abstract. Proportionality, accepted as a general principle of constitutional law by many countries, requires that government intrusions on freedoms be justified, that greater intrusions have stronger justifications, and that punishments reflect the relative severity of the offense. Proportionality as a doctrine developed by courts, as in Canada, has provided a stable methodological framework, promoting structured, transparent decisions even about closely contested constitutional values. Other benefits of proportionality include its potential to bring constitutional law closer to constitutional justice, to provide a common discourse about rights for all branches of government, and to help identify the kinds of failures in democratic process warranting heightened judicial scrutiny. Earlier U.S. debates over “balancing” were not informed by recent comparative experience with structured proportionality doctrine and its benefits.

Many areas of U.S. constitutional law include some elements of what is elsewhere called proportionality analysis. I argue here for greater use of proportionality principles and doctrine; I also argue that proportionality review is not the answer to all constitutional rights questions. Free speech can benefit from categorical presumptions, but in their application and design proportionality may be relevant. The Fourth Amendment, which secures a “right” against “unreasonable searches and seizures,” is replete with categorical rules protecting police conduct from judicial review; more case-by-case analysis of the “unreasonableness” or disproportionality of police conduct would better protect rights and the rule of law. “Disparate impact” equality claims might be better addressed  through more proportionate review standards;  Eighth Amendment review of prison sentences would benefit from more use of proportionality principles. Recognizing proportionality’s advantages, and limits, would better enable U.S. constitutional law to at once protect rights and facilitate effective democratic self-governance.

author.Thurgood Marshall Professor of Constitutional Law, Harvard Law School. An earlier version of this paper was given as my Chair Lecture in October 2013 at Harvard Law School. With deep thanks to friends and colleagues for helpful comments on drafts and for discussions over the years of these topics, including Aharon Barak, David Cole, Richard Fallon, David Fontana, Nancy Gertner, Tom Ginsburg, Jamal Greene, Dick Howard, David Law, Jud Mathews, Martha Minow, Iddo Porat, Judith Resnik, Fred Schauer, Kim Lane Scheppele, Mike Seidman, Greg Shaffer, Joseph Singer, Gerry Spann, Carol Steiker, Nick Stephanopoulos, Geoff Stone, David Strauss, Laurence Tribe, Mark Tushnet, Barbara Underwood, and Katharine Young; with thanks for opportunities to present precursors to this paper in George Washington University Law School’s Comparative Constitutional Law Roundtable, at a Law and Society meeting in Boston, and at a faculty workshop at the University of Chicago Law School; and with great appreciation for excellent research assistance from Harvard Law students, including Tom Burnett, Abby Collela, Jess Eisen, Dylan Lino, Ezra Marcus, Karthik Reddy, Andres Salinas, and Jason Shaffer.

Introduction

“Proportionality” is today accepted as a general principle of law by constitutional courts and international tribunals around the world.1 “Proportionality review,” a structured form of doctrine, now flows across national lines, a seemingly common methodology for evaluating many constitutional and human rights claims.2 The United States is often viewed as an outlier in this transnational embrace of proportionality in constitutional law.3 Yet some areas of U.S. constitutional law embrace proportionality as a principle, as in Eighth Amendment case law,4 or contain other elements of the structured “proportionality review” widely used in foreign constitutional jurisprudence,5 including the inquiry into “narrow tailoring” or “less restrictive alternatives” found in U.S. strict scrutiny.6

Justice Stephen Breyer has suggested that there are other areas in which the appropriate standard of judicial review would involve examining the proportionality of government regulation.7 For example, in United States v. Alvarez,8 Justice Breyer’s concurrence, joined by Justice Kagan, associated proportionality review with intermediate scrutiny and applied this standard to evaluate a First Amendment challenge to the Stolen Valor Act.9 In his dissent in District of Columbia v. Heller,10 Justice Breyer explicitly invoked the idea of proportionality as a guide to permissible regulation under the Second Amendment.11 This explicit invocation of proportionality led some scholars to begin to consider, critically, the prospects of proportionality review, as it has developed elsewhere in the world, being more fully embraced in the United States.12

Given developments within and outside the United States, the time is ripe to take a fresh look at proportionality, both as a general principle in constitutional analysis and as a structured doctrine of potential benefit to discrete areas of U.S. constitutional law. In 1987, T. Alexander Aleinikoff criticized U.S. constitutional law for its overreliance on balancing in doctrines like strict scrutiny and in cases like Tennessee v. Garner13 or Mathews v. Eldridge,14 where the Court aimed to strike a balance among different interests.15 Other work soon followed, contrasting more categorical and rule-like approaches, on the one hand, and standards, on the other.16 The scholarship of the late 1980s may have influenced case law in some areas towards more categorical rules.17 But these earlier U.S. debates could not have been informed by the subsequent course of proportionality review in other countries. Foreign courts’ experience with proportionality review casts new light on these enduring questions in ways that suggest that U.S. constitutional law would benefit from a moderate increase in the use of proportionality.

Proportionality can be understood as a legal principle, as a goal of government, and as a particular structured approach to judicial review. As a principle and as a goal of constitutional government, proportionality is a “precept of justice,”18 embodying the idea that larger harms imposed by government should be justified by more weighty reasons and that more severe transgressions of the law be more harshly sanctioned than less severe ones.19 Proportionality as a principle is embodied in a number of current areas of U.S. constitutional law: for example, in Eighth Amendment “cruel and unusual punishments” and “excessive fines” case law; as a limit imposed by the Due Process Clause on the award of punitive damages; and in Takings Clause cases requiring “rough proportionality” between conditions on zoning variances and the benefits of the variance to the property owner. In each of these areas, the principle of proportionality imposes some limit on otherwise authorized government action, a limit connected to a sense of fairness to individuals or a desire to prevent government abuse of power. Proportionality is centrally concerned with how, in a “democratic society, . . . respect for the dignity of all men is central,”20 reflected in “our Nation’s [longstanding] belief in the ‘individuality and the dignity of the human being.’”21

Proportionality as a structured legal doctrine is used by some (not all) courts that treat proportionality as a general principle. In countries like Germany, Canada, and Israel, courts use a similar multi-part sequenced set of questions;22 elsewhere, such questions are considered but in a less sequenced way.23 In Canada, for example, structured proportionality review begins with attention to the scope of what a right is intended to protect; if a right has been infringed, the inquiry turns next to the authority for the action, and to the importance and legitimacy of the government purpose. If an infringement on interests protected by a right is shown, and if the challenged action has been “prescribed by law” sufficiently precisely and for a legitimate and sufficiently important purpose, then the constitutionality of the means used are examined through a three-fold inquiry into: (a) rationality; (b) minimal impairment; and (c) proportionality as such.24 Several of these criteria correspond with elements in U.S. “strict,” “intermediate,” or “rational basis” scrutiny: the need for a sufficiently important or “compelling” government purpose; the rational connection required between the means chosen and the end; and the “minimal impairment” inquiry into whether there are less restrictive means towards the same goal.

Structured proportionality analysis in countries like Canada, Germany, or Israel includes an additional stage—“proportionality as such”— asking whether the intrusion on the challenger’s rights can be justified by the benefits towards achieving the important public goal. This step calls for an independent judicial evaluation of whether the reasons offered by the government, relative to the limitation on rights, are sufficient to justify the intrusion. While this step is sometimes referred to as involving “balancing,” the “proportionality as such” question in structured proportionality doctrine differs from “balancing” tests that tend to focus primarily on quantification of net social good, as in Dennis v. United States25 or Mathews v. Eldridge.26

Take Canada as an example of structured, sequenced proportionality analysis. First, “proportionality as such” is a part of a doctrine that, as a whole, prioritizes the right, putting the burden of justification on the government.27 In this respect, structured proportionality analysis differs from “multi-factor” analyses of proportionality, as one sees in some countries, including South Africa,28 or from some U.S. “striking a balance” case law. Second, Canadian-style proportionality review is a logically sequenced set of inquiries that limits the need to consider whether the government interests justify the intrusion on interests protected by rights. It does so by first examining whether the challenged action is authorized by law, and then whether the government’s purpose is sufficiently important to serve as a basis for limiting the right at all. If these first tests are met, Canadian proportionality review examines the rationality and necessity of the means chosen, all before reaching the final “proportionality as such” inquiry. In this way, if the means chosen are not suitable or necessary to advance the government’s interest, the case can be resolved at one of these stages:the courts need not reach the “proportionality as such” question unless there is a genuine conflict between the government’s interest and the interests protected by the right.29 Third, “proportionality as such” returns courts to considering both the infringed-on right and the government’s purposes, not just in terms of their theoretical gravity, but in terms of the relative weight or bearing of the government’s reasons in relation to the harm to the challenger’s rights, in a particular context and in light of constitutional values. In this way, courts are not “substituting” their judgment for that of the legislature.30 They are playing a valuable judicial role—checking to assure appropriate attention to rights within a framework of constitutional justice.

Part I provides background for considering proportionality in the United States. It notes several areas of U.S. constitutional law in which proportionality already is an element of constitutional analysis and argues that one of the goals of the Constitution was to produce a just government, one likely to avoid arbitrariness and to act proportionately. As further background, Part I goes on to describe in more detail the structured form of proportionality review as it exists in several foreign countries, with special focus on Canada.

Part II explores why proportionality has not been used as a general principle of constitutional law in the United States. It suggests that the aversive impact of Lochner v. New York31 and Dennis v. United States,32 as “negative precedents,”33 led to a search for categorical approaches to constrain judicial discretion. Moreover, the age of the Constitution and related interpretive practices help account for the absence of any general embrace of proportionality. For example, the Constitution’s brevity and, relatedly, the relative dearth of rights that are viewed as in tension with each other, have tended to reinforce a view of rights either as trumps34 or as prohibited reasons for government action.35 These conceptions contrast with alternative understandings of rights as presumptive protections of human interests36 or as values to be optimized,37 which some leading theorists link with proportionality review. And, unlike European countries, which have incentives to harmonize national constitutional law with international rights regimes that rely on proportionality, the United States has not been comfortable treating its international human rights obligations as judicially enforceable domestically.38

Part III makes an affirmative case for greater use of proportionality as a principle and for structured proportionality as a standard of review in the United States. I begin by looking at discrete areas of U.S. constitutional law, starting with Fourth Amendment cases like Atwater v. City of Lago Vista,39 with rigid rules allowing police to detain and search regardless of the severity of the offense—rules that facilitate humiliating and badly intentioned police conduct. Excluding proportionality considerations neither fulfills the purpose of the Fourth Amendment nor promotes respect for the Constitution as law. Canadian case law on analogous rights offers an alternative approach. I then consider a recent First Amendment case, Holder v. Humanitarian Law Project,40 that appears to depart from existing categorical rules. Applying structured proportionality analysis in this case, I suggest, would require more disciplined attention both to free speech and national security interests, in order to clarify which considerations control.

Next, I discuss some general normative arguments in favor of structured proportionality review and proportionality principles. First, Canadian-style proportionality review promotes structured and transparent decisions through a stable methodological framework. Second, proportionality as a principle helps bring constitutional law closer to constitutional justice. Third, proportionality principles and structured proportionality review provide a better bridge between courts and other branches of government, offering criteria for constitutional behavior that are usable by, and open to input from, legislatures and executives. Fourth, proportionality analysis can reveal process failures, including departures from impartial governance, warranting heightened judicial scrutiny.

Part IV takes up several objections to proportionality review—that it is irrational, insufficiently protective of rights, unduly intrusive on legislatures, or overempowering of courts—and responds to each. I give special attention to the concern that proportionality review might focus too much attention on governmental justifications for its means and not enough on deontological understandings of rights. I suggest that more deontological understandings of rights, and attention to particular constitutional texts and lines of cases, is appropriate both in initially defining whether a right has been infringed and what ends are legitimate, and also in evaluating “proportionality as such.” Part IV also considers arguments from American exceptionalism that would preclude greater use of proportionality review. Exceptionalist claims, however, cannot be made or answered in broad brushstrokes; indeed, I argue, U.S. history and experience support greater use of proportionality.

Although some scholars view case-by-case application of proportionality analysis as almost always normatively superior to other approaches to rights adjudication,41 Part V takes a different view. Text, history, and precedent matter. Not all rights have the same structure nor serve the same purposes; free speech claims, which benefit from a presumptively categorical structure, are different from police behavior or criminal sentences, both of which would benefit from greater attention to proportionality. Even in adjudicating a single claim, different issues may call for different treatment. In equal protection law, paying more attention to disproportionate effects need not imply embrace of all elements of structured proportionality doctrine. Moreover, sometimes the most “proportionate” results will be achieved through categorical rules, especially when remedial frameworks are considered.42 At least some of these rule-like regimes can be justified in terms of proportionality analysis at the level of the rule. Being proportional about proportionality means recognizing that history and text have roles to play, and that proportionality as a principle is not always served by proportionality as a doctrine.

