: that, as a result, « there is a reasonable probability » that an outcome would have been different, « absent the deficient act or omission[178] ». Further, habeas corpus cases could rarely be the basis for developing new readings of constitutional obligations[179].

In 1996, Congress codified a good many of the judge-made restrictions and added some of its own. Included was the directive that federal judges accord finality to state court judgments denying prisoners relief on a claim, if

adjudicated on the merits in State court proceedings unless the adjudication … resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States[180].

Once a state issues a decision, this statute’s « highly deferential standards kick in[181] ». In 2011, the Supreme Court explained that this provision ensures that federal review is limited to « extreme malfunctions in the state criminal justice systems » rather than « ordinary » errors[182]. Federal overrides are permitted only when « there is no possibility fair minded jurists could disagree that the state court’s decision conflicts » with the Court’s precedents[183].

Under this test, explained as avoiding intrusion on « state sovereignty »[184] as well as respecting states’ « sovereign power to punish offenders and their good-faith attempts to honor constitutional rights[185] », a state prisoner has to show that the state court’s ruling « was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fair-minded disagreement[186] ». Federal judges therefore have to puzzle about what kinds of interpretations merit the characterization of « fair-mindedness » and the differences between « ordinary » errors and « reasonable » misapplications of federal law, as well as when to characterize processes or outcomes as « extreme malfunctions[187] ». That charter became more demanding when the Supreme Court held that even if state courts issue only a few written lines (« postcard » judgments, which are commonplace) and do not give reasons for denying petitions, federal judges must assume that reasons exist, supply them, and then accord them deference[188].

How is one to assess the impact of rulings that federal law formally applies but that leave states largely unsupervised? In terms of individuals, the case law is replete with instances leaving habeas petitioners with federal rights deprived (but not « unreasonably » so) and incarcerated for life or subject to the death penalty[189]. In terms of the rights themselves, the discounts largely tolerate dilution of rights, undermining their import and licensing further deviations or under-enforcement[190]. In terms of costs to federal judges, the labyrinth that the doctrine requires them to explore results in long decisions focused on parsing the kinds of errors made rather than the underlying justice of letting a conviction or a sentence stand[191]. Lower court judges often write lengthy opinions, reflecting the time required to parse records to decide whether a petitioner (usually self-represented) has successfully surmounted the various federal procedural hoops[192]. In addition to workload, some federal judges bridle at being tasked with upholding judgments that have federal constitutional failings[193].

In contrast, the Supreme Court’s majority describes its system as respectful of state courts, on the front lines of criminal justice, of state citizenry protected by their criminal justice system, and of federal judges who, despite the current labor-intensive review if dealing thoroughly with files, are freed from even a greater torrent of filings. In addition to buffering federal court from potential cases, the Court has also protected its judiciary from the potential struggle with state systems over implementation of federal constitutional rights.

Whether states benefit from the under-enforcement of federal constitutional rights requires more inquiry. For example, federal court mandates – such as requiring adequate prison conditions – have sometimes worked to augment criminal justice budgets by providing leverage in some state legislatures to gain resources[194]. Moreover, on occasion, states have argued for the Court to expand federal constitutional rights. In the 1960s, twenty-three states filed an amicus brief in Gideon v. Wainwright – urging the Court to read the Sixth Amendment as requiring states to provide lawyers for unrepresented indigent felony defendants [195]. Even as many states had, before 1963, a nascent system of defender services[196], the federal ruling provided a basis for expansion[197]. And even when a federal judge rules on behalf of a habeas petitioner, the order is limited (in my terms, state regarding), because it requires the release or retrial of an individual but does not include structural injunctive relief. The question of compliance, going forward, is left to the states to fashion.

Under-enforcement of federal constitutional rights may have imposed another form of cost on states. Commentators across the political spectrum now embrace the view that the country has been too punitive; some two million people are in jails and prisons[198]. State budgets are swamped by the expense of housing this aging population, in need of medical and mental health care.

Some states have responded by obtaining funds from criminal defendants, charged if put on probation, or fined. State and local prosecutors send scores of individuals through low-cost prosecution « mills » – putting them into jails or extracting sums of money and rolling them into debt. The racially disparate impacts are well known[199]. In 2008, 1 in 100 adults were under state control; « one in 30 men between the ages of 20 and 34 » was « behind bars », but for « black males in that age group the figure [was] one in nine[200] ». By 2015, police shootings of black men in several cities and the tactics of local prosecutors and judges dealing with poor individuals subjected to fines had produced new documentation of the distrust of law enforcement and declining social cohesion[201].

