Jus Cogens

pp 1–17 | Cite as

Political-Liberal Legitimacy and the Question of Judicial Restraint

  • Frank I. MichelmanEmail author
Original Article


The term “judicial restraint,” applied to courts engaged in judicial constitutional review, may refer to any one or more of three possible postures of such courts, which we here will distinguish as “quiescent,” “tolerant,” and “weak-form.” A quiescent court deploys its powers sparingly, strictly limiting the agenda of social disputes on which it will pronounce in the constitution’s name. A tolerant court confirms as valid laws whose constitutional compatibility it finds to be reasonable sustainable, even though it independently would conclude to the contrary. A weak-form court acts on the understanding that its pronouncements on matters constitutional will be duly open to considered rejection by other political agencies. Theory commonly tends to treat the question of judicial restraint as turning on a bedrock political value of democracy. We may also, however, understand debates over judicial restraint in the light of a different bedrock value, that of political legitimacy. Where democracy is the focal concern, debaters may tend toward conflating into one measure the three dimensions of judicial restraint. A focus on legitimacy rather tends toward a dis-bundling of the three dimensions, thus complicating the choices while also clarifying the stakes. The political philosophy of John Rawls helps us to see how and why this occurs.


Constitutional law Judicial restraint Political legitimacy Political liberalism John Rawls 

1 Locating the Paper: the Constitution’s Complex Role in Liberal Constitutional Democracy

This essay comes as one of a series concerned with a tangle of functions with which liberal constitutional democracies invest the substantive parts of their constitutional laws. These are, as I here will name them, a “regulatory” and a “justificational” function.1 The regulatory function is forward-looking in time. It is to constrain political outcomes yet to come in directions preselected by the constitution’s authors. The justificatory function, by contrast, fixates on the present. It is to provide a basis on which free and equal citizens, perhaps finding deeply wrong or misguided some of the laws right now issuing from the duly constituted authorities, can nevertheless freely and reasonably accept those laws and be prepared normally to abide by them. (Of course, this dual classification does not, in itself, conjure up the full array of reasons and motives by which observers and theorists would variously explain the introduction into a country’s legal practice of a layer of substantive constitutional law.2 The classification only sorts out these sundry possible aims along one axis of differentiation among them.)

Regulation-by-constitution roughly says, “We the constitution’s progenitors have certain wishes regarding the future conduct of government in this country. By putting these wishes into a constitution, we mean to secure the efficacy of our wishes over time.” Justification-by-constitution roughly says, “Allowing as we must for the frequency and depth of expected disagreement in this country over sundry legislative policy choices and directions, it is still the case that everyone here has reason to accept and respect as law the legislative outputs of our political order in force—not just because it is, in fact, the one in force, but given also the assurance that those outputs issue in conformity to certain instructions contained in this constitution to which we are just now pointing.” Seeing thus how justificational force is coupled to regulatory effect, we might suppose that normally, at least, the two functions will coincide, so that the pursuit of one is also the pursuit of the other. On a closer look, though, one finds that in some respects and in some contexts, the two pursuits are not obviously compatible.

The two can come apart, it seems, as guides to exercises of power by citizens, judges, and officials at various points in a country’s constitutional history. Awareness of the co-imbrication of the regulatory and justificatory functions in constitutional-democratic thought thus can help us—so I have argued—to understand and to cope with sundry problems and debates seemingly also chronic in that thought. First of all, such awareness sharpens and clarifies certain large questions faced by constitutional instigators, starting with (i) the choice whether to have a legal constitution at all (as opposed, say, to one that is purely “political”),3 and then (ii) if “legal” the constitution is to be, the choice between one that is wholly or partly written and one that is all and only unwritten4; and then going on to (iii) how far constitutional framers rightly can go (or must go) in seeking to cement into constitutional law their own deepest convictions of the demands of political justice,5 (iv) the appropriate level of generality at which to cast the constitution’s substantive guarantees,6 and (v) the extension of such guarantees beyond civil rights and liberties to “social” matters such as access to social position and material needs.7 A due regard to the justificatory function—always alongside the regulatory—furthermore sheds light on (vi) constitutional democracy’s chronic troubles over judicial review,8 and (vii) the recent burgeoning of interest in “dialogic” and “weak-form” review9; and then on to contestations (viii) over approaches to constitutional interpretation—say, originalist/textualist as opposed to purposive\constructivist,10 and (ix) over the meaning and value of constitutional fidelity,11 and then ultimately, (x) over the resilience of claims for justification-by constitution—and so even, perhaps, of claims on behalf of liberal constitutional democracy tout court—in conditions of not just “reasonable” pluralism but, increasingly in our countries, of “hyperpluralism.”12

Into that foregoing topical enumeration, this current essay will fit just following item (ix) on constitutional fidelity (but it will also, as we shall see, reach back strongly to item (v) on constitutionalization of social rights). Its topic is the question of judicial restraint, specifically with regard to courts at work in the constitutional-legal field.