I. proportionality in u.s. constitutionalism and abroad

Proportionality as an element of constitutional doctrine has already been recognized in several areas of contemporary constitutional law in the United States. This is not surprising, since well-designed constitutions are generally intended to promote proportionate, non-arbitrary government behavior. What the United States does not presently use is the structured “proportionality doctrine” described in Part I.C.

A. Proportionality Principles Already Recognized in U.S. Constitutional Law

Americans are already familiar with the legal principle of proportionality in constitutional law. The Eighth Amendment’s case law has long recognized that punishments grossly disproportionate to the severity of the offense are prohibited as cruel and unusual punishment,43 although the Court’s willingness actually to scrutinize the proportionality of sentences has varied over time and contexts.44 The Excessive Fines Clause of the Eighth Amendment has also been understood to impose proportionality limits.45 Since the 1990s the Court has invoked proportionality in several other constitutional contexts. For example, under the Due Process Clause, courts must now ensure that the measure of punitive damages in civil cases “is both reasonable and proportionate to the amount of harm to the plaintiff and to the general damages recovered.”46 Under the Takings Clause, conditions for zoning permits must have “rough proportionality” to the effects of the proposed use of the property.47 Furthermore, the “undue burden” standard is now the controlling inquiry in the Court’s abortion cases, invoking in its language and application a concern for the reasonableness of regulations affecting women’s choices to abort their pregnancies prior to viability.48 All of these standards invoke proportionality in resolving individual rights questions, as do Justice Breyer’s First Amendment opinions.49 Moreover, the Court has extended proportionality standards to federalism issues: as of 1997, legislation under Section 5 of the Fourteenth Amendment must have “congruence and proportionality” to conduct that Section 1 prohibits.50

As these examples suggest, U.S. courts have found the concept of proportionality increasingly attractive in resolving interpretive challenges, prompting scholars to identify the roots of proportionality doctrines in U.S. constitutional law. Richard Fallon, for example, has drawn comparisons between European proportionality doctrine and U.S. strict scrutiny as it emerged in the 1960s (and applied thereafter),51 while Alec Stone Sweet and Jud Mathews see proportionality review in nineteenth century Dormant Commerce Clause cases.52 Attraction to proportionality in both the courts and the academy is no surprise, since an aspiration to proportionate government, as an important aspect of justice, is implicit in the constitutional design.

B. Proportionate Government as a Goal of Constitutional Design

The Constitution’s Preamble states that one of its goals is to “establish Justice,” echoing the defining commitments of leading state constitutional instruments of the time.53 “Justice” has, at least since the time of Aristotle, been associated with proportionality.54 Although the Preamble does not contain independently operative grants of power, it nonetheless provides important background for understanding constitutional purposes relevant to the interpretation of the operative provisions that follow.55 Similarly, there are allusions to proportionality in the Federalist Papers, where the constitutional design is described more generally as aimed to produce “a wise and well-balanced government for a free people”56 in a way that will help control “abuses”57 and avoid the exercise of “arbitrary and vexatious powers.”58

These sorts of commitments to government that is just, and to proportionality in the government’s treatment of citizens, have deep roots in antecedents to the U.S. Constitution, including the Magna Carta. The Magna Carta’s articles on “Amercements” plainly expressed a demand for proportionality in the imposition of fines;59 other provisions of the Magna Carta called for “justice” to be provided through the law courts.60 As Dick Howard has shown, the Magna Carta’s influence was felt in the American colonists’ demands for recognition of their rights as English citizens in accordance with colonial charters;61 the influence of the Magna Carta and the English Bill of Rights62 is further reflected in founding period state constitutions, in requirements that no “cruel and unusual” punishments nor “excessive fines” be imposed,63 as well as in the right to open courts.64 Both Massachusetts’s and Virginia’s post-revolutionary constitutions emphasized “justice” and “moderation” as among the first virtues of the governments they sought to establish.65 Similar requirements are evident in most modern constitutions in constitutional democracies, and even when not explicit, the goal of proportionality is implicit in any constitution that aims to produce justice by limiting as well as empowering government.66

Proportionality bears a special relationship to government in a constitutional democracy. For an essential idea of constitutional democracy is that in confrontations between citizens and government, government is restrained and avoids oppressive and arbitrary action.67 The means to achieve this goal are varied, but requiring proportionality of action is one way in which the idea of limited governnment can be realized.68 Second, constitutional democracies’ legitimacy is based on accountability to the people, including but not limited to majoritarian consent. Elections provide one source of accountability, but ensuring that government has justified reasons for action (whether legislative or executive) helps promote accountability on an onging basis.69 Third, constitutional democracies are not only limited governments; they are goverments limited by a commitment to fundamental human equality. It is on that commitment to the normative equality of all members of the polity that democratic self-governance rests.70 Recognizing each person as endowed with a quality of humanity equally deserving of respect arguably calls for reasoned justification for the imposition of special burdens or intrusions.71

Recognizing proportionality as a goal of constitutional government does not necessarily imply that judicial review is the best method for achieving proportionate decision making. For example, McCulloch v. Maryland concluded that the principal protection against abusive taxation is the link between representation and the taxed constituency.72 Legislators and executive actors may be understood ordinarily to have obligations to act proportionately, even if those obligations are not justiciable. What, then, is the role of judges in implementing the constitutional value of proportionality? I consider this question first outside the United States and then within it.

C. Proportionality Elsewhere: The United States in Comparative Perspective

Having suggested that the principle of proportionality is part of the U.S. Constitution, I turn now to proportionality as a structured doctrine developed in the post-World War II period in Germany, Canada, Israel, and elsewhere.73 Although there are differences in doctrinal terms and applications among different courts,74 for purposes of comparison to U.S. approaches, I focus primarily on Canada,75 drawing from other jurisdictions to illustrate particular points. In 1982, after a long public process, Canada adopted as part of its constitution the Charter of Rights and Freedoms, which in Section One guaranteed the rights set forth therein “subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.”76 This provision may be referred to as a “limitations” clause because it recognizes that rights may be limited by strong enough reasons, or as a “savings” clause, because statutes otherwise infringing on rights may be preserved from invalidation by meeting the standards of Section One. Canadian doctrine has developed a proportionality test to determine whether this standard is met.77 Limitations clauses in other countries have also been understood to invite courts to review the justifications for government action through proportionality analysis.78

In Canada, when government action is challenged as violating a Charter right, the challenger bears the burden of showing a rights infringement, and Canadian judges first inquire into the scope of the interests that the right protects. In so doing, the court typically adopts a generous view of the scope of what is protected by the right.79 The court then considers whether the government has shown that it is acting under clear legal authority80 and for reasons that are “pressing and substantial in a free and democratic society”;81 if not, inquiry is at an end.82 If the infringement of right is pursuant to government action authorized by law and has a “pressing and substantial” purpose, the Court then considers whether the government has shown that the challenged action is “demonstrably justified.”83

At this justificatory stage, the courts employ a three-part inquiry, focusing on the means used to advance the government’s purpose and asking whether (1) the means chosen are rationally related to the legitimate object; (2) the means chosen “minimally impair” protected rights; and (3) the benefits towards achieving the government’s objective are sufficient to warrant the harm to interests protected by rights (a step called “proportionality as such”).84 The rationality step is similar to U.S. rational basis review.85 Although this element is normally found to be satisfied, in Oakes the Canadian Supreme Court concluded that a rebuttable presumption that one who possessed any amount of a drug was also trafficking in the drug was not rational.86

If the statute is found (as most are) to be a “rational” means of advancing the government’s purpose, the courts go on to consider whether it impairs the right “as little as reasonably possible in order to achieve the legislative objective.”87 This minimal impairment step has sometimes been described as a cognate test to the U.S. “least restrictive alternative” requirement in strict scrutiny; this second step is sometimes described in scholarly literature as a “necessity” test.88 However, the minimal impairment test does not necessarily imply that if any less restrictive approach can be imagined, the law is invalid;89 the government “is not required to pursue the least drastic means of achieving its objective,” so long as it “adopt[s] a measure that falls within a range of reasonable alternatives.”90 The Canadian courts will look to see whether there is an obvious and workable alternative, sometimes drawing on approaches already in use by governments, as in a recent case involving procedures for secret evidence in immigration proceedings.91 Chief Justice McLachlin has emphasized that the “important point” is whether proposed alternative (and less rights-impairing) means would be “less effective” in advancing the government’s goal.92

In cases involving more polycentric interests, “minimal impairment” scrutiny can allow considerable latitude to legislative choices. In Edwards Books,93 the Canadian Supreme Court upheld an Ontario statute establishing Sunday as a common day of recreation in which most retail businesses had to close. The statute had an exception for employers who closed for Sabbath on Saturday and had fewer than seven employees working on Sunday,94 but several Ontario retailers, including some owned by observant Jews, challenged the statute. They argued that the different approach taken in New Brunswick was less impairing of religious freedom rights; New Brunswick provided an exemption for any retailer with a sincere religious belief that it needed to close on a day other than Sunday.95 The Court was not persuaded that New Brunswick’s approach was less impairing: New Brunswick made the exemption available regardless of the number of employees, but Ontario did not require the employer to claim a sincere religious belief. So a small shopkeeper employing observant Jews could benefit from the exemption regardless of the employer’s beliefs.96 Likewise, the Court found, another proposed alternative—allowing an exemption to be invoked by individual employees—was not necessarily more minimally impairing, because of subtle social pressures on employees not to assert such claims.97 Given the complexity of the rights-holders’ interests—as owners, employees, and consumers—the Court could not conclude that one approach was less impairing of rights than another; the infringement on religious freedoms was found not disproportionate to the legislature’s objective; and so Ontario’s law stood.98

The last stage of analysis is sometimes called “proportionality as such.”99 In this phase, the court asks whether the government’s reasons for regulating and the degree to which they are likely to be served can justify the harm to constitutionally protected interests. By going beyond rationality and minimal impairment, the “proportionality as such” test can make the doctrine more rigorous than U.S. strict scrutiny, which ends after the “least restrictive means” test. In Oakes, Chief Justice Dickson explained that:

Some limits on [Charter] rights and freedoms . . . will be more serious than others in terms of the nature of the right or freedom violated, the extent of the violation, and the degree to which the measures which impose the limit trench upon the integral principles of a free and democratic society. Even if an objective is of sufficient importance, and the first two elements of the proportionality test are satisfied, it is still possible that, because of the severity of the deleterious effects of a measure on individuals or groups, the measure will not be justified by the purposes it is intended to serve. The more severe the deleterious effects of a measure, the more important the objective must be if the measure is to be reasonable and demonstrably justified in a free and democratic society.100

Minimal impairment analysis is defined by the scope of the government’s objective; only proportionality as such “takes full account of the ‘severity of the deleterious effects of a measure on individuals or groups.’”101

Canadian cases rarely turn on this third step, generally finding laws unconstitutional on minimal impairment grounds.102 Other jurisdictions, however, sometimes find that a statute that passes minimal impairment nonetheless fails “proportionality as such.” In Germany, for example, “proportionality as such” has been used more often than in Canada.103

While “proportionality review” requires an initial determination of whether the government’s purpose is sufficiently important to warrant restricting rights at all, in the final stage the relative strength of that interest is evaluated in relation to the specific harm to rights;104 the greater the intrusion on rights, the greater must be the need and justification for the challenged measure. Consider an example from Israel, whose case law sometimes adopts a particularly rigorous form of analysis of this last prong. In the Beit Sourik case,105 the Israeli High Court of Justice found that the government had a legitimate purpose in building a fence to protect Israelis from violent attacks from occupied territory.106 The Court found that the government’s choice for the fence location, near the top of a mountain, was a rational step towards the goals of surveillance and protecting security forces and travelers on a nearby highway.107 The line drawn was also minimally impairing of the rights of Palestinians fenced off from their lands because no other route could achieve an equivalent level of security.108 The court explained that a “less restrictive means” referred only to an alternative that equally advances the law’s purpose while intruding less on rights.109 However, the Court held, the fence had to be moved to a less elevated location, allowing Palestinians more access to their lands, because the initial location failed the final, “proportionality as such” test: the marginal improvement to security—and protection of the life of the Israeli civilian—from the line that the military chose, as compared to a line in a lower location, was, in the Court’s view, far less than the marginally greater intrusions on Palestinian humanitarian rights.110

Not surprisingly, the U.S. case law on “less restrictive means” sometimes obscures the distinction between “less restrictive means” that are as effective and those that are not, in part because of the absence of any separate analysis of “proportionality as such.” Differing formulations can elide whether a “less restrictive means” must be one that achieves equivalent progress towards the government’s legitimate goal.111 Indeed, U.S. courts referring to “least restrictive alternatives” tend not to specify whether this analysis requires that the measures being compared “equally advance” the compelling government interest. In United States v. Alvarez (the Stolen Valor Act case), the plurality accepted that an online database against which false claims could be checked was less restrictive than a criminal prohibition on lying about receiving the Medal of Honor.112 The analysis left unclear whether the plurality had concluded that a database would be equally effective in carrying out the government’s legitimately relevant interests, or instead that even if the database were less effective, the database would be a sufficient alternative given the relatively greater importance of free speech concerns.113 Similarly, in McCullen v. Coakley,114 the U.S. Court, in concluding that a thirty-five-foot buffer zone was not sufficiently tailored to achieve the government’s legitimate goal of maintaining public safety and preserving access to abortion clinics, left unclear whether there were equally effective alternatives or whether the marginal additional benefit towards the government’s goal under the statute, as compared to alternatives, was unjustified in light of the degree of intrusion on rights.115 By contrast, the relative importance of the rights and values at stake can be distinctly evaluated in structured proportionality analysis at the “proportionality as such” stage.