The federal courts could have helped by insisting decades ago on closer compliance to federal obligations requiring real investments in criminal defense services, obliging prosecutors to disgorge exculpatory information, and insisting that judges screen for racial bias in grand and petit juror selection and in trials and sentencing. In contrast, the Court has refused the opportunity to use habeas corpus to monitor quality, just as it has shaped other doctrines that limit federal judges from overseeing state policing and prosecution when claims of violations of rights are made.

Instead of particularized decisions that at times provide for temporizing accommodations, for example, in response to an unusual spike in a state’s criminal caseload or other special circumstances, the Supreme Court has accorded discounts to state criminal procedures across the board, justified through essentializing criminal prosecution as an arena of state governance. The result has been to underprice what states ought to have to spend, were they to convict and sentence people in full accordance with constitutional obligations. The costs are also borne in the federal system, as judges have to respond to tens of hundreds of unrepresented prisoners, tasked with navigating detailed procedural requirements.

Moreover, the Supreme Court has undervalued its own method of engaging with state criminal justice systems. If and when an individual succeeds, the judgment formally relates only to that person, leaving states free to shape methods to implement the rules prospectively through what Robert Cover and Alexander Aleinikoff called « dialectical federalism[202] ». The posture again has parallels in Europe. The ECtHR requires its member states to provide protection from violence for vulnerable individuals through avenues for adjudication and sanctions. But the method of doing so is left to the national level[203].

Returning to the United States, the « unreasonably wrong » approach is not only an artifact of federalism. The same doctrines imposing barriers to habeas are applied to federal as well as to state prisoners, alleging violations of their constitutional rights by federal prosecutors, judges, or ineffective lawyers[204]. In this sense, the discounts are not genuinely respectful of states as specially situated actors, but predicated instead on concerns about crime control that license expanded government control, whatever the level at which it occurs. The claimed deference to distinctive political communities to respect their autonomy collapses into an undifferentiated toleration of a landscape of punishment. The result is that neither federal and state judges nor legislatures are required to spend their time assessing how to respond to the failings of constitutional pronouncements of rights, even as policing, prosecution, and prisons are widely regarded as profoundly deficient, in legal, moral, and economic terms[205].

Just as in Morrison, when the Court’s ruling took federal courts out of addressing the relationship of violence to equality, the habeas doctrines cut off explorations of the Constitution’s relationship to the system of criminalizing behavior and to the questions of when to term that process just. The most important juncture of individuals’ relationship to the state is left with little by way of substantive constitutional content about the boundaries of government intrusion. Federal court involvement is no panacea, but its absence has not only reduced the pressures that federal rulings could have imposed but also reduced input from the country’s most well-resourced jurists.

 

B. Declining to Discount: Voting by Prisoners under the European Convention

Having provided a sense of the complexity of parsing what harms and benefits flow from authorizing deviations from centrally-avowed rights in the United States and the defects of blanket discounts, I turn now to an example of an initial refusal to provide a federalism discount and some of its ramifications. My exemplar is the holding by the European Court of Human Rights in 2005 in Hirst v. U.K.[206] that the U.K.’s ban on voting rights of prisoners violated Convention rights – followed by that court’s 2011 pilot judgment continuing to find the U.K. ban unlawful; the ECtHR’s 2012 approval of an Italian prisoner voting ban for those incarcerated for five years or more; and the ECtHR’s refusal to award filing costs in 2015 and in 2016 to prisoners complaining about the U.K.’s refusal to alter its ban[207]. The details make plain the choices when characterizing rights, assessing jurisdictions’ rules if searching for consensus, calibrating the kind of discount given[208], and repeatedly assessing the impact on both courts and subunits of the decisions made.

The debate about the obligation to provide some form of voting rights to prisoners centers around Article 3 of Protocol No. 1 to the European Convention on Human Rights, which provides that « the High Contracting Parties undertake to hold free elections at reasonable intervals by secret ballot, under conditions which will ensure the free expression of the opinion of the people in the choice of the legislature[209] ». This text prompts several questions including whether it creates an individual entitlement to its enforcement, which elections are covered[210], who constitutes « the people », and what if any restrictions are permissible. During the last two decades, the ECtHR has responded by reading this formulation to create an implied individual right-to-vote and by ruling that restrictions of this right must be proportionate[211]. Thus, limits on universal suffrage have to have legitimate aims, translated to include enhancing civil responsibility, respect for the rule of law, and assuring the proper functioning of democratic regimes[212].