2 Judicial Restraint, Democracy, Legitimacy

2.1 Modalities of Judicial Restraint

In most constitutional democracies today, courts of law—one or more of them—pronounce on the constitutional compatibility of acts of other political agencies. In what ways and degrees (if any) should we wish courts in such matters to proceed with restraint? But then what do we mean by restraint? Constitutional lawyers and theorists use “judicial restraint” to refer to any one or more of three distinct postures of reviewing courts, which we can name as “quiescent,” “tolerant,” and “weak-form.”

2.1.1 Restraint as Quiescence

Courts differ in their dispositions to widen or to narrow the range of social disputes on which they will see fit to pronounce in the constitution’s name. We commonly call “activist” a court that uses loosened notions of standing and justiciability, or widened readings of the substantive reach of constitutional guarantees, to expand that range to the utmost.13 A quiescent judicial posture is the opposite from that.

2.1.2 Restraint as Tolerance

A court’s posture is tolerant, I shall say, when the following conditions hold: (i) The court works within a political community which accepts that questions of a law’s compatibility with constitutional requirements can sometimes be open to answers pro and con that are honestly and competently defensible on both sides, so that those on the losing side need not see the loss as a rupture of a regnant constitutional pact; and (ii) the court upholds as constitutional laws that it, for its own part, would judge to be non-compatible, as long as it finds the opposite conclusion to fall within bounds of what is competently defensible. (A tolerant court thus, it may seem paradoxically, upholds as constitutional some number of laws whose enactment its members would have felt obliged to oppose as sitting parliamentarians sworn to support and defend the constitution.)

2.1.3 Restraint as Weak-Form

A weak-form court accepts that its pronouncements on matters constitutional are legitimately open to considered rejection by other constituted political agencies—the court’s voice to that extent becoming one (more or less respected) voice among others—and conducts itself accordingly.14 A court’s posture is oppositely “strong-form” insofar as the court counts on the country to treat its pronouncements as fixtures thenceforward in the country’s constitutional law-in-force, binding as such on other political agents and revisable only by formal constitutional amendment or subsequent reversal by that court.15 (By “pronouncements,” here, we mean not just a court’s punctual thumbs-up or thumbs-down on this or that law’s constitutionality but its full ratio decidendi, the steps in its reasoning that would be considered binding (say) on lower courts in the system.)

Mark Tushnet (to whom we owe the “strong form”/“weak form” terminology) writes that the mark of a weak-form court is that its constitutional-legal pronouncements will be “expressly open to legislative revision in the short run.”16 “Short run” is important here. Even in undoubtedly strong-form systems, over courses of historical time, conditions undergo change and perhaps, along with them, widespread societal understandings of the point of constitutional guarantees. New legislative initiatives may call forth new and possibly modulated strong-form judicial responses. Those responses may themselves then exert a steering effect on the next run of legislative initiatives. The process over time may end in a somewhat reconfigured consolidation of constitutional doctrine from where the strong-form court had left it some time back. All of this, as experience shows, inevitably occurs within systems of strong-form judicial review. As long, however, as the understanding prevails at all times that only the court itself can displace its latest pronouncements from their force as law binding throughout the system, the practice is strong-form, not weak-form, in the classification I have in view.17

Note that our three modalities of judicial restraint are notionally independent. A court that weakly renounces authority to impose its own answers conclusively across the system as a whole may still be disposed to treat expansively the range of controversies coming within its purview to speak at all, and it may still refuse to allow that there might be more than one true answer (its own) to a question of constitutional application. A court that tolerantly allows for a possible plurality of competent answers may still expect the country to accept its own answers as operationally conclusive on the system as a whole, and may still define quite broadly the range of matters on which it will speak at all; and so on around the circle.