A striking feature of Canadian jurisprudence has been the stability of the proportionality doctrine and its utility as a method for a structured decisional analysis in which the Justices generally focus on the same questions in the same order.116 (As we shall see, Canadian concerns for proportionality are found not only in formal Section 1 analyses but also in definitions of the scope of certain rights.117) Although the three doctrinal components of proportionality review of means are similarly framed in most jurisdictions that use the doctrine, these elements may be applied somewhat differently by different courts or judges.118 Nonetheless, proportionality doctrine has shown itself capable of providing a stable framework across many controversial issues, in jurisdictions widely recognized as free and democratic constitutional states.

II. of older texts, clause-bound interpretation, and negative precedents

Despite proportionality’s appeal in other countries and its partial presence in some areas of U.S. constitutional law,119 the Supreme Court treats proportionality in different constitutional arenas as unconnected. Multiple accounts of the relative absence of proportionality from U.S. constitutional law have been offered.120 As later Parts will argue, this relative absence does not mean that the current situation must remain as it is, nor are the historic reasons for its relative absence reasons against expanding its use today.121 In this Part, I try to account for why proportionality as a general principle or doctrine has not emerged in the United States.

There are many factors contributing to the relative dearth of proportionality analysis in U.S. jurisprudence, among them a general propensity for what John Hart Ely critically referred to as “clause-bound” interpretation.122 Unlike some European courts, U.S. constitutional case law has for the most part not aspired to general theoretical connections linking constitutional doctrines in one area to those in another.123 Moreover, from a comparative perspective, scholars have observed that balancing or proportionality in Germany is associated with rights protection in a frame of constitutional perfectionism, while in the United States balancing is associated with pragmatic ad hocery and limitations on rights.124

Several additional reasons relating to the age of the U.S. Constitution also help account for why proportionality has not emerged as an articulated general constitutional principle or doctrine. The Constitution’s age affects both the timing of case law development and the contents of constitutional text. Unlike the Canadian Charter of Rights (1982) or the German Basic Law (1949), many of the Constitution’s rights provisions date to the late eighteenth and mid-nineteenth centuries. They regularly became the subject of the Court’s interpretation in the late nineteenth and early twentieth centuries. An evolving body of U.S. case law had already developed well before the atrocities of World War II and the subsequent explosion of international human rights law. By contrast, in Germany (after 1949) and Canada (after 1982), the highest courts were faced with new rights-protecting instruments, framed by international commitments to human rights, which provided an occasion for affording some degree of coherent interpretation to new constitutional instruments.125 The U.S. Constitution, moreover, has no general limitations clause, unlike many modern constitutions.126 Such limitations clauses can provide a textual basis for a general doctrine of how to justify the infringement of rights, though they are not necessarily the foundation for courts doing so.127

As an older constitution, moreover, the U.S. Constitution (as conventionally understood) contains fewer rights and thus gives rise to fewer occasions for conflicts between constitutional principles than many newer constitutions. This is especially true for modern constitutions that enforce both older liberal rights and newer positive rights.128 Where constitutional rights are many and are viewed as “principles” requiring optimization, as in Germany, approaches that seek to give each principle its proportional due are likely to be of great appeal.129 In the United States, conflicts between constitutional values—like free press versus fair trial—exist but are perceived to arise less often. This in part reflects the relative terseness of the Constitution and its failure to include positive rights as such. But it also reflects the predominantly negative contemporary view of those rights that do exist.130 The Court has resisted arguments that would impose positive obligations on the government to enable the realization of rights, except in limited categories, such as the rights to counsel and to appeal in criminal cases. There are accordingly fewer perceived conflicts in rights and thus less felt need to find ways of reconciling such conflicts.131 The absence of positive obligations also affects other aspects of U.S. doctrine, in ways that call for caution in considering methodological shifts that are more than incremental in character.132

There are other contributing factors apart from the text and age of the U.S. Constitution. Unlike countries in Europe, the United States is not nested in a tightly woven supranational structure of economic union, nor deeply embedded in an effective regional human rights convention, enforced by a transnational court. Courts in Europe have incentives to draw on, and to anticipate, rulings of the two European courts, each of which relies on forms of proportionality review. Unlike Canada, the United States is not part of the Commonwealth, which has arguably promoted more sharing of jurisprudences across national lines. U.S. courts thus have not experienced to the same degree the flow of cases from national to supranational courts that is common in Europe, nor the regular interchange that occurs among judges of the Commonwealth nations.133 Its relative isolation from these influences, or those of international tribunals, is reflected both in the hesitation of the political branches to ratify human rights conventions,134 and in the Supreme Court’s recent case law.135

Over time, moreover, the U.S. Supreme Court has developed distinctive discourses around rights. U.S. law does not generally discuss rights as being subject to external limitation; when U.S. jurists, lawyers, or scholars say a “right” has been “infringed,” this is typically the end of analysis.136 In Canada, the scope of interests that the right protects is determined first from the perspective of the rights-holder; if the “right” is infringed, analysis does not end, but instead the government’s reasons for limitation are then separately considered. Likewise, in Germany, according to a leading scholar of proportionality review, rights that are “principles” are understood to be “optimization” requirements which must be protected to the maximum extent possible but which may be limited if there are strong enough reasons for the government to do so.137 In the United States, courts often blend the two ideas—which personal interests a right protects, and how the government may legitimately act to limit freedom—and articulate a “right” only after internally accounting for limitations deemed warranted by the government interests.138

At the same time, there are distinctively American fears about judging and the role of judges, in part an inheritance of legal realism and critical legal studies (CLS). This kind of skepticism about law, judging, and judges contrasts with German (and European) forms of optimism about the possibility of law as a practice distinct from politics.139 If legal realism and CLS contribute to a general skepticism about the capacity of law to constrain, then fears of judging were also reinforced by what we might call the ghosts of Lochner140 and Dennis,141 two cases that have come to be viewed as “negative precedents,” or cautionary notes of what not to repeat.142

As Richard Fallon has argued, Carolene Products laid the foundation for the Court to develop bifurcated categories of review, including more deferential review of economic regulation and heightened review of laws adversely affecting discrete and insular minorities, the representative process, or the protections of the first eight amendments.143 The vices of Lochner are debated,144 but Carolene Products, and the ensuing bifurcation of standards of review into rational basis and strict scrutiny, responded to two major critiques of Lochner by creating a clear hierarchy of rights: it rejected liberty of contract as an object of heightened attention and seemed to limit judicial intrusion on political choices, confining judicial discretion by “committing” the Court to two discrete standards of review,145 each of which was close to outcome determinative—strict scrutiny almost always fatal; rational basis rarely so.146

Yet over time, the persuasive, predictive, and constraining force of this bifurcation diminished. The concept of a rigid division in standards of review was implicitly challenged in Justice Thurgood Marshall’s 1970 dissent in Dandridge v. Williams,147 which argued that defining the level of benefits for children in poor families was not the kind of economic regulation of commercial enterprises on which the Carolene Products distinction rested.148 Questioning of the rigid tiers of review has extended to more recent debates about whether sexual orientation is a suspect or quasi-suspect category.149 With the addition of intermediate scrutiny,150 as well as hard-to-account-for variations in the application of the various tiers of review,151 the predictability of these categories has been somewhat diminished.152 Recent years have also seen some resurgence of enhanced constitutional protection for economic rights, such as in takings jurisprudence153 and commercial speech cases.154

If a reaction to Lochnerism helps explain the initial development of the two-tiered structure of review signaled in Carolene Products, the perceived failure of balancing to provide appropriate protection to First Amendment interests in Dennis155 may have contributed to the development of more categorical approaches to restrictions on speech inciting violence, as in Brandenburg v. Ohio.156 Such developments in turn have contributed to the notion that U.S. constitutional law more generally rests or should rest on categorical rules.157 Concerns for proportionate government action may, however, have informed the development of Brandenburg’s categorical rule.158 Exceptions to free speech rules in recent years, including Holder v. Humanitarian Law Project,159 have created a more complex and less determinate overall structure.160 Other categorical constitutional rules adopted, for example, in criminal procedure, including Fourth Amendment law, have been followed by arguably even more complex exceptions.161

As even our more categorical constitutional rules have become increasingly uncertain and complex,162 is this a time for some reorientation of U.S. law towards proportionality?

III. benefits of proportionality review for u.s. constitutional law

In this Part, I argue that in at least two areas of constitutional law, greater reliance on proportionality would beneficially enhance the protection of individual rights. Working from the facts to the law in the common law tradition, Part A considers recent Fourth Amendment case law in which the Court rejected arguments that arrests, or searches related to pretrial detention, should be limited by proportionality principles, and it contrasts such decisions with Canadian case law. Part B explores how the absence of a “proportionality as such” inquiry diminishes the force of U.S. rules against content-based regulation under strict scrutiny, using Humanitarian Law Project as an example. Finally, Part C advances some more general, theoretical arguments for increased use of structured proportionality review and proportionality as a principle in constitutional adjudication.

A. Regulating Police Behavior Under Constitutional Norms

1. Atwater v. City of Lago Vista and Fourth Amendment Case Law

In Atwater v. City of Lago Vista,163 the Court found no Fourth Amendment violation in the arrest of a motorist for a non-jailable traffic offense.164 Atwater was driving her two young children in their neighborhood when she was stopped by a police officer for not wearing her seatbelt and not having her children in seatbelts.165 Arresting Atwater, the officer denied her request to ask a neighbor to care for the children, indicating that he would bring them to the police station.166 Atwater’s hands were cuffed behind her back; she was placed in the back of the police car—without a seatbelt—and driven to the station.167 She was released about an hour later, paid a $50 fine for the seat belt offense, and discovered her car had been towed.168 She sued for damages, including distress-related medical costs for herself and one child.169

The Court described the police officer’s conduct in arresting the motorist as involving “merely gratuitous humiliations” and inflicting “pointless indignity and confinement.”170 Indeed, the Court wrote, her claim “clearly outweighs anything the City can raise against it specific to her case.”171 Acknowledging that “[i]f we were to derive a rule exclusively to address the uncontested facts of this case, Atwater might well prevail,” the Court noted that Atwater was an “established resident of Lago Vista with no place to hide and no incentive to flee, and common sense says she would almost certainly have buckled up as a condition of driving off with a citation.”172

Yet the Court rejected Atwater’s Fourth Amendment challenge: history suggested and functional concerns required that police officers be treated as having lawful discretion to arrest for any offense with probable cause.173 To hold otherwise, according to the Court, would impose unwarranted burdens on police officers of knowing details of criminal codes and anticipating likely charging decisions, thereby creating incentives to under-enforce criminal law by officers making split-second decisions.174 (A similar structure of analysis is found in Florence v. Board of Chosen Freeholders,175 involving visual strip and cavity searches at pretrial detention facilities.176)

If a case like Atwater had arisen in Canada, the first question the Canadian Court would likely have addressed is whether the plaintiff had interests protected by the provisions analogous to the Fourth Amendment.177 The first question in fact addressed by the Atwater Court was the scope of common law authority to make an arrest.178 Had the U.S. Court followed the structured proportionality review approach, it would have considered whether Atwater’s interests were within the scope of interests protected by the Fourth Amendment before going on to consider whether the search or seizure was justified, that is, reasonable. The amendment’s text plainly suggests that searches and seizures must be reasonable. It provides: “The right of the people to be secure in their persons . . . against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause . . . .”179 The U.S. Supreme Court did not conduct its analysis in this order. Moreover, it did not address potential harm to Atwater’s children.180

Whether “the people” can feel “secure in their persons” knowing that any traffic infraction can result in their being jailed deserves more attention. Justifications that sound only in authority, based on common law practice, are not so persuasive to the modern ear; and proportionality tests do not stop with the question of authorization.181 The Court’s methodology, which defined the rights at stake only in relation to an ambiguous common law history and its analysis of the government’s interests, left an essential aspect of the question under-explored.