The focal point of the current debate is the U.K. practice that has varied over time. From 1870 until 1967, England did not let prisoners convicted of felonies vote but permitted individuals « serving sentences of less than 12 months » for misdemeanors to vote[213]. In 1983, the U.K. changed its rule and prohibited all persons convicted of crimes and incarcerated from voting in « any parliamentary or local government election[214] ». In 2000, the U.K. modified that provision by restoring voting to people detained in prison but awaiting trial[215]. In 1998, the U.K. enacted its Human Rights Act, making European Convention rights domestically applicable[216]. In 2001, the English courts – relying on a proportionality analysis – denied a challenge to the prisoner voting ban[217]. In 2005, in Hirst v. U.K., the ECtHR first ruled on whether the 1983 provision violated the Convention right to vote[218].

Before the ECtHR, the U.K. argued that it was owed deference in constituting its own electorate and therefore that it could make the determination to exclude, temporarily, individuals imprisoned for violating its laws[219]. Moreover, the U.K. explained that its voting restrictions were tailored, in that they were limited to those whose violation of the laws was « serious enough » as to result in incarceration, rather than applied to civil contemnors or pretrial detainees; further, that the disenfranchisement lasted only during the period of incarceration[220]. But the 2005 Hirst ruling rejected the U.K. ban.

When doing so, the ECtHR discussed practices around Europe to ascertain the degree to which the U.K. stood apart. The court counted eighteen countries that permitted prisoners to vote without restriction, thirteen in which prisoners could not vote, and twelve that authorized voting with some limits[221]. The ECtHR also widened its lens, to consider court decisions in Canada and in South Africa, both of which had mandated that prisoners be permitted to vote[222]. As for the rights at stake, the ECtHR characterized the Convention right as « vitally important[223] » and chastised the English Parliament for providing no evidence of efforts to « weigh the competing interests or to assess the proportionality of a blanket ban [224] ». Noting that « the margin of appreciation is wide, [but] it is not all-embracing[225] », the Court held the U.K. in violation of the Convention[226]. The ECtHR then left « it to the legislature to decide on the choice of means for securing » the voting rights guaranteed by the Convention[227].

Given that the statute at issue dated from 1983, that it was modified to exclude pre-trial detainees in 2000, and that the U.K. had acquiesced in other high-saliency ECtHR judgments (such as the obligation to end discrimination based on sexual orientation in its navy[228]), one might have expected compliance. But resistance followed. In the years after the Hirst decision, political parties hostile to involvement in Europe gained power in the U.K. and groups within Scotland, arguing for greater independence, challenged the authority of the U.K. In 2010, other prisoners returned to the ECtHR to seek relief; the applicants in Greens and M.T. argued that the British ban had prevented them from participating in European Parliament elections in 2009 and that they would lose their opportunity in 2011 to vote in elections to the Scottish Parliament[229].

The ECtHR admonished the U.K. for its continuing violation of Convention rights and its failure « to abide by the final judgment[230] ». The Court directed that, « in light of the lengthy delay in implementing that decision and the significant number of repetitive applications », the U.K. was, within six months, to propose legislation to amend its felon disenfranchisement laws to be « Convention-compliant » and thereafter to enact such legislation « within any such period as may be determined by the Committee of Ministers[231] ».

That decision was again met with resistance. « Britain must stand firm against this growing abuse of power by unaccountable judges[232] » was what a member of Parliament said, while the Prime Minister opined that it made him « physically ill to contemplate giving the vote to prisoners. They should lose some rights including the right to vote[233] ». In 2011, the country’s second most senior judge provided assurance that « the domestic courts would not interfere if Parliament chose to reject the controversial decision », as the issue was a « ‘political decision’ and if the Government chose to ignore a Strasbourg ruling there would be ‘nothing objectionable’ in British law[234] ».

As Britain resisted, the ECtHR ruled in 2012 on prisoner voting in Scoppola v. Italy, in which the Court upheld an Italian provision, imposing a permanent disenfranchisement (with a slim chance of its removal) on individuals serving prison sentences of five years or more[235]. The Scoppola ruling distinguished Hirst by noting that the Italian ban did not impose « disenfranchisement in connection with minor offenses » and was not « general, automatic, and indiscriminate[236] ». Rather, individuals had the option to apply to the government to obtain exemptions and regain voting rights.