2.2 Grounds for Judicial Restraint: Democracy and Legitimacy

American legal pundits—to take us as an example—typically frame the question of judicial restraint as one about democracy—or, more precisely, as one about choice among various conceptions of democracy. In any and in all of its modalities, “restraint” presents itself as an offset to impairments to rule by current political majorities that necessarily must flow from judicial policing over permissible legislative content. But such impairments are deemed problematic in the first place—a “difficulty,” as they have been called18—just and only insofar as we count such a curbing of current majorities as “deviant” from democracy rightly understood.19 Some do and some do not so count it. Contentions over judicial restraint thus have tended, with us, to end up as contentions over the best understanding of a bedrock political value of democracy.20

My interest here will be in tying debates over judicial restraint to a different bedrock political value, to wit, the value of political legitimacy in its sense of moral justification for the force of laws enacted over strong opposition and dissent. A constitution’s presence in a country’s legal order can and does often serve as a main prop for such justification. (When someone complains bitterly about the immorality or stupidity of some legislative act or policy, Americans think it highly germane to respond that the act or policy has after all been (or predictably will be) judicially upheld as constitutional). That sort of reliance on a country’s constitution for political justification receives philosophical support in the work of John Rawls. It will be a part of my design here to show how the Rawlsian exposition pushes toward a dis-bundling of judicial restraint into the three component postures I have described.

A suggestion of judicial restraint out of Rawls may possibly come as a surprise to readers calling to mind Rawls’s well-known remarks on the supreme court’s role as protector of the higher law in a scheme of dualist democracy,21 and, relatedly, on the court as “exemplar of public reason.”22 I mean to show that there is, even so, something to the suggestion. I undertake this effort not out of any settled conviction that Rawls has completely right the matters covered here. Rather, my belief is that he has enough right to make some arguable implications of his views for the judicial-restraint question a matter of interest for friends of liberal constitutional democracy.

I write “arguable” implications, because teasing out these implications will involve me in a schematic reconstruction of a so-named liberal principle of legitimacy (hereinafter sometimes “LPOL”) as proposed and developed by Rawls in his book Political Liberalism (hereinafter sometimes PL), against a background assumption of a society that is approximately well ordered by Rawlsian criteria. Mine here is a deliberately aggressive reconstruction, perhaps destined to strike some readers as excessively formal and procedural, but that is a part of the point of the exercise. Seeing what results from the LPOL under my aggressive reading could then become a part of a case (should that be your view) for finding that not to be the best reading of Rawls.23

3 The Constitution as Public Platform of Justification

3.1 Justification-by-Constitution: the Liberal Principle of Legitimacy

Rawls brings forth the LPOL in answer to an urgent concern—-Rawls in fact calls it “the problem of political liberalism”24—about the basis on which, in conditions of what Rawls called a “reasonable pluralism” of moral, philosophical, and religious doctrines and outlooks,25 democratic citizens can possibly hope to justify to each other as free and equal their coercive impositions—political majorities over dissenters—by votes touching fundamental matters over which the citizens are morally, philosophically, and religiously divided.26 The principle reads as follows:

Our exercise of political power is proper and hence justifiable [to others as free and equal] only when it is exercised in accordance with a constitution the essentials of which all citizens may reasonably be expected to endorse in the light of principles and ideals acceptable to them as reasonable and rational.27

According, then, to the LPOL of PL, assurance of every law’s compatibility with a country’s constitution can supply justification for a citizen’s acts in support of laws that are otherwise deeply and reasonably contested. Of course, not any old constitution can thus serve as a public platform of justification among citizens free and equal. Necessarily required by the LPOL is a constitution to which the citizens all conscientiously can point, as a sufficient basis for reciprocating expectations among them for normal compliance by each with laws that issue in accordance with its terms. Presupposed, then, by the LPOL is the possibility of such a constitution, among rational citizens responding reasonably to their social situation.28

By “constitution,” here, Rawls plainly has in view a legal one: a body of higher-level laws within a country’s positive-legal order.29 He envisages a publicly legible script of institutionally mandatory terms, such that (i) citizens see in common what are those terms,30 and (ii) any citizen reasonably disposed can find those terms acceptable by all so disposed as an institutional framework for politics going forward, despite a virtual certainty that from that framework will issue some outcomes that one or another fraction of citizens will find to be morally, philosophically, or religiously repugnant. As long as—so the thinking goes—outcomes are guaranteed to fall within bounds of a good-faith application of those terms, such repugnancies should not be such as to warrant cessation of cooperation by any citizen both reasonable and morally conscientious.