The Atwater Court did engage in some balancing or weighing of government needs in deciding between a case-by-case or rule-based approach, and it chose a categorical rule. The Court treated police officers as needing prophylactic protection, reasoning that “a responsible Fourth Amendment balance is not well served by standards requiring sensitive, case-by-case determinations of government need, lest every discretionary judgment in the field be converted into an occasion for constitutional review.”182 The Court made empirical judgments—concerning the supposed dearth of abusive arrests and the need to avoid “a systematic disincentive to arrest”—in order to strike “a responsible . . . balance” through its categorical rule.183

As the dissenters argued, qualified immunity doctrine already protects officials from monetary liability under unclear legal standards.184 Some focus on the proportionality of the officer’s conduct, examining the reasons for this conduct, would have little potential for interference with law enforcement and would better protect citizens’ rights to be secure in their persons.185 Yet the Court offered little discussion of the scope of the interests protected by Fourth Amendment rights or of why the police officer did not use less restrictive alternatives reasonably available to him; its suggestion that the political process could control abuses,186 and its reference to a possibly different approach in “extraordinary” circumstances,187 left the decision only partially justified and partially transparent.

2. A Canadian Comparison

For comparison, let’s turn briefly to a recent Canadian decision188 concerning the Canadian Charter’s constitutional protections of the “right to be secure against unreasonable search or seizure”189 and “the right not to be arbitrarily detained or imprisoned.”190 In Aucoin, a Canadian police officer made a traffic stop because of a license plate irregularity; on questioning the nineteen-year-old driver, the officer found that he had consumed alcohol in violation of traffic laws prohibiting new drivers from drinking.191 Having decided to give the driver a ticket, the officer also decided to place the driver in the back of the police car while he wrote up the citation.192 For safety reasons, the officer conducted a pat-down search before putting the driver in the back of the patrol car and during that search discovered illegal drugs.193 The parties and the Court agreed that the initial detention of the driver in the traffic stop was lawful. The question was whether the decision to put the driver in the back seat of the patrol car was a reasonable exercise of the authority to detain.194

In the Canadian Court’s words, the issue was not whether there was authority to detain, but whether the officer was justified in exercising the authority as he did.195It was the “shift in the nature and extent of . . . detention” for “two relatively minor motor vehicle infractions” that created the constitutional violation.196 Placing the driver in the back seat of the police car, especially with the accompanying pat-down, “increased restrictions on the appellant’s liberty interests . . . [and] altered the nature and extent of the appellant’s detention in a fairly dramatic way—especially when one considers that the infractions for which he was being detained consisted of two relatively minor motor vehicle infractions.”197 Given the minor character of the offense, the decision to detain in the car did not meet the test of being “reasonably necessary” under all the circumstances, and so the detention and accompanying pat-down were not constitutional.198 The Canadian Court was unanimous in this holding.199

Canadian law thus adopts an alternative approach, insisting on a more case-by-case approach to examining whether a police authority has been exercised in a reasonable and proportionate way.200 A comparison with Atwater suggests that some form of more individualized proportionality analysis may produce decisions that are both better reasoned andmore protective of rights than the “categorical approach” employed by the U.S. Court.201

B. “Strict Scrutiny” and the First Amendment

The First Amendment is an area in which U.S. law is typically described as being based on presumptive or definitional categories.202 Would U.S. First Amendment law be improved by more attention to proportionality? If, for example, in applying the categorical presumption against content-based regulation, courts used as an additional test the question of “proportionality as such” from structured proportionality doctrine? Or if, in defining exemptions from the categorical presumption against content-based regulation, more attention were given to the principle of proportionality? To begin to answer these questions, consider first the Court’s recent decision in Holder v. Humanitarian Law Project.203

The case involved a challenge to a criminal statute prohibiting material support to designated terrorist groups. The challenge was brought by U.S.-based NGOs that sought, inter alia, to provide training to certain designated terrorist groups (such as the Kurdistan Workers’ Party (PKK) in Turkey) about how to invoke international law processes to advance their claims. Concluding that the statute involved a content-based regulation of speech, the Court nonetheless upheld the statute in light of the government’s interest in combatting terrorism.204

The protective power of the categorical approach is called into question by this decision. The Court in Humanitarian Law failed even to mention an arguably controlling decision from 1969, Brandenburg v. Ohio,205 which had held that speech believed to incite violence could be banned only when the speech’s character was an incitement to imminent action and likely to cause imminent lawlessness. Under Brandenburg, it would have been difficult to uphold the material support statute as applied to speech designed only to promote lawful invocations of international procedures, as the speech had neither the purpose of inciting nor a likelihood of causing imminent lawlessness.206

As noted, the Humanitarian Law Court concluded that the statute regulated speech based on its content; it therefore subjected the statute to strict scrutiny, rejecting the government’s argument for intermediate scrutiny.207 The Court indicated that the correct standard to apply was “the more rigorous scrutiny” found in such cases as Cohen v. California,208 Texas v. Johnson,209 and R.A.V. v. City of St. Paul.210 Although the Court was less than clear on precisely what that standard was, it appeared to be “strict scrutiny.”211 The parties all agreed that combatting terrorism was a compelling government interest. The Court emphasized that the prohibition was narrow, insofar as it did not prohibit “independent” advocacy, and applied only to “knowing” support.212 As applied to teaching terrorist groups how to petition international agencies, the Court concluded, the ban was sufficiently connected to combatting terrorism for three reasons: to prevent the freeing up of “fungible” resources that could be directed to unlawful acts; to obstruct terrorist groups from acquiring “legitimacy”;213 and to avoid difficulties in relationships with allies in the fight against terrorism.214 In responding to plaintiffs’ argument that there was no need to prohibit their nonviolent education and training activities, the Court further accepted Congress’s finding that “any contribution” to a terrorist organization facilitates its terrorist conduct—a finding the Court found was “justified” in an area where concrete information was often unavailable but serious risks were real.215 This deference to the government raises echoes of Dennis and casts further doubt on the constraining character of the “categorical” approach to free speech.

There is much to debate about the Court’s analysis in Humanitarian Law; I focus here only on two methodological points relating to structured proportionality analysis.

First, it is possible to understand the Court as saying that the statute was sufficiently narrowly tailored to the government’s compelling interest in combatting terrorism.216 It is not clear, however, how seriously the Court took the idea of narrow tailoring (which is analogous to the minimal impairment step); it did not, for example, explain how the “contribution” of training in international law could be “fungible” with support for terrorist activities, in the way other forms of contribution (such as money) could be. It arguably applied a less stringent means-ends test of whether the prohibition could be said rationally to serve the government’s asserted interests at all.217 What the Court may really have been conveying was the overriding importance of the government interest relative to the free speech interests affected by the specific statutory prohibition. Had the Court followed a more structured analysis218 it would be easier to understand whether the Court was modifying (or abandoning) narrow tailoring as a requirement in some class of national security cases. Second, addressing all of these elements might not only clarify the doctrine but also better protect free speech, which is always under particular stress during times of war or perceived security threats. Governments that will be held accountable for failures of security may in good faith believe that broad prohibitions on “support” are needed to provide the greatest assurance against future terrorism, without necessarily considering whether any marginal gain in security by prohibiting peaceful speech, in the form of teaching foreign groups about international law, justifies the harm to free speech values. “Least restrictive alternative” analysis might be understood to accept the government’s goals (assuming they are “compelling”) without evaluating their relative force vis-à-vis intrusion on rights. The added question of “proportionality as such” enables a court, even as it defers to government expertise on the nature of security risks, to exercise independent judgment on whether the risk reductions justify the harm to free speech rights.219 Because U.S. courts do not use structured proportionality doctrine in their constitutional jurisprudence, they may not even consider the appropriate relationship of government goals to free speech rights, captured by “proportionality as such,” or may do so sub silentio, to the detriment of both rights protection and the transparent and consistent development of constitutional law.

Consider, again, United States v. Alvarez,220the Stolen Valor Act case. A separate evaluation of the “less restrictive means” and “proportionality as such” tests might have clarified the decision. Although both the plurality and Justice Breyer asserted that the criminal statute could not be upheld because the government’s interest in protecting the integrity of military medals could be advanced by other means, the plurality opinion, at least, was unclear about whether other mechanisms were or needed to be viewed as equally effective.221 Arguably, both the plurality opinion and Justice Breyer’s combined the “less restrictive means” test with a sub silentio evaluation of “proportionality as such.”222 Important as the integrity of military honors may be, it may not have warranted an ex ante effort to suppress even false speech through a broad criminal sanction, if the goal of protecting military honors could have been served through less restrictive measures—even if those alternative measures were not quite as effective as a criminal sanction in deterring false claims. If this captures what the Justices in the majority were thinking, then “proportionality as such” might have better explained what motivated the decision.223 Even if the outcome were not changed through the adoption of a structured proportionality approach, the Justices’ reasoning would have been clarified.

C. Theoretical Benefits of Proportionality Review in Deciding Rights Claims

This Part will now identify at a more general level several benefits to be derived from judges applying proportionality doctrine or principles in evaluating rights claims. First, experience elsewhere suggests that structured proportionality review provides a stable framework for persuasive reason-giving, thereby enhancing the transparency of decisions, unlike more opaque forms of balancing.224 Second, proportionality analysis helps to bridge the roles of courts and legislatures. It requires legal authorization for infringement of rights; it also identifies criteria—to which legislatures are competent to speak—that form part of the justificatory process. Third, reliance on proportionality principles can help bring law closer to the community’s sense of justice, in part by cultivating the art of judgment by judges and lawyers. Fourth, attention to proportionality can help identify, and respond to, process deficiencies in governance.

1. Structured and Transparent Reason-Giving with Broad Justificatory Appeal

Proportionality analysis in Canada and some other jurisdictions provides a structured and transparent mode of reason-giving that produces justifications likely to be meaningful, or at least understandable, to the parties and other audiences for constitutional courts’ decisions. The sequencing and defined order of proportionality review of constitutional rights claims in Canada has provided a more or less stable doctrinal framework within which disagreements are conducted.225 It also contributes to the relative accessibility and transparency of the Court’s reasoning. The stability of the methodology, and its widespread acceptance, enables the Canadian justices’ disagreements to focus on matters that are understandable by the parties as substantively relevant to the contested issue; such opinions also make accessible to readers the nature of the justices’ disagreement, and the divergent evaluations they may give to the same factors.226 The sequencing of analysis may be contrasted with more “free form” evaluations in well-known U.S. balancing cases.227 In the United States, however, different Justices may well continue to deploy different methodologies, and so some of the structured transparency and consistency gains of a Canadian-style approach might not be realized.