The lone dissenting jurist in the Scoppola decision viewed his colleagues’ judgment as a rejection of Hirst. He argued that the Italian rule was as « blunt » an « instrument » as the U.K. provision struck in Hirst, and he complained that by affording a margin of appreciation to the Italian ban, his colleagues had « stripped the Hirst judgment of all of its bite as a landmark precedent for the protection of prisoners’ voting rights in Europe[237] ». As another member of the Court put it in 2014, it was « difficult » to read Scoppola as « anything other than a partial overruling » of Hirst[238].

Another kind of cutback came in the 2015 decision in McHugh, ruling on more than 1,000 applicants and finding that the U.K. had not complied with its Hirst ruling. The ECtHR refused to order the payment either of damages to the individuals or the legal costs of the submission[239]. In 2016, the Court responded to another 22 applicants, and again concluded that a « finding of a violation constitutes in itself sufficient just satisfaction », and awarded no other forms of relief[240] – thereby giving little incentive to lawyers to bring it more claims. And, while the majority in McHugh justified its refusal to provide those costs on the basis that the rules were so clear and therefore the cases were so easy to bring that no lawyer was needed, the two dissents stated that the Scoppola judgment had opened the question of whether the Court could be persuaded to reject the Hirst rule[241].

Indeed, dissenters in 2014 in another application argued that the Court had been wrong on the merits and that national governments, and not the ECtHR, were charged by the Convention with the « further realisation of human rights and fundamental freedoms[242] ». Additional arguments that European law permitted restrictions on prisoner voting came from a 2015 decision by the Court of Justice of the European Union, which held that a French law (of narrow application) had not deprived the relatively few prisoners affected by rights protected under the Charter of Fundamental Rights of the European Union[243].

How is one to assess the decade of interactions and the impact of the refusal to provide a discount (via the margin) in Hirst and the interactions thereafter? After its ruling in Scoppola v. Italy, the ECtHR gave the U.K. extensions to comply. That 2012 decision seemed to provide an easy route for compliance; all that was needed was to relax the absolute ban and differentiate among prisoners. Thereafter, a Parliamentary Committee received testimony on what it could do, with options such as banning voting for prisoners serving four months or more, or for prisoners serving four to six years, or dropping the ban, or ignoring the European Court’s ruling and continuing to apply the prohibition to all prisoners in custody[244]. The hundreds of pages of submissions to the Parliament made plain that, as former Home Secretary Jack Straw put it, the U.K. has « paid huge attention » to the European Court’s ruling[245], even as it has not, as of 2016, altered its rules on prisoner voting.

Straw argued that the U.K. ought not change its practice because a decision to incarcerate was itself an approach to voting that was not an untextured ban; custody was, in his view, a reasonable proxy for the « seriousness » of an offense[246]. (As John Finnis later opined, because English sentencing practices result in fewer than ten percent of those convicted being sent to prison, disenfranchisement was not automatic but applied selectively to a targeted subset[247]). Further, in a 2016 report to Parliament on how the U.K. has responded to « Human Rights judgments », the Lord Chancellor and Secretary of State for Justice described the « UK’s record on implementation of ECtHR judgments » by noting the country was, as of 2015 « responsible for 19 (0.2%) of a total of 10,652 pending judgments before the Committee of Ministers », putting it 31st of out of 47 countries[248]. That report, with detailed appendices, both described the plan to set out proposals for a Bill of Rights and seemed intent on documenting that, for the most part, the U.K. was compliant with, rather than antagonistic to, human rights precepts.

Yet to focus on the U.K./ECtHR interaction solely through the lens of what happened to prisoners’ voting is to miss (again borrowing from Jack Straw) that « at the heart of it ... the issue is the constitutional relationship between this Parliament and Strasbourg[249] ». And, for those in the U.K. eager to limit their relationships with Europe, the ECtHR’s refusal to give a federalism discount for prisoner voting helped them to promote their claims of the need for more autonomy. On the particular issue of voting, Parliament had not, as of 2016, voted on any of the bills altering the blanket ban[250]. More generally, U.K. officials led efforts that resulted in the « Brighton Declaration », a 2012 statement that, while reconfirming commitments to the European Convention, admonished the ECtHR for its failure to respect the « sovereign equality of the States », to appreciate the importance of deference through subsidiarity, and to accord wide enough margins of appreciation[251].