3.2 Proceduralism, Thick and Thin, the Scheme of Liberties and their Central Ranges

Justification-by-constitution works by a deflection of divisive questions of legislative policy and value (does this law or policy merit the respect or rather the contempt of a right-thinking person?), to a different question (is this law or policy constitutional?) for which the answer is to be publicly apparent, or at any rate is to be ascertainable by means that are an order of magnitude less open to divisive dispute than are the deflected substantive disagreements. It thus offers itself as a procedural response to the special challenge of democratic political justification. It is true, of course, that the Rawlsian constitutional essentials are in some part substantive, setting restrictions or requirements on the goals to be sought or effects to be wrought by procedurally authorized acts of legislation and administration31; but still, they work as part of what remains overall a procedural device. (The political system, writes Rawls, “would not be a just procedure if it did not incorporate” liberty of conscience, freedom of thought, liberty of the person, and equal political rights.32).

This proceduralizing project of justification-by-constitution (as we may call it) sets up a Goldilocks dilemma. A justification-bearing constitution’s terms of assurance will have to be not too thick, but also not too thin. In order to sustain the regime’s acceptability to all reasonable and rational citizens, those terms will have to stop short of express and permanent foreclosure of questions over which reasonable citizens divide (“not too thick”), leaving those questions for future continuing examination in the democratic political venues of daily life. To that end, in the view of Rawls, the roster of justification-supporting constitutional essentials is to be kept short and its items cast at accommodating levels of abstraction.33 But that roster still must carry a core of common meanings (“not too thin”) sufficient to render both mutually coherent and widely persuasive the claims of citizens to each other of the worthiness of any conforming regime for continued support. The challenge is to cater for a sufficiency at all times, but never at any time an excess, of publicly settled meanings for the constitutional essentials.

Rawls explains how this challenge can possibly be met. Among free and equal citizens, any justificational load-bearing constitution will include guarantees respecting certain liberties under abstract names such as “liberty of conscience” or “the right to hold (personal) property.” But since the liberties thus named can all without strain be extended in ways that will sometimes bring them into conflict, “the institutional rules which define these liberties,” as Rawls writes, “must be adjusted so that they fit into a coherent scheme of liberties secured equally for all citizens.”34 The “scheme,” then, will have to be unified by some known, single, overall governing aim that can consistently guide the adjustments (of which guarantees? in which particular respects?) as the needs become manifest by courses of events. That unifying aim, then, must itself, in effect, compose a term of the reasonably and rationally acceptable constitutional pact.

As an example: Rawls defends for this purpose a guiding aim to secure for each citizen the conditions of a full and adequate development and exercise, over a complete life, of certain moral powers of the “reasonable” and the “rational.”35 The point to see for now, though, is that the liberties listed in the underlying conception are never reasonably to be understood as “absolute;”36 rather they all stand subject to being institutionally adjusted as experience may show is required to hold them together as a unified expression of political ends and values. Exposure to such institutional adjustment of the named liberties thus becomes itself a key clause in the deal. But then we have a problem. As one horn of our Goldilocks dilemma (we said), the substantive terms of a justification-bearing constitution in the LPOL must at all times carry a core of commonly agreed meanings, sufficient to render coherent and persuasive the claims of citizens to each other of the worthiness of any conforming regime for continued support. Is that requirement compatible with attachment of a “subject to institutional adjustment” codicil to the constitution’s requisite terms of guarantee?

It would be, Rawls points out, as long as we can assume that (i) there resides within each named essential guarantee some widely agreed, fixed core of meaning or “central range of application,” and (ii) there is always “a practicable scheme of liberties ... in which the central range of each is protected.”37 Supposing those conditions satisfied, we can then understand the justification-bearing constitution of the LPOL to include, as one further stipulation, a guarantee that all institutional adjustments of the extensions of the liberties will “preserve intact” the central range of application of each.38 The problem then would stand solved.

3.3 The Supreme Court as Referee

Except, that is, for one further wrinkle. Owing to the intrinsic reasonable contestability of the issues,39 but especially in the visionary pluralist conditions posited by Rawls, disagreement among reasonable citizens cannot be expected to stop at the water’s edge. Inevitably, disagreement will extend even to setting the bounds of the central ranges of the justificational load-bearing constitution’s named essential guarantees.40 The procedural project of justification-by-constitution then will require a further convergence of citizens on an institutional service whose judgments regarding such questions they trust to fall within the bounds of reasonable acceptability. It is with a view to fulfilling this role of trusted service that Rawls defends the use of courts as authoritative public arbiters of the fulfillment of the constitutional essentials.41 It is not that Rawls assumes the requisite institutional service must necessarily be a law-court. Rather, Rawls in this work finds no cause to upend such an arrangement when once it has become a settled part of a country’s political practice. He takes that choice to be a reasonable one in line with the constitution-centered proposition on legitimacy.42