In addition to its benefits in structuring and making more transparent the reasoning of the different justices, proportionality review—by embracing a wider range of reasons than those that resort to text, precedent, and history alone—may increase the persuasive value of the decisions to both the parties228 and the broader public. As Cass Sunstein has written, “[i]n American constitutional law, government must always have a reason for what it does.”229 Frank Michelman’s work emphasizes the connection between government reason-giving and equality of persons.230 Authority to act is not the same as a reason to act; authority alone does not meet demands for reasons. Furthermore, varying reasons may appeal to different audiences.231 Even in the formulation of categorical rules, as in Atwater, the Court typically invokes at least some consequentialist understanding—there, of the need to allow unimpeded law enforcement. Notwithstanding the sometimes-expressed view that proportionality involves only arbitrary evaluations,232 there is nothing “non-legal” about efforts to promote the proportionality of government action by considering its effects on relevant constitutional values.233

2. Bridge Between Courts and Legislatures

A second benefit of structured proportionality analysis is that it can provide a bridge between decision making in courts and decision making by the people, legislatures, and public officials. Proportionality doctrine arguably invites more participatory deliberation over constitutional rights, and it may achieve more compliance by legislatures and other officers with constitutional values by offering a rubric for decision making that is accessible to those other decision makers.234

Preliminary inquiries into whether challenged action has been authorized by law and has a proper purpose can be seen not simply as judicial checks on government action but as opportunities for the legislature to reflect on and improve its own legislative product. Insisting on proper purpose and legal authority focuses attention on the central role of legislatures in authorizing, and limiting, government conduct that affects rights.235 Assuming authority and proper purpose, legislative decision making may also take into account and thereby influence courts’ determinations of whether the proportionate means tests have been met. Proportionality doctrine can thus be seen as a reflection of the dual commitments of constitutional democracies—to the protection of rights and to democratic self-governance, which itself can be conceived of as a right.236

Moreover, in situations of epistemic or normative uncertainty, legislatures may be more empirically competent and democratically legitimate than courts in making prognostic factual determinations and in making accommodations among competing values.237 As Robert Alexy put it, when judgments about “suitability” (rationality) or “necessity” (analogous to minimal impairment) are in a zone of “epistemic uncertainty,” the fact that the legislature is democratically elected is a reason to accept its determination of these issues.238 When there is epistemic uncertainty—for example, whether decriminalizing marijuana would be as effective as criminalization in preventing dangers associated with that drug’s trade and use—legislative judgments about the necessity of the criminal prohibition prevail.239 When there is a “normative” stalemate—involving, for example, competing principles worthy of optimization as in the protection of workers’ rights and those of small employers in lay-off situations—legislatures have normative discretion to make different choices.240 And, in theory, the sequenced structure of proportionality doctrine allows for judicial deference to legislative resolution of some questions, such as minimal impairment, even if not on all questions.

One of Laurence Tribe’s critiques of John Hart Ely’s representation-reinforcement theory of judicial review was that Ely’s theory offered no guidance on constitutional meaning to legislators or executive branch actors.241 By contrast, the questions of proportionality analysis resonate with the competences of legislatures, especially in its inquiries about rational relation and minimal impairment, both of which have “predictive” factual components about the connection between the means chosen and the legislative goal.242 Legislators who understand that statutes will be evaluated under proportionality standards if challenged as infringing on individual constitutional rights will have reason to give attention to the rationality of the means, to whether there are other means less likely to intrude on rights, and to whether the gains to be achieved are weightier and of such a character as to warrant intrusions on protected freedoms.243 As Mattias Kumm has written, focusing public actors on the elements of proportionality review can have a

disciplining effect on public authorities and help[] foster an attitude of civilian confidence among citizens. The legal institutionalization of Socratic contestation helps keep alive the idea that acts by public authorities that impose burdens on individuals must be understandable as reasonable collective judgments about what justice and good policy requires to be legitimate.244

Proportionality considered in courts and in legislatures may differ: legislatures can focus on finding the best achievable solutions; “proportionality analysis” by courts can serve as a check against serious disproportionalities.245 For courts, the sequenced structure of proportionality doctrine offers benefits of consistency and transparency in methodology; but for both legislatures and courts, there are benefits from considering proportionality, even in less structured ways, as a principle of justice.

3. Justice, Law, and Judgment

Proportionality as principle or doctrine is a way to bring the demands of justice into greater harmony with the law of constitutional rights.246 Justice is not synonymous with law; it provides a critical platform from which to evaluate law. There is value in a legal system’s aspiring to do justice, as understood in its society. Attention to different factual contexts, as well as the need to confront the impact of general rules on particular cases in terms of proportionality, can help hone a juridical and political community’s sense of justice.247

Legal systems whose decisions do not resonate with widely held conceptions of justice may not be able over the long run to perform their basic functions. Such decisions undermine respect for law and for the legitimacy of courts. In the context of Fourth Amendment law, scholars have observed that the Court has tended to reject categorical rules and apply totality of the circumstances tests where the proposed categorical rule would benefit those who are the subjects of police searches, and to embrace categorical rules where they are permissive of police behavior.248 This pattern, together with the exclusion of officers’ intent (or pretext) and of state or local law in defining what is reasonable,249 cannot but tend to contribute to the lack of trust in police now prevalent in many minority communities. In this area, moving towards doctrine that permits a fuller range of the factors that people in ordinary life consider reasonable would help re-establish the law’s connection to justice. To be sure, constitutional justice will often be contested. Even so, proportionality doctrine helps clarify the grounds for decision and the relative importance of different components of justice, thereby providing a framework of analysis for resolving what is most importantly at stake.250

A related advantage of proportionality is the opportunities it provides for the development of what we might call constitutional judgment or “situation sense.”251 Mark Tushnet has argued that Justice Breyer’s dissent in Heller should not be understood as primarily about proportionality or balancing, notwithstanding its use of “proportionality” language; rather, it should be understood as about the application of legal judgment to complex settings.252 Tushnet has also argued that the Court’s First Amendment decisions in such cases as Snyder v. Phelps,253 United States v. Stevens,254 and Sorrell v. IMS Health Inc.,255 represent a form of “judicial pathology,” consisting of overestimating the harms that prohibitions on speech would cause and an insensitivity to the distribution of those harms.256 This pathology, Tushnet suggests, is connected to the “rule-ification” of the area and a related “fear” of making obvious judgment calls on issues of degree.257 Once a “rule” is announced, its function is to focus judicial attention only on the “rule” (that is, for example, asking whether a regulation is “content-based”), rather than on the purposes of the constitutional provision the rule is intended to implement. The arguments for “rule-ification” are stronger with respect to multiple decision makers, like lower court judges and executive officials, than with respect to the Supreme Court, which can always consider introducing an “exception” to a rule. In recent free speech cases, however, the Court has arguably deprived itself of the opportunity to engage with the purposes behind the presumptive rule against content-based regulation, reaching results that may well be inconsistent with the long-term constitutional judgments of the people.258

The Court in Stevens, for example, rejected the government’s argument for a “‘categorical balancing of the value of the speech against its societal cost,’” denying that the Court has “a freewheeling authority to declare new categories of speech outside the scope of the First Amendment.”259 Consider instead if the question had been whether the statute met a multi-part proportionality standard like that used in Canada. Presumably the Court would have found that it did not, because the statute was so “overbroad” it would have failed “minimal impairment.” But the Court would also have had to address such questions as whether the government’s purpose in reducing animal cruelty was legitimate and of sufficient importance to warrant some limitation of expressive activity, and whether prohibiting the commercial development and distribution of videos featuring animal cruelty was a rational means of achieving that purpose.260 The guidance provided by analysis of these questions might have assisted subsequent legislative efforts to address the problem through more narrowly tailored legislation.261 Proportionality analysis, in short, could help promote judicial engagement with basic questions of constitutional justice, reflected in judgments about legitimate or compelling purpose—and its relationship to the harms from limiting expressive activity presumptively protected by the First Amendment.

4. Process Failures Warranting Heightened Scrutiny

A different kind of argument arises from considering whether disproportionalities in the effects of government action may be a signal of failures in the legislative process that warrant increased scrutiny by the courts.262 On John Hart Ely’s theory, process failures resulting from conscious prejudice and intentional discrimination against minority groups warrant higher levels of justification and judicial scrutiny.263 A wider range of process failures might be signaled by disproportionalities in the application of law. Disproportionalities—such as those that occur when a law is more intrusive than necessary to serve the stated purposes—may signal an underlying problem, relating not only to conscious prejudice but also to failures of equal regard. Some may arise from lawmakers’ insufficient concern with disproportionate effects on the relatively powerless; some may reflect unconscious or unarticulated prejudices; some may arise from the simple inability to anticipate legislation’s effects. Each of these might be understood as a process failure: a failure, in Justice John Paul Stevens’s terms, to fulfill the government’s duty of impartiality to the people.264 Some such process failures may warrant heightened judicial attention or intervention.265

In the United States, the tiered structure of review applicable to rights claims under equal protection and due process already embodies to a certain extent the idea of proportionality, because more is required to justify laws in categories deemed likely to be of greater constitutional concern; so, too, does the role of less intrusive alternatives in areas of U.S. doctrine.266 There can be a large gap between “strict scrutiny” and “rational basis” review, however, seen in the contrasting treatment of overt racial classifications and neutral laws with a disparate impact based on race. The principle of proportionality supports Justice Thurgood Marshall’s suggestion that whether a classification violates equal protection should depend not on rigid ex ante categories but on a more flexible, more proportionate approach,267 as will be discussed further in Part V below.

***

Some of these benefits relate primarily to use of the structured proportionality doctrine of Canada and similar systems. But others—bringing law closer to justice—derive from greater use of proportionality principles even in differently or less structured contexts.268

IV. objections and responses

Some objections to proportionality as a standard for review would apply, generally, in any constitutional democracy. These objections—including to its indeterminacy, its asserted intrusiveness, its potential for inconsistent applications, its asserted irrationality, and its claimed incompatibility with strong conceptions of rights—are discussed in Part A. Other concerns about proportionality relate to particular aspects of U.S. constitutional law and culture; these are addressed in Part B.

A. General Objections to Proportionality as a Standard of Review

Although structured proportionality review’s responsiveness to legitimate government justification could help to protect rights while maintaining effective self-government, some argue that this very flexibility detracts from its quality as law, creating an unacceptable level of indeterminacy.269 The weight to give the indeterminacy critique depends to an important degree on what proportionality review would replace. It is one thing if it replaces a seemingly determinate categorical test;270 but if proportionality doctrine replaces a less structured “all things considered” approach, or an exception-riddled set of categorical rules, it might produce a more disciplined jurisprudence. And structured proportionality can co-exist with understandings of rights as “principles” requiring optimization,271 or as “shields,” requiring legitimate and strong reasons to interfere.272

Another argument is that proportionality review is too intrusive on legislatures, establishing a standard that cannot realistically be met.273 A version of this argument, which Alexy refers to as the “highest point thesis,” contemplating single right answers,274 is inconsistent with the recognition by leading proponents of proportionality of the existence of significant “zones” of legislative “discretion,”275 in which the legislature’s judgment will control. It is moreover inconsistent with widespread recognition that proportionality review, with its sequenced steps, is capable of being applied with “variable intensity.”276 The related claim that “proportionality,” like balancing, is more a legislative than a judicial competence277 ignores the degree to which while legislatures and executives may have particular knowledge and competence about, for example, the scope of national security risks and the best means to minimize those risks, courts have more capacity fairly to decide questions of individual rights.278

There are, to be sure, institutional concerns with using “standards” like proportionality, rather than “rules.” Non-judicial actors, like police, may find it easier to implement a rule than a standard.279 Rules, however, can lose their ease and clarity as their exceptions proliferate.280 Even if “categorical” rules would result in fewer errors, moreover, a standard may result in fewer “serious” errors, or departures from a common sense of constitutional justice, than its “categorical” counterpart.281

To the extent proportionality analysis allows courts to consider more factors, however, the range of reasonable applications may be broader, which may result in more consistency problems in lower courts in the decentralized system of U.S. judicial review.282 Recent experience with categorical rules in the United States suggests that neither determinacy nor respect for legislative outcomes is necessarily protected through such rules.283 Moreover, the U.S. Supreme Court’s “‘shrunken docket’”284 suggests that it has substantial unused capacity to control errors and promote consistency in the lower courts: the Court’s docket remains roughly half of what it was decades ago.285 The Court may be able to expand its docket and use some of that capacity to minimize inconsistencies in the lower courts’ application of proportionality.