That Declaration helped to produce what became Protocol No. 15 – the first amendment to the Preamble of the Convention on Human Rights and Fundamental Freedoms since its inception in 1949[252] and which, as noted at the outset, calls on the ECtHR to deploy the margin of appreciation. That 2013 Protocol, which is not effective until ratified by all 47 members[253], states that, under « the principle of subsidiarity » the High Contracting Parties « have the primary responsibility to secure the rights and freedoms defined in this Convention, and the Protocols thereto, and that in doing so they enjoy a margin of appreciation, subject to the supervisory jurisdiction of the European Court of Human Rights[254] ».

As in the context of the essentialist federalism stance reflected in debates about the U.S. Violence Against Women Act, aspects of Protocol No. 15 are symbolic. Subsidiarity in the European Union (as distinct from the Council of Europe) requires that the EU address « areas which do not fall » within its « exclusive competence », assessed by deciding if « the objections of the proposed action cannot be sufficiently achieved by the Member States, either at the central level or at the regional or local level, but can rather, by reason of scale or effects of the proposed action, be better achieved at Union level[255] ». As the description’s own circulatory suggests and as many commentators have explained, this injunction to European legislators does not weight a decision in favor of any particular level of governance but creates a presumption of action at the scale in which it is needed[256].

Yet the point of Brighton and of Protocol No. 15 is to link the EU’s doctrine of subsidiarity to the ECtHR’s work and to its doctrine of margins, so as to press the ECtHR to accede to a wider array of domestic decisions[257]. Moreover, in addition to limiting the ECtHR, critics in the U.K. argued for the repeal of its Human Rights Act of 1998 and for withdrawal from the jurisdiction of the ECtHR (a proposition that the U.K.’s own federated state made complex, given that the devolution of powers to Northern Ireland, Scotland, and Wales also required that they comply with Convention and other international human rights obligations of the U.K.)[258]. After the elections of 2014 in Scotland, pressures on the U.K. intensified. In 2016, the popular referendum on membership in the European Union resulted in Brexit; the legality of doing so by way of referendum as contrasted with a decision of Parliament is, as of this writing, before the Supreme Court of the U.K.[259] Brexit is an exit from the EU and not the Council of Europe, but popular pressure against « Europe » does not always draw distinctions among the distinct forms that « Europe » takes.

One could review these developments and argue that, by refusing to provide a discount, the ECtHR has paid a substantial price, in terms of its case law and its stature. One can thus see that federalism discounts are court discounts, which – when deployed – may try to protect judiciaries from conflicts with subunits. If in a self-protective mode, the ECtHR may have miscalibrated its own authority when it took up voting rights of prisoners, as well as given fodder to those within the U.K. who sought to distance the country from Europe. And, of course the tumultuous times in Europe are a multi-variable phenomenon, of which the case law of the ECtHR is but one factor.

Defenders of Hirst argue its importance as instantiating a democratic commitment that all persons in a population who are subject to sovereign power ought not to be precluded from exercising authority within it[260]. Hirst is also a landmark in prisoners’ rights law for an insistence on the personhood of prisoners. Understanding its impact requires considering more than the sequence of decisions detailed thus far, so as to look beyond the conflict between the European court and the subunit. As of 2005, the estimate was that about one million European prisoners were prohibited from voting[261]. Once Hirst was decided, Ireland responded by increasing efforts to enable already enfranchised prisoners to vote[262], and other countries amended legislation on prisoner voting[263]. Further, in a series of judgments, the ECtHR also concluded that other countries (including Russia[264] and Bulgaria[265]) were obliged to permit voting by prisoners.

These rulings are in turn part of criminal justice and prison condition reforms. In 2013, the ECtHR issued an important decision on the « right to hope »; Vinter v. U.K. held that the U.K.’s « whole life tariff », which appeared to admit virtually no opportunity for release, violated the Charter. Even if sentenced for gruesome crimes, prisoners have to be given genuine consideration for release at some point[266]. Once again, and even as the impact would have been limited (in March of 2014, 55 people had whole life sentences[267]), the U.K. resisted, albeit by arguing that its process for reconsideration on « compassionate grounds » does provide the requisite « hope » required by the Vinter ruling[268]. In 2015, a five-member panel of the ECtHR concurred[269].