3.4 Liberal Justice Conceptions as a “Family”: Trouble for the LPOL?

PL was republished in paperback form in 1996. Keeping the book’s main text from 1993 unrevised, Rawls did also, however, include a new “Introduction to the Paperback Edition.” Among a number of new emphases,43 Rawls there introduced the idea of a “family” of somewhat conflicting but all-of-them reasonable liberal conceptions of justice. A society, Rawls now wrote, can be reasonably well-ordered in the constitutional-democratic tradition while reflecting, in its political arrangements, any one of an indefinitely numerous family of underlying liberal conceptions of justice. The family members have in common that they pick out certain “rights, liberties, and opportunities” to which they assign a special priority.44 But there can be “different and incompatible” reasonable such conceptions, and conscientiously reasonable citizenship cannot demand more of you or of me than to draw our resolutions of fundamental political matters from reasons consorting with that member of the family which is for you or for me “the most (more) reasonable.”45 For John Rawls, that member is the conception he calls justice as fairness and develops at length in the unaltered main text of PL; but other liberal-minded thinkers, as Rawls now expressly allows, can and do reasonably disagree.46

This new emphasis on reasonable disagreement within the liberal fold, not just at the level of constitutional-legal application but at the level of a constitution’s underlying conception of justice, would seem initially to spell trouble for the LPOL of PL. Recall our pending solution to the problem posed by the insertion of a “subject to institutional adjustment” clause into the justificational load-bearing constitution of the LPOL.47 We could insert, we said, a further clause to the effect that no such adjustment would invade a “central range of application” of the liberties guaranteed by the constitution. That solution would depend, we said, on fulfillment of two conditions: (i) that there is a practicable scheme of aptly adjusted liberties that protects the central ranges of each, and (ii) that the ranges treated as central by this scheme are ones on which all reasonable views converge. But are these conditions satisfied in point of fact? Regarding the first, Rawls has offered historical experience as evidence of satisfaction, or at least of the possibility thereof.48 Regarding the second, however, the new Introduction’s admission of a family of reasonable but conflicting underlying liberal justice conceptions must open a space for doubt.

The hitch comes with the need, affirmed by Rawls, to supply an underlying normative conception by which to guide reciprocal adjustments of the ranges of the liberties listed in a constitution, as required to maintain them as a unified, coherent scheme.49 As we have noted, the criterion advanced by Rawls is to keep the liberties, in combination, conducive to the development and exercise by citizens of certain powers of moral agency.50 Thus, for example, the basic liberty to hold property will have as its core mission the assurance to each citizen of “a sufficient material basis for a sense of personal independence and self-respect” to allow for the development and exercise of these agency powers.51 That liberty, then will be open to possible institutional curtailments in respects deemed to fall outside that core (but to which other liberal views of the moral basis for property rights might take exception)—for example, denial of inheritance rights, or a move to socialized ownership of means of production.52

The troubling question is whether any guiding criterion that you or I or John Rawls might propose will not be subject to reasonable intra-liberal disagreement in conditions of pluralism. If we ask from whence Rawls draws the agency-powers criterion, the answer will be: from a certain conception of the person “as political” or “as citizen,” embedded in an idea of justice as the pursuit of fair terms of social cooperation, put forth and defended by Rawls in PL under the name of justice as fairness.53 Defended by him, that is as “the most reasonable” conception, for such a purpose, within bounds of the broad tradition of constitutional democracy. The most reasonable, that is (as we now must add), among other, conflicting but still reasonable, liberal conceptions.54

That last concession places new pressure on what we have been seeing as a necessary condition for a “central ranges” kind of solution for the Goldilocks dilemma: that is, a convergence of the reasonable on trust in a supreme court (or other designated institution) to resolved disagreements over the strongly guaranteed constitutional essentials, in a constitution sufficiently thick with such essentials to sustain the claims of citizens to each other of the worthiness of support of any conforming regime.55 With that observation, the table is finally set for our consideration of the dimensions of judicial restraint in Rawlsian constitutional democracy.

4 Judicial Restraint for the Rawlsian Supreme Court?

As laid out so far, the Rawlsian program for justification-by-constitution does not yet tell us all we need to know about how a reviewing court ought to approach its dealings with legitimately “hard” constitutional cases (to borrow Ronald Dworkin’s term for them),56 or about the level of conclusiveness with which the court in such cases should expect its pronouncements to be received by other political agencies. In terms of our agenda for this article, we still have to locate the Rawlsian supreme court’s position along each of our three axes of judicial restraint: quiescent, tolerant, and weak-form.