The “proportionality as such” element of proportionality review has been most widely subject to critique, as unconstrained “balancing” of often incommensurable values and based only on the preferences of the judges.286 Indeed, some, including Jürgen Habermas, view the “proportionality as such” test as essentially irrational because it requires the weighing of incommensurables lacking a common metric.287 Even absent a common metric, however, judgments about the relative priority of two values can be rational.288 An example is “large-small trade-offs” involving a small sacrifice of one value for a large gain in another.289 It is a mistake to understand balancing in mathematical terms: rather, “proportionality as such” balancing should entail a reasoning process about the priority of one constitutional value as it relates to another in a particular setting.290 It is also worth noting that “proportionality as such” is the last in a sequence of inquiries and therefore is part of a more structured decisional process than “all things considered” balancing.291

A final and significant set of concerns is that applying proportionality doctrine is incompatible with the basic concept of a constitutional right,292 or might undermine the distinctively principled character of rights.293 Carol Steiker, for example, has suggested that understanding proportionality to be a necessary condition for government action intruding on rights might lead to the idea that proportionality is sufficient, thereby “[o]ccupying the justificatory field.”294 But on some accounts, even in jurisdictions applying proportionality analysis, one can recognize “core” aspects of rights that are viewed as entirely non-abrogable and not subject to limitation by arguments from proportionality.295 Judicial elaborations of human dignity in Germany, for example, striking down a law authorizing the shooting down of hijacked civilian aircraft, or in Israel, prohibiting privatization of prisons, show that deontological analysis can coexist with extensive use of proportionality doctrine.296 Moreover, structured proportionality analysis itself leaves room for the conclusion that a statute has an impermissible goal, one ruled out by the commitments to maintaining rights in a free and democratic society.297 Beginning the analysis with an inquiry into purpose and then focusing on the nature of the right and the severity of its infringement can help mitigate important concerns with narrowing of the “justificatory field.”298

B. Arguments from Lack of Fit with U.S. Constitutionalism

It is sometimes argued that Canadian or European approaches to rights analysis do not fit well with already developed U.S. constitutional law. To be sure, a highly contextualized analysis is necessary in evaluating whether approaches in one legal system can usefully be adapted in another. At the same time, it is important to recognize the multiple strands of possibilities for change within particular legal cultures.299 U.S. constitutional case law already includes several lines informed by the basic idea, and several of the doctrinal components, of proportionality review. Although the United States is unlikely to adopt proportionality as a general principle applicable to all challenges to government action, there is good reason to think that, in some discrete areas, U.S. constitutional law could benefit from greater use of both the principle and the structured doctrine of proportionality.

While the United States does not have the kind of limitations clause found in post-World War II constitutions, U.S. jurisprudence recognizes that limits on matters ordinarily understood as protected by rights can sometimes be constitutionally justified. Indeed, U.S. constitutional law in many areas contemplates “triggering rights” that generate strict review but that in the end, may not be “final rights” because the “triggering right” may be subject to limitation.300 Influences on contemporary “limitations” clauses are many, but among them is the provision of Article 29 of the Universal Declaration of Human Rights (UDHR),301 whose language was influenced by a proposal from an American Law Institute (ALI) committee.302 The German Constitutional Court’s influential proportionality doctrine did not derive primarily from the express limitations clauses of the Basic Law but rather from judicial elaboration of constraints on government regulation in the course of interpreting police law in the nineteenth century.303 And, as Stephen Gardbaum has argued, notwithstanding the absence of an explicit limitations clause, basic approaches to rights interpretation in the United States have much in common with those in countries explicitly using proportionality review.304

Nonetheless, some scholars have suggested that U.S. legal culture is hostile to proportionality review. They argue that balancing in the United States developed as an effort to limit the power of courts (acting on behalf of rights) to interfere with legislative outcomes, rather than, as in Germany, as a way of formalizing and protecting rights.305 Dennis is sometimes described as “symboliz[ing] to this day the most troubling risk of balancing: the danger of judicial capitulation to the legislature’s determination of the balance of interest in times of national security crisis,” a case whose “stigma” led the Court thereafter to “dissociate[]” itself from balancing.306 This adverse reaction to balancing was, in important part, historically contingent,307 and may now be weakening, at least in national-security inflected First Amendment case law.

Even if we assume that the predominantly categorical conceptual structure of free speech law will survive, there are a number of areas of contemporary constitutional rights law in which the U.S. does use balancing, or even proportionality.308 Richard Fallon, in describing strict scrutiny, notes that in addition to sometimes functioning as a close-to categorical rule, at other times strict scrutiny is applied as if it were “a weighted balancing test, similar to European proportionality inquiries.”309 Moreover, outside of cases governed by strict scrutiny, balancing tests are alive and well, and not necessarily hostile to rights protection. In Hamdi v. Rumsfeld,310 the plurality drew on the 1976 decision in Mathews v. Eldridge311 for “[t]he ordinary mechanism that we use for balancing such serious competing interests” to decide what process was due an American citizen detained as an enemy combatant.312 The invocation of balancing was rights-protecting insofar as the government had argued that “‘[r]espect for separation of powers and the limited institutional capabilities of courts in matters of military decision-making in connection with an ongoing conflict’ ought to eliminate entirely any individual process, restricting the courts to investigating only whether legal authorization exists for the broader detention scheme.”313 The Court rejected this and other arguments.314 A more accurate way to describe U.S. constitutional law is thus that in important areas the Court relies on balancing tests but does so in a less systematic way than its Canadian or German counterparts.

Some scholars argue further that U.S. constitutional law focuses on the “intent” of government actors, not the “effects” of their actions, in defining constitutional rights, an approach claimed to be incompatible with proportionality’s concern both with a challenged act’s purpose and its effects.315 But in some areas, narrowly focused intent tests have only recently—and contestedly—replaced more effects-oriented aproaches.316 One should not mistake a phenomenon that is no doubt present in some areas for a more general state of affairs.317 There are other significant swathes of U.S. constitutional law that are or have been effects-oriented. This is so not only in the Dormant Commerce Clause area,318 but is also characteristic of the second (effects) and third (entanglement) prongs of the Lemon v. Kurtzman test for Establishment Clause claims.319 Further, in Takings Clause jurisprudence, one inquiry focuses primarily on the effect of the challenged regulation on the property owner.320 In the First Amendment context, when “incidental” burdens on free speech result from content neutral regulation, the Court still applies an intermediate form of scrutiny.321 Although for some purposes the Sixth Amendment right to counsel looks to the intent of state actors,322 for other purposes the effects of action are the significant factors, as in determinations of ineffective assistance of counsel.323 Since 1997, Casey’s “undue burden” test has asked whether a regulation has the “purpose or effect” of creating a substantial obstacle to a woman’s choosing to abort a pre-viable fetus.324 And in Fourth Amendment law, the Court has determinedly turned away from intent. The Court has insisted that the actual intent of police officers in making a stop or arrest is irrelevant; what matters is whether there was an objective basis for a “reasonable suspicion” or for probable cause; the fact that the police may actually have been motivated to make a stop because of the suspect’s race, or preexisting bias against the suspect, is not relevant under the Fourth Amendment.325In this light, it is not accurate to describe U.S. law as having a general propensity only to be concerned with intent and not with effects.

Recent scholarship has also suggested that the United States is more skeptical about the possibilities of law in the hands of judges (and thus of proportionality review) than are Canada or Germany.326 The United States is, to be sure, more willing to leave to democratic processes decisions that, elsewhere, would be made by more expert, elite decision makers (as in the popular election of judges in many states within the United States). The U.S. Supreme Court is an empowered, activist Court, however, even without proportionality review; it has invalidated a significant number of federal statutory provisions since the early 1980s, in cases that include INS v. Chadha,327 NFIB v. Sebelius,328 Citizens United v. FEC,329 City of Boerne v. Flores,330 and several other First, Eleventh, and Fourteenth Amendment cases.331 Meanwhile, public confidence in Congress is at astonishingly low levels;332 a recent Harvard Law Review Foreword commented on the Supreme Court’s apparent “disdain” for Congress.333 Even if people believe their elected representatives are more legitimate decision makers than judges, they surely would not intend for legislators to act without reason, or to act in an abusive way.334 If judicial doctrine on proportionality can better focus legislators on good reasons for their action and at the same time encourage courts to take more seriously legislators’ reasons for acting, it may be a net gain for democratic decision making.335

Other kinds of objections to proportionality review flow from general interpretive approaches in the United States. Originalist claims are a distinctive feature of contemporary U.S. constitutional law; their force owes much to a historically specific reaction to the Warren Court’s legacy.336 Those committed to resolving constitutional controversies only by resort to the “specific meaning” of constitutional provisions at the time of the founding would presumably make less use of proportionality-like analyses. For most Justices, however, original understandings are only the beginning and not the end of analysis,337 and some questions simply cannot be resolved by resort to specific original understandings.

Some academic proponents of proportionality go too far in suggesting that text and precedents do not matter.338 In so doing they ignore important foundations of law’s legitimacy. Texts and their history and purpose matter: proportionality alone cannot provide a substantive theory of what interests are within the scope of rights. Specificity matters: a constitution requiring payment of “just compensation” for the taking of property imposes constraints that may not be enforced in its absence. Stare decisis emphasizes the role of precedent in constitutional adjudication (except where departures are sufficiently justified), thereby linking past, present, and future in a stable but flexible continuity. The long lines of precedent in many areas of individual rights, the different character of different rights, and other factors discussed below, caution against any massive reconstruction of U.S. constitutional law through the lens of proportionality.

V. defining the boundaries for proportionality review

I do not argue that the United States should embrace proportionality across the board. For one thing, the U.S. Constitution does not provide as clear a textual basis as exists in Canada for the adoption of proportionality as a pervasive test.339 The U.S. Supreme Court is not in the position of the Canadian courts interpreting the 1982 Charter, nor the German Court interpreting the 1949 Basic Law; the United States has no new charter of rights subject to interpretation for the first time. And the United States is a large country, with highly decentralized opportunities for judicial review in multiple court systems; a greater need may exist for categorical rules to achieve acceptable levels of consistency in the law (even if the Supreme Court were to expand its docket). Moreover, where reasonably well-functioning lines of law exist, developed over time, there may be insufficient reason to unsettle the law. Not all rights protected by the Constitution involve the kinds of principles that can best be applied through ideas of proportionality. Some rights may be better understood as concrete rules, requiring particular procedures to legalize the government’s use of coercive power.340 Other rights can be better viewed as normatively nonderogable guarantees.341

Finally, even when rights have components concerned with promoting proportionate government conduct, case-by-case application of proportionality standards may not be the best approach; formal application of a categorical rule over the course of cases may result in a better group of decisions overall.342 Given the draw of consistency in adjudication, moreover, rules are likely to emerge even from case-by-case applications of a proportionality standard;343 and what some call “definitional balancing” or “categorical balancing” might be reconceptualized to reflect conceptions of proportionality in light of the purposes of the right and its implementation in a decentralized system of justice. The goal of proportionality in government action, in the sense of justice and good governance by actual institutions, may sometimes be better served by more categorical rules.344 How then should judges determine whether an area calls for a more categorical, or case-by-case application of proportionality standards?

A. Different Rights, Different Roles, Different Texts

Not all rights have the same conceptual structure.345 Nor do all rights play the same role within the constitutional system. Some rights, like those associated with the Establishment Clause, have been viewed by some as concerned primarily with “excluded reasons” for government action.346 Doctrine implementing rights, like those secured by the Fifth Amendment Takings Clause, may on occasion draw on proportionality principles to analyze nonphysical actions of government that are claimed to constitute takings,347 but at the same time treat even minor permanent physical invasions as per setakings for conceptual or historical reasons.348 Further, the text of that right specifically provides the remedy for when a taking of property occurs—that is, payment of just compensation.349

The First Amendment’s protections of freedom of speech and association function as broad guarantors of democracy, securing freedom of political competition; they prohibit government conduct motivated by a desire to suppress dissent; and they secure a host of individual expressive freedoms. The First Amendment is also arguably emblematic of a particular form of constitutional identity for the United States.350 Application of proportionality analysis in an individual case-by-case way might be considered inconsistent with the symbolic importance of treating the First Amendment as providing strong protections. But there is no conceptual obstacle to providing strong rights protection through proportionality analysis by treating a government purpose to suppress ideas as per se illegitimate and by treating the value of freedom of expression as presumptively stronger than reasons for suppression in the “proportionality as such” stage. Still, categorical statements of presumptive rules might be thought to accomplish this in ways more consistent with symbolic or expressive aspects of this amendment (and its “shall make no law” text). But such categorical rules—including categorical exclusions for regulation for obscenity351 or fighting words352—can themselves be informed by considerations of proportionality. The possibility of identity-reinforcing benefits in framing First Amendment jurisprudence in the form of presumptive categorical rules does not answer what those rules should be or what exceptions to a categorical presumption against content-based regulation should be recognized.