Hirst and Vinter are part of a broader picture of prisoner rights and criminal justice reform, and thus a question is how the U.K./ECtHR exchanges have affected that social movement aiming to change the import of incarceration, the parameters of sentences, and the treatment of prisoners. According to one submission in Hirst, ending disenfranchisement was part of systemic efforts to revise the treatment of people serving long-term sentences. In 2003, Europe’s Committee of Ministers had called for prisoners to be dealt with in ways to maximize « individualisation », « normalisation », and « responsibility[270] » in the « management of life and long term prisoners[271] ».

That proposal grew out of the European Committee on Crime Problems (CDPC), which had recommended to the Council of Europe that changes were needed to counteract the « negative effects of long term imprisonment[272] ». The CDPC’s proposals included prisoner voting. The point was to underscore that, while prisoners lost their liberty, they did not lose « their citizenship[273] ». The idea was that if and when prisoners could reenter the community, they needed to understand themselves as valued participants. Voting was a means of getting prisoners to invest in the social order and thereby to make both prisoners and communities safer. The goal was to « approximate as closely as possible the realities of life in the community ». Doing so entailed keeping the « values of citizenship alive », which were in turn the « basic principles of democratic government[274] ».

Even as Europe became more punitive and sentencing rates rose[275], voting was and is a piece of a story about grappling with the harms of the current expansive structure of criminal justice[276]. The insistence on prisoner voting is a means of calling into question the panoply of deprivations imposed by incarceration. Rather than assuming the entailments of deprivations, the goal of « normalization » is to put each aspect of liberty-deprivations up for scrutiny. On this account, the ECtHR has been an important participant in insisting on reforms. And its activities have not been limited to voting but include a host of rulings, such as on prison conditions in Italy and access to visitors for those in solitary and more generally for prisoners as part of the right to family life[277].

 

IV. De-Essentializing Federalism(s)

 

Having detailed the limits of essentialist approaches and accounted for the potential of federalism discounts as a form of dialectical federalism leaving room for legal and political exchanges across the dizzying map within and beyond a federation, I close by underscoring the need to expand the attributes identified with this political configuration and to open up discussions of its forms. The three-part schematic provided by Andreas Auer[278] – identifying autonomy, participation, and superposition as the key elements of federalism – requires additions. My account has added another three: de-essentializing compentencies to underscore the changing import of rights; translocal-transnationalism to capture fluid, disaggregated reconfigurations of political authority across boundaries; and federalism discounts that provide for temporizing accommodations, to acknowledge that the discounts which federalism begets require revisiting regularly.

I also have used federalism(s) as an umbrella to underscore the dynamic features within federations and the variations among federated forms. The diversity of positions taken in the name of federalism, the multiple actors in federated systems and the range of interactions undermines efforts to impose neat grids to map the interactions. By pluralizing the form, I invite reflection on whether the jurisdictional essentialism and federalism discounts that I have explored in the context of the United States and Europe are prominent features of other federated systems, also in need of mediating mechanisms. Using federalism(s) as the category reflects the variety within the genre and thereby buffers against essentialist presumptions that federalism per se provides answers to questions about the allocation, the legitimacy, and the justice of power and the shape of individual rights.

 

Judith Resnik

Arthur Liman Professor of Law, Yale Law School. J. Resnik is the author of Representing Justice: Invention, Controversy, and Rights in City-States and Democratic Courtrooms (with Dennis E. Curtis, New Haven, Yale Univeristy Press, 2011); Migrations and Mobilities: Citizenship, Boarder, and Gender (co-editor Seyla Benhabib, New York, New York University Press, 2009); « Bordering by Law: The Migration of Crimes, Sovereignty, and the Mail », in J. Knight (dir.), Nomos LVII: Immigration, Emigration, and Migration, New York, New York University Press, 2017, p. 79-201, and « Diffusing Disputes: The Public in the Private of Arbitration, the Private in Courts, and the Erasure of Rights », Yale Law Journal 124, 2015.

 

Pour citer cet article :
Judith Resnik «Accommodations, Discounts, and Displacement: The Variability of Rights as a Norm of Federalism(s) », Jus Politicum, n° 17 [https://juspoliticum.com/article/Accommodations-Discounts-and-Displacement-The-Variability-of-Rights-as-a-Norm-of-Federalism-s-1141.html]