We can start by reminding ourselves of the reason why—in terms of the Rawlsian program—we are, in the first place, setting up a supreme court with powers of constitutional review. In pluralist conditions, operation of a constitutional-justificational procedural pact apparently requires at least one trusted agent to nail down decisive applications of its terms in uncertain or disputed cases, and the supreme court is to serve as such an agent. Accordingly our question here will be: What is there in that assignment of judicial role and responsibility to impel us in one or the other direction along any of our three axes of judicial restraint? My suggestion from the start has been that the answers will not necessarily accord with those we would draw purely from democracy-based concerns, whether those be to secure the regulatory will of popular constitutional-legislative framers or to heed the interpretations of that will by current political majorities. Our aim is clarify the independent bearing on judicial restraint (vel non) of a placement on constitutional law and constitutional courts, in the manner herein described, of a burden of justification for current collaboration in the force of law.

By way of posing the question more sharply still, we can suppose the system before us to be one that sets up the supreme court—assisted, as it may be, by lower courts—as its one and only trusted final decider of applications of the constitutional essentials, on the expected satisfaction of which political justification solely depends.

4.1 Quiescent or Activist Court?

Consequences then might follow that would strain acceptability in our current constitutional cultures. On those perhaps somewhat drastic assumptions, the oversight of the supreme court logically should extend to all potentially divisive, publicly salient disputes raising colorable claims of violation of the essential guarantees of a constitutional-justificational pact. That logic points toward a judicial posture that we would see as distinctly not quiescent but rather as activist and expansionary, disposed toward broadening rather than narrowing the range of disputes over which the court will pronounce in the constitution’s name. Seemingly implied would be a rollback from the more restrictive norms of justiciability currently entrenched in at least some of our legal cultures. These would be the commonplace norms restricting law courts to decision of live (not hypothetical) controversies, between parties having material stakes in the outcomes, for which the law supplies conventionally justiciable standards.57

The motivation for these restrictions is always, in part—and especially so in the constitutional field—the protection of majoritarian democracy against overreach by headstrong courts. The restrictions, alas, do not fit so neatly with the constitution’s justificational function. The problem then is that justiciability constraints can block the supreme court—on our assumption, the uniquely trusted institutional arbiter of constitutional compliance—from resolution of disputes over application of constitutional essentials on which legitimacy might well be thought to depend, to that extent undoing the settlement envisioned by the LPOL. A clear case in point, from the USA today, is judicial suspension of our constitution’s political-equality guarantees when faced with so-called political gerrymanders.58 Without denying the fact or urgency of the problem, our supreme court has so far blocked our federal lower courts from addressing it, citing reasons of a lack of a judicially manageable standard59 or (sometimes) of a plaintiff with standing.60

More widely known around the globe is uncertainty over constitutional inclusion of guarantees respecting some measure of equality of access to basic material needs or to other components of what Rawls calls the social bases of self-respect,61 and subsumes largely under the principle of justice he calls “fair equality of opportunity.”62 It looks impossible to draft such guarantees in terms that are both sufficiently open to local and temporal contingencies of policy choice and sufficiently incisive to satisfy conventional ideas of justiciability (another Goldilocks problem).63 Now the LPOL, it may appear, can succeed at its proceduralizing mission only if we rigorously exclude from constitutional-essential status any guarantee that will not fit comfortably under the conventional justiciability standards. If so, that would mean exclusion from the constitution of a guarantee respecting fair equality of opportunity, never mind if it is (as Rawls affirms) a basic principle of justice. Rawls himself has so concluded, arguing in defense that a constitution limited to conventionally justiciable guarantees can still suffice as a reasonably acceptable starting framework for government. The right collection of such guarantees, he has suggested, can provide assurance enough for reasonable citizens to accept the resulting political framework as the one through which they thenceforward will work to satisfy the more complex and debatable demands of economic and social justice.64

That suggestion, however, runs into substantial and credible resistance—as represented, say, by political philosopher Tommie Shelby. Shelby maintains that a political regime’s lack of an anchoring commitment to the constant and vigorous pursuit of fair equality of opportunity for all can render that regime more than barely or marginally unjust—can render it, indeed, “intolerably” unjust—to some fraction of its citizens.65 That failure, in Shelby’s view, cannot be cured by the regime’s faithful observance of any more narrowly constricted set of constitutional essentials. In sum, by Shelby’s reckoning, a regime that fails of an essential commitment to deliver on fair equality of opportunity thereby forfeits its claim to be a regime reasonably acceptable to all reasonable and rational citizens, and thus (by force of the LPOL) its claim to the loyalties of all who live their lives within its domain.66 A fair-equality guarantee thus becomes a constitutional essential for a justification-bearing constitution; from which it would follow that a Rawlsian supreme court must necessarily be—to that extent, anyway—an activist supreme court, not bound to conventional justiciability norms. Any possible culturally entrenched resistance will have to be overcome.