Use of proportionality doctrine to review the reasonableness of a search is a different matter than its use to review free speech claims. The Fourth Amendment’s text protects “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures . . . .”353 Strong considerations of the rule of law and of popular conceptions of justice would support a proportionality approach to some Fourth Amendment issues now governed by categorical rules. Some years ago a scholar wrote: “When an officer acts reasonably, it would torture the English language to condemn his action as an unreasonable search.”354 Likewise, when an officer acts “unreasonably,” as the officer did in Atwater, it tortures any popular sense of what the Fourth Amendment means to find no violation. To treat the Fourth Amendment as favoring categorical rules to the same degree as the seemingly more absolute language of the First Amendment is to suggest that the text does not matter.355 And to assume that the proliferation of categorical rules will help constrain rather than liberate official discretion of police officers may be more heroic than realistic.356

B. Remedial Constraints

Adjudications of liability are always nested in particular remedial systems. The remedies available or required, and their consequences, may have constraining effects on how courts are willing to define the underlying right.357

1. The Exclusionary Rule

In the United States, Fourth Amendment remedial rules requiring exclusion of evidence have been applied, at least for a time, in a seemingly categorical manner.358 By contrast, in Canada the consequence of finding a violation of Charter rights is not necessarily exclusion of the evidence: under Charter Section 24, courts decide, case by case, whether admitting the evidence would “bring the administration of justice into disrepute.”359 The apparent rigidity of the U.S. exclusionary rule may thus militate against more generous interpretation of the right because of the consequences to criminal justice administration. Yet proportionality approaches might also support modifications in the U.S. approach to the exclusionary rule.360

William Stuntz suggested that the absence of attention to proportionality—including the “blindness to differences among crimes”—is one of the deepest problems in Fourth Amendment law.361 The trans-substantive doctrine of the Fourth Amendment, he argued, created a “reasonableness” gap in the application of the Fourth Amendment’s substantive standard.362 Stuntz also suggested that while the U.S. version of the exclusionary rule serves many useful purposes, the remedy has adversely affected the crafting of substantive Fourth Amendment doctrine and misdirected resources away from more fundamental questions of guilt or innocence.363 Given the number of existing exceptions to the exclusionary rule, it is possible that more might be gained than lost by adopting a more proportionate approach both to the substantive standards of the amendment and, possibly, even to the consequences of illegality.364

2. Equal Protection

Distinct remedial challenges are posed by successful equal protection claims, as their redress may require changes adversely affecting nonparties. For example, if a benefit is made available on terms found discriminatory, there may be options to equalize down, as well as up.365 Remedial complications may help explain why courts that apply proportionality principles in equality cases do so more deferentially in evaluating challenges to economic or commercial regulations.366

Given respect for democratic decision making, interests in the stablity of law, and concern for the reasonable expectations of third parties, there are reasons for caution in the application of equal protection standards to the great mass of legislation.367 Indeed, in Washington v. Davis,368 the Court rejected disparate impact based on race as a trigger for strict scrutiny, expressing concern that many statutes could not meet the standards of justification required by strict—then usually fatal—scrutiny. Experience with proportionality review elsewhere suggests that equal protection review could be implemented in a more proportionate way, one that does not automatically invalidate laws with such disparate impacts and that can recognize differences in the severity of impacts, especially on historically disadvantaged groups.369

Proportionality in equal protection review resonates with Justice Thurgood Marshall’s “sliding scale” view of equal protection law.370 In Dandridge v. Williams, the Court used relaxed “rational basis” review to uphold a state welfare law imposing a cap on benefits for families with dependent children that had the effect of giving less per child for children in families above a certain size.371 For Justice Marshall, who dissented, there were differences of constitutional magnitude between classifications affecting businesses and classifications affecting poor children. These differences could be explained by reference to proportionality and a form of “process” theory that Justice Marshall explicitly invoked.

A case involving “‘the most basic economic needs of impoverished human beings,’” Justice Marshall wrote, should not be reviewed under a mere rationality standard.372 Such a rationality standard accepted “extremes . . . in dreaming up rational bases for state regulation” because of “a healthy revulsion from the Court’s earlier excesses in using the Constitution to protect interests that have more than enough power to protect themselves in the legislative halls.”373 Here, Justice Marshall drew an implicit contrast between the interests of businesses, which can “protect themselves in the legislative halls,” and the interests of much less powerful, poor human beings, including children.374 He explained that where “the literally vital interests of a powerless minority[,] poor families without breadwinners,” are involved, “the relative importance to individuals in the class discriminated against of the governmental benefits that they do not receive” required more careful analysis of the government’s asserted reasons for the law.375 Justice Marshall’s emphasis on the relative importance of the rights is a plea for more proportionality in reviewing standards and in the justifications governments must proffer for the distinctions that their laws create.376

Justice Marshall’s rejection of “a priori definition[s]”377 in defining the standard of review reverberates with Justice Stevens’s later argument that “[t]here is only one Equal Protection Clause,”378 with a common standard: whether a legislature acting in good faith rationally could believe that the harm it was imposing was justified in support of a greater good. A single standard can be implemented with varying degrees of seriousness depending on the impact of the classification. An example of this kind of approach may be found in Plyler v. Doe.379 Striking down a Texas statute denying public education to children who reside illegally in the country, the Court wrote:

In determining the rationality of [the statute], we may appropriately take into account its costs to the Nation and to the innocent children who are its victims. In light of these countervailing costs, the discrimination contained in [the statute] can hardly be considered rational unless it furthers some substantial goal of the State.380

The idea in this passage from Plyler is that where the harm is great, a rational legislator would need much more convincing evidence of likely effectiveness towards a “substantial goal” before she could conclude that it was rational to impose the harm.381 This idea is consistent with both Justice Stevens’s and Justice Marshall’s approaches, as well as with the central idea of proportionality.

As noted earlier, disproportionality in the effects of laws, especially where laws have particularly adverse impact on traditionally discriminated against groups, may be a signal of process failures tainted by prejudices.382 It may reflect a “deliberate indifference” that is a close cousin to more active forms of prejudice.383 Rather than relying on tiers of review as on-off switches indicating when reasons must be more substantial, courts might view disparate impact on historically disadvantaged groups (especially if less harmful alternatives towards the asserted goals exist), as signalling a potential process failure requiring higher levels of justification.384

Indeed, the rigidly separated “tiered” standards of review may have led to the narrowed understanding of the substantive scope of the Equal Protection Clause in Washington v. Davis.385 Although some have suggested that U.S. constitutional law is generally oriented towards the prohibition only of intentionally violative acts,386 this is not a sufficient explanation for Washington v. Davis. At that time, as the Court noted, “there [were] some indications” in the Court’s own case law that intent was not the critical factor in making out Fourteenth Amendment violations.387As the Court also indicated, various courts of appeals had treated disparate impact alone as triggering heightened scrutiny388: such court of appeals decisions, the Court fair-mindedly said, “impressively demonstrate that there is another side to the issue”389 within the existing interpretive resources of U.S. constitutional law.

An important element in the Court’s interpretation of equal protection law was its concern that allowing equal protection claims “based solely on [a] statistically disproportionate racial impact” would have sweeping effects on a wide range of important laws.390 For example, the Court wrote, such jurisprudence might eliminate “various provisions of the Social Security Act” and “a whole range of tax, welfare, public service, regulatory, and licensing statutes that may be more burdensome to the poor and to the average black than to the more affluent white.”391 For the Court, “‘acceptance of appellants’ constitutional theory would render suspect each difference in treatment among the grant classes, however lacking in racial motivation and however otherwise rational the treatment might be.’”392 Thus, it concluded, “[d]isproportionate impact . . . [s]tanding alone” does not trigger strict scrutiny.393

It seems clear that the Court did not reach its conclusion because it thought racially disproportionate effect was of no concern; disparate impact was “not irrelevant.”394 Motivating the decision in Washington v. Davis in important part were the remedial consequences for a broad range of statutes under the then-usually fatal “strict scrutiny” tier.395 Given the categorical structure of two-tiered review that existed when Washington v. Davis was decided, this concern is understandable: intermediate scrutiny had not yet been identified at this time.396 With the Court’s decision in Personnel Administrator of Massachusetts v. Feeney,397 the gap between strict scrutiny for facial uses of race (or intermediate scrutiny for gender classifications) and rational basis review for facially neutral laws with known disparate impacts widened.398

Some of the Court’s recent cases have moved away from reliance on rigid tiers, without clearly indicating what is in its place.399 A standard focused not only on the nature of the classification but also on the relative nature of the harm complained of and its relationship to the particular government interests at stake would allow courts the flexibility to hold legislatures accountable without invalidating most legislation. Such a change in approach would help answer critiques of the Court’s position on disparate impact. Failing to recognize disparate impact obscures some invidiously motivated conduct that does take place, and does not recognize the constitutional harm to equal protection of the law that can result from unconsciousbias or from deliberate indifference to the situation of minority groups by members of more privileged groups.400 A more flexible standard for reviewing equal protection claims could treat disparate impacts differently from overt or intentional uses of race, without suggesting that disparate impact on a racial minority group, or other historically discriminated-against group, creates no greater constitutional concern than distinctions between businesses for tax purposes.

More specifically, the use of neutral criteria, claimed to have a disparate impact on already disadvantaged groups, need not be treated as presumptively unconstitutional in order to require some real scrutiny of the reasons for the practice under a single standard of review. Rather, a substantial “disparate” impact on a minority group long subject to discrimination might be viewed as a signal of a possible process failure, reflecting the operation of unconscious bias or deliberate indifference. Such a finding might be viewed as requiring—not the kind of scrutiny that overt uses of race require—but some degree higher than that applied to challenges to economic legislation that is not claimed to impair fundamental rights or rely on suspect classifications.401 Indeed, similar elements were proposed in the scholarly literature in the decades after Washington v. Davis was decided.402

Experience elsewhere has demonstrated that the burdens of justification, when significant disparate effects on a group historically the subject of discrimination trigger heightened scrutiny, are not necessarily fatal or even that difficult to meet. In the European Court of Justice (now the Court of Justice of the European Union), violations of an anti-discrimination rule may arise from facially neutral policies which have the “effect” of treating women and men differently.403 Under this quasi-constitutional standard for “indirect discrimination,”404 the European Court entertains challenges to neutral employment practices with disproportionate adverse impacts on women, including in evaluating wage scales providing lower hourly wages to part-time workers than to full-time workers,405 or in requiring minimum periods of full-time work to establish eligibility for pensions.406 In such cases the European case law took into account that women are disproportionately part-time workers and required employers to provide specific justifications for the exclusion of part-time workers from higher wages or other benefits; this justification ensured that there was no purpose of discriminating against women and that the discriminatory effects were justified as needed for legitimate economic purposes.407 Once a disparate impact on women (over-represented as part-time workers) is proven, the employer has the burden of showing that the difference in treatment satisfies the principle of proportionality.408

In the Danfoss case,409 the European Court held that a criterion of “mobility . . . to reward the employee’s adaptability to variable hours and varying places of work,” could “work to the disadvantage of female employees, who, because of household and family duties for which they are frequently responsible, are not as able as men to organize their working time flexibly.”410 The mobility-related compensation criterion thus arguably violated the “principle” of EU law “that men and women should receive equal pay for equal work.”411 Employers could seek to “justify the remuneration of such adaptability by showing it is of importance for the performance of specific tasks entrusted to the employee.”412 By contrast, in the Cadman case the Court accepted that pay scales based on experience would ordinarily be regarded as justified, rejecting the argument that their disproportionate effect on women would require further justification (absent a showing that duration of experience was not in fact job-related in particular settings).413 This EU case law suggests that a disparate impact standard, sensitive to the different social and life experiences of women and men, can be applied to existing laws without undue economic disruption.414

To the extent that equal protection violations in the United States were defined narrowly to focus on intent because of concerns about the risks of judicial intrusion on the existing legal structure,415 experience elsewhere with a more proportionate, flexible set of inquiries in equality cases suggests that the risk may have been overstated.416 Washington v. Davis led to a large gap between the stringent standard of review of facial or intentional uses of race and the lax “rational basis” standard of review for laws having disparate impacts on minorities—a puzzling feature of U.S. equality law. Many aspects of contemporary equality law might be thought to be implicated by this distinction.417

Is it too late for U.S. constitutional equality law to reconsider Washington v. Davis, in light of experience elsewhere? Perhaps not.418 A substantial disparate impact on historically discriminated-against groups could be treated as raising an inference of a prohibited motive (including “deliberate indifference”), which could be rebutted on a showing less stringent than “strict scrutiny” but more rigorous than “rational basis.”419Although the Court now treats any overt use of race as subject to the same standard of review (whether challenged by majority or minority group members), there are arguably constitutionally relevant differences between the intentional use of race to classify persons and the use of neutral laws that are “race-consciously” designed toward some legitimate end.420 Disparate impacts that adversely burden minority groups might be regarded as of greater constitutional concern than “disparate impact” harms to members of a majority—if not on a substantive theory of racial nonsubordination then on an evidentiary theory that such disparate impacts are likely to result from bias, whether conscious or not.421 A more proportionate approach to equal protection could allow courts to probe laws with substantial disparate impacts on racial minorities or women under the more flexible standards proposed by Justices Marshall and Stevens.422