Now, an activist court, we have already taken pains to say, may at the same time be a weak-form court.67 Perhaps in that direction—activist + weak-form—the Rawlsian salvation lies.

4.2 Weak-Form or Strong-Form Court?

Suppose you are a worker in the field of constitutional law, in a country that does in fact, in the manner of the LPOL, look to a higher-law constitution to serve as that country’s public procedural pact on political justification, and that furthermore looks to courts of law to provide the needed confirmations of constitutional compliance in cases of good-faith public disagreement. And suppose, in those conditions, you also hold to the belief that a fair-opportunity commitment is nevertheless a requirement for any truly justification-worthy state regime. By the argument of Section 4.1, you are, then, in a bind.

A recent explosion of interest by broadly speaking liberal-minded constitutional scholars in varieties of weak-form judicial constitutional review—under names such as “dialogical,”68 “experimentalist,”69 and “catalytic”70—may be understood as a response to the bind.71 A shift of your reviewing court from strong-form to weak-form status obviously eases the pressure for justiciability constraints born of concerns about judicial overreach in a democracy, given that a weak-form court is one whose pronouncements on constitutionality are not binding on the country but rather invite response from other constituted political agencies.

It is not clear, though, how weak-form judicial review can satisfy the proceduralizing ambition of the LPOL. If the constitutional-procedural pact includes designation of a supreme court to serve as its uniquely trusted institutional protector,72 that court’s authority, when acting in that capacity, must be strong-form, not weak-form. To allow other agencies then to override that court’s judgments of constitutional compliance and the major reasons behind them would be to undermine the procedural pact itself.

This seemingly ineluctable deduction may, however, be subject to refinement that can open the door to some varieties, at least, of weak-form review. Weak-form review, we said, means the court’s constitutional-legal pronouncements are rejectable by legislative authorities “in the short run” (meaning not just in the inevitable, glacial way over the long pull of constitutional time and change).73 But the short run need not be instantaneously short. Case-bound determinations of constitutional meanings and applications from undoubtedly strong-form court need not take full and final legal control from the instant of their first utterance.74 System and practice can cater in various ways for a court’s further short-term exchange with a legislature.75 A system can, for example, make allowance for judicial orders of remand to parliament for its further consideration, following upon a court’s initial finding of a constitutional violation but still pending finalization of that finding. What the LPOL’s proceduralizing strategy requires is that the court’s constitutional determinations, once finalized, be treated as controlling system-wide unless and until revised by the court. Perhaps we can find there enough wiggle room—weak-form review, but not too weak (or maybe it is strong-form review but not too strong, either way it is Goldilocks again)—to make social guarantees admissible as essential terms of a justificational load-bearing constitution. If not, then maybe a turn toward judicial tolerance of reasonable disagreement can do the trick.

4.3 Tolerant or Assertive Court?

To what extent should a Rawlsian supreme court—activist and strong-form as it otherwise might have to be—be disposed toward a tolerant treatment of judgments of constitutionality from other political agencies subject to democratic accountability?

Recall from Section 3.2 the general Rawlsian program for judicial constitutional appraisals: In the face of plausible conflicting claims for protection from one or another of the constitution’s essential guarantees, restrict the set of admissible solutions to those aimed at maintaining the overall scheme of liberties in shape conducive to each person’s full and adequate development and exercise of his or her moral powers. An assertive (non-tolerant) court applies that rule by imposing on the country what it (meaning typically a majority of its members) finds to be the singularly best solution of the balance of constitutionally dictated principles and values and condemns any legislative choice not squaring with that. But given the plurality of reasonable liberal conceptions of justice,76 Rawls does not think the set of (“at-least”) reasonable solutions will regularly be restricted to a single member. He expects that often, there will be multiple solutions credibly claiming support from a reasonable balance of public reasons reasonably ascribed to the constitution. By assertively imposing its own “most reasonable” balance of the “most reasonably” applicable reasons, a strong-form court would be writing into the justification-supporting constitutional pact a foreclosure of the contest of reasonable balances, thus running straight against Rawls’s argument for keeping thin and sparse that pact’s substantive guarantees.77 By that argument, it rather seems, a law supportable by any (“at-least”) reasonable balance must be held within a democratic legislature’s power to enact or not as it sees fit.78