3. Criminal Sentencing

In contrast to the remedial challenges of equality law and the exclusionary rule, proportionality could play more of a role in criminal sentencing without such complications.423 Prior to the enactment of the Sentencing Guidelines in the 1980s, there was very little appellate review of sentences in the federal system. Appellate review for compliance with the Guidelines was authorized on appeal by either the government or the defendant. Since 2005, when the Guidelines ceased to be mandatory, federal appellate courts have been authorized to review sentences for reasonableness.424

The Court has repeatedly considered the proportionality of death sentences and held them to be unconstitutional under the Eighth Amendment under various circumstances, as when for example, the crime was not intended to and did not result in death; but its non-capital case law has been parsimonious in reviewing prison sentences under the “gross disproportionality” standard.425 Indeed, the Court’s Eighth Amendment case law on non-capital sentences for adult offenders is sparse. In Weems v. United States, the first case in which the Court found a punishment to violate the Eighth Amendment, the Court referred to that Amendment as embodying a “precept of justice that punishment for crime should be graduated and proportioned to offense.”426 Although some Justices have argued that the Cruel and Unusual Punishments Clause bans only particular methods of punishment and not excessive sentences,427 majorities since Weems have found that it does ban severely excessive sentences. In Solem v. Helm,428 the Court reaffirmed that the Eighth Amendment prohibits not only barbaric punishments, but also punishments that are disproportionate to the offense, holding unconstitutional a life without parole sentence imposed on a petty offense recidivist. In Harmelin v. Michigan, a majority of the Court, three of whose members emphasized that the Eighth Amendment encompassed only “a narrow proportionality principle,” refused to apply Solem to invalidate a mandatory life without parole sentence for possession of more than 650 grams of cocaine.429 The case law suggests that while the proportionality of death sentences is subject to serious scrutiny, for non-death sentences of imprisonment the standard of “gross disproportionality” will rarely be met. Yet Canada, interpreting a very similarly worded provision in its Charter430 and applying its own judge-made rule of “gross disproportionality,” has taken a harder look at criminal punishments. For example, in R. v. Smith,431 the Canadian Court held that a mandatory minimum of seven years for all offenses involving the distribution of narcotics was grossly disproportionate because it applied regardless of distinctions in degrees of seriousness of the offense. The sentencing practices of some foreign nations,432 international tribunals,433 and some states,434 suggest that a more just and consistent approach to sentencing would be possible with greater attention to the proportionality of the sentence in light of the offense, the offender, and the treatment of comparable offenses and offenders. Proportionality as an outer limit based on the offense severity, as well as proportionality as a form of comparability with similarly situated offenders, have widespread support in the scholarly literature.435

Judicial resistance to reviewing sentence lengths as purely subjective436 is puzzling, as the availability of information on sentences for comparable offenses and offenders both within and outside of the jurisdiction provide objective anchors for gross disproportionality determinations based on treatment of others.437 In two recent cases, the Court has drawn from its death penalty jurisprudence and more closely scrutinized juvenile life sentences. It has held that a life without parole sentence for non-homicide offenses violates the Eighth Amendment as applied to minors, because of their characteristics; it also held that a mandatory life without parole sentence for a homicide crime is impermissible for juveniles, because an individualized sentencing determination, like those required for adults in capital cases, is required before imposition of the most severe lawful penalty.438 Whether these cases foreshadow a broader willingness to take a harder look at the constitutional proportionality of noncapital sentences is uncertain. Likewise uncertain is whether federal courts’ growing familiarity with review of sentences for reasonableness will contribute to further development of constitutional standards of proportionality under the Eighth Amendment. There is reason to think, however, that both law and society would benefit from more real attention to the problem of grossly disproportionate prison sentences than has occurred to date.439

C. Fragile Rights, Fragile Regimes

There may be a distinctive need for prophylactic rules either to protect a right that is particularly fragile or to protect the performance of particularly sensitive government functions.

It is widely believed that some rights are particularly sensitive to threats from the possibilities of enforcement and accordingly require prophylactic protection.440 In the United States, First Amendment rights of freedom of speech are understood to have this kind of fragility, and various doctrines have developed, including overbreadth, that constitute departures from ordinary adjudicatory practice.441 New York Times Co. v. Sullivan442 can be understood in this way: perhaps it is not that there is a First Amendment right to be negligent in reporting adverse facts about a public official,443 but that there is a First Amendment right to engage in robust critical discussion of such public figures. That right might be threatened if reporters’ actions were adjudicated under only a negligence standard, and thus plaintiffs must show malice or reckless disregard for the truth—as a prophylactic rule.444

In other areas the First Amendment’s reach has been narrowed by categorical rules, arguably reflecting some form of proportionality analysis behind the rule, and sometimes qualified by further categorical exceptions to the categorical rule. For example, the First Amendment has been interpreted to allow statutes that ban “fighting words.”445 The Court explained in R.A.V. v. City of St. Paul that fighting words do not entirely lack expressive content but could be prohibited based on aspects of the speech unrelated to their content that justify the limited restriction on speech.446 (On the majority’s account, a statute singling out racist forms of fighting words fell within a “content-based” exception to the exception for fighting words.) For the concurring Justices, prohibiting hateful fighting words posed little threat to First Amendment values because the “expressive conduct . . . is evil and worthless in First Amendment terms.”447 On either rationale, one can identify a categorical rule arguably resting on a relative evaluation of the harms from the speech and the harms from its prohibition.

Scholarly literature identifies distinctive U.S. doctrine protecting “hate speech” from punishment as an important aspect of the U.S. constitutional tradition.448 I do not here address the correct constitutional treatment of hate speech regulations. I contrast the majority’s analysis in R.A.V. v. City of St. Paul,449 which rested on a view that even “fighting words” had expressive value and that the ordinance was a “content-based” or “viewpoint-based” regulation of speech, with a Canadian case analyzing a similar issue through proportionality analysis. In R. v. Keegstra, the Canadian Supreme Court upheld a statute making it a crime willfully to engage in public speech calculated to bring hatred upon a group defined by race, color, ethnicity, or religion.450 The justices were sharply divided on the outcome, but the majority and dissenting opinions all applied proportionality analysis and addressed the same questions in the same sequence. The dissenting opinion, while agreeing with the majority that the statute had a pressing and substantial objective, argued primarily that it did not minimally impair free speech values (noting the risks of its misapplication and the severity of a criminal sanction), while also questioning the statute’s rationality and proportionality as such (noting its potential chilling effects and the possibility that criminal prosecutions would draw more attention and support to the racist views).451 Both majority and dissent in the Canadian case gave substantial weight to the special harms of hateful speech based on race or religion in ways that went well beyond the U.S. Court’s brief mention of the reprehensibility of the cross-burning in R.A.V.452 As the competing opinions in Keegstra suggest, the issue is one on which there are powerful arguments on both sides.

As Mark Tushnet has suggested, the U.S. Court may be hesitant to recognize exceptions to the general rule against content-based regulation because of a “fear” of judgment.453 Courts cannot, however, escape judgment; they can sometimes obscure the character of their judgments through nonpurposive extensions of rules, as arguably has occurred in several areas of First Amendment law.454 Structured proportionality analysis can help make more transparent the arguments for and against recognition of further categories of analysis of First Amendment claims. The Canadian opinions suggest that, to the extent the First Amendment is construed to prohibit even narrowly drawn hate speech statutes, it should not be because such expression is somehow viewed as having more value than obscene speech, or fighting words in general, but because of the risks of misapplication that such statutes historically present.455

This piece has for the most part focused on constitutional rights. But just as prophylactic rules are sometimes necessary to protect unusually fragile rights, prophylactic rules may also be designed to protect important government functions. Many such rules exist.456 Immunity for judges from civil liability for their adjudicatory acts, which is an absolute immunity under federal law, is famously justified on the basis that without it the costs of defending nonmeritorious suits would be too high, and fear of lawsuits would threaten judicial independence.457 Qualified immunity rules have been justified as necessary to protect officials’ ability to function.458 Such general categorical rules, designed to protect the government’s ability to carry out functions that could be jeopardized by intrusive remedies, might well be upheld through structured proportionality analysis notwithstanding their adverse effects on rights-remediation. Consider these questions from structured proportionality analysis: is the goal of protecting officials from the predictable effects of suits, most of which are nonmeritorious, a substantial one? Most would say yes. Is the provision of a high degree of immunity a rational means towards doing so? Plainly yes. But do the means “minimally impair” the rights of those who claim officials have acted unlawfully? This is a closer question, depending on empirical estimates of the effects of different forms of protection that might shift over time.459 Finally, application of immunity rules would probably be found proportional in all but the most unusual of cases, because to freely make exceptions would undermine the protective purpose of the rule.

Conclusion

Embracing proportionality as a principle does not necessarily support its doctrinal use in all areas of adjudication.460 Proportionate justice concerns could, in some areas, lead to categorical rules rather than to contextualized case-by-case determinations. Using proportionality to define violations, of course, does not dictate remedies or exclude definitions of rights based on separate deontological or historical questions. However, greater use of proportionality, as a principle and as a structured form of review, has several potential benefits. It could enhance judicial reasoning by clarifying justifications for limitations on freedoms. Proportionality might also improve the outcomes of adjudication by bringing U.S. constitutional law closer to (admittedly varied) U.S. conceptions of justice, in ways consistent with the demands of effective government. Finally, proportionality may be democracy-enhancing, both in providing a shared discourse of justification for action clamed to limit rights and in providing more sensitivity to serious process-deficiencies reflecting entrenched biases agaist particular groups.

Justice Scalia has famously argued that the rule of law is a law of rules.461 But sometimes the rule of law requires attention to the “reasonableness” of conduct. Sometimes considerations of degree will bear on the formulation of a categorical rule in ways that the questions of proportionality analysis can help answer. If we are proportional in our application of proportionality, we may be able to improve much criticized areas of constitutional law while retaining an important role for presumptive categorical rules. Expanding existing proportionality review of criminal sentences would bring more justice to the criminal justice system. In equality case law, the Court’s definition of the constitutional right as excluding injuries to minorities from “disparate impacts” of neutral laws may have resulted from fear of applying the usually fatal “strict scrutiny” test;462 it might be reconsidered in light of case law from other jurisdictions involving more proportionate approaches to defining the violation. Asking the “proportionality as such” question might clarify and strengthen the “compelling interest” test used in some First Amendment areas. Where the Constitution itself uses “reasonableness” as a criterion, as in the Fourth Amendment, embracing proportionality in place of some of the more categorical existing approaches offers substantial benefits.

Consider Atwater again: if there were no less rights-impairing alternative that as effectively advances legitimate law enforcement interests, and if the harm to protected constitutional rights were outweighed by the need to advance those law enforcement interests, then a categorical investigatory rule could be upheld. An opinion following the Canadian approach, though, would take fuller account of the individual’s constitutional rights claim than did the Atwater Court. In so doing, the Canadian approach might well be more convincing than opinions that focus primarily on authority for the law enforcement action, rather than the reasons for it. Unlike trial courts’ decisions, police officers’ decisions made in their daily interactions with citizens are highly unlikely to be redressable through the means of subsequent review. Further, when police officers accustomed to unreviewable exercises of discretion make unnecessary arrests or commit gratuitous violence, the harm to the subjects of police abuse—as well as to respect for the rule of law—can be high.463

Perhaps the Court’s decision in Atwater could be viewed as a form of empirical humility about the presumed expertise of police as compared to courts. Perhaps it could be regarded as manifesting respect for democratic federalism, notwithstanding case law treating authority under state law as irrelevant.464 At the same time, given very high incarceration rates in the United States and evidence that the criminal justice system falls with greater severity on members of already disadvantaged groups,465 it is by no means clear that law enforcement officers need more, rather than less, insulation from judicial review of the constitutionality of their actions.

Where the Constitution itself uses “reasonableness” as a criterion, as it does in the Fourth Amendment, the use of categorical rules that treat patently unreasonable conduct as constitutional does a disservice to the rule of law and fails to protect express constitutional values.466 When the Constitution requires the vindication of such large-scale commitments as “equal protection of the laws,” or “due process of law,” embracing a more flexible, proportionality-based approach may better protect constitutional justice than will existing categorical approaches, by offering a check on governmental indifference or blindness to acute harms caused to those less able to protect themselves in political processes. Incorporating concerns for proportionality across larger areas of constitutional law may also allow for more meaningful participation by all branches of government in the ongoing process of working the Constitution to achieve effective and human rights-protecting governance.467