That conclusion squares with our definition, near the start, of a tolerant court as one that upholds as constitutional laws that it, for its own part, would judge to be non-compatible, as long as it finds the opposite conclusion to fall within bounds of what is competently defensible.79 A tolerant court thus accepts the view that there may be more than one resolution consistent with good-faith adherence to the constitutional-justificational pact—“not all reasonable balances are the same.”80 It accepts that two or more differing balance-judgments can sometimes all be seen to be reasonable by citizens, and so be seen by them as consistent with a credible commitment to the extant constitutional deal on justification.81 The Rawlsian supreme court, so far, is a tolerant court.

To say so is not to contradict our deduction (in Sections 4.1 and 4.2) that the Rawlsian supreme court is activist and strong-form. A judicial posture of tolerance is easily combinable with both an expansionist (not quiescent) inclination and a strong-form (not weak-form) anticipation of pan-systemic submission to settled judicial pronouncements of constitutional law. A court that speaks widely and often to questions of constitutionality can still, when it speaks, be tolerant of legislative choices that the court finds to reflect a reasonable (whether or not the most reasonable) balance of relevant public values. Judicial opinions upholding as at-least reasonable one or another legislative balance of applicable constitutional values must still contain their statements of the acceptability (vel non) of major and intermediate supporting premises—and those judicial statements, then, can then stand, in strong-form style, as the country’s settled resolutions of the bounds of acceptability of claimed public values and principles, unless and until revised by the court. Thus, even for a court that is activist in its outreach and strong-form in its expectations of submission to its pronouncements, the dimension of tolerance can open some space for interbranch constitutional dialog—perhaps, again, enough space for an allowance of social guarantees as constitutional essentials.82

Our developing picture of a tolerant (if still activist and strong-form) Rawlsian supreme court is not, however, yet complete. It has conveniently been skipping past the Rawlsian program’s first-priority demand to keep intact the central ranges of application of the various constitutional guaranties. Assurance against invasions of these common cores of guarantee sits at the very base of the constitution’s claim to eligibility as a justification procedure that everyone reasonable should find acceptable. We cannot yield on the demand that the bounds of these common cores be publicly settled without fear of upset by tolerated controversy. Granted, these bounds are to align with values that all citizens now alive supposedly, as reasonable and rational, can endorse as terms of the pact.83 And granted, those values can modulate over courses of temporal change and social learning. Still, an at-least moderately firm public settlement of the constitutional-essential cores must be the constant aim, because a procedural pact with core terms unsettled is no pact at all. That argues more toward assertiveness than toward tolerance in the Rawlsian supreme court’s determinations, at least, of the bounds of the central ranges.

So now we have: assertive judicial decision of the bounds of the central ranges, but still tolerant review of legislative balances, taking those central ranges as given. Granted, that is not an easily or transparently practicable program. Take an illustrative case. The constitution, say, contains guarantees respecting liberties of “conscience,” “association,” “expression,” and “religion,” and also of “the equal protection of the laws.” Say further, a newly enacted antidiscrimination law makes it a punishable offense for commercial purveyors of goods and services to vary their terms of service according to the buyer’s sexual orientation. Can this law constitutionally be applied to purveyors of wedding-related goods and services engaged in faith-based refusals to serve or supply same-sex lawful weddings? In this case, which of these would you say is to be the decisive question for the court: Is it to be (i) the locations of the bounds of the central ranges of the liberties respectively claimed by the parties? Or is it to be (ii) which balances can be sustained as at-least reasonable? How can we say which is the decisive question before the court hands down its opinion in the case? It seems the upshot must be that, for the strong-form Rawlsian supreme court, impulses for both assertiveness and tolerance are locked in close dialectical embrace, which only skill and sensitivity can sort out from cases to case and context and context.84

Such thoughts do not yet drive Rawlsian aspiration from the field. Here, in brief, is some of what I think Rawls might say in response. First, in the nearly well-ordered constitutional-democratic societies that Rawls has in view, divisive disagreements over the bounds of the central ranges should be rare. Understandings in regard to them are cemented, he would say, into the broad historical culture of constitutional democracy.85 Second, to the extent that common understandings of the central ranges remain fluid and shifting over time, they shift in response to social and cultural conditions in which judges and citizens alike will all be living their lives. Owing to these first two factors, most of the hard cases, in which opinions will seriously divide, should fall into the marginal-not-core space of tolerant judicial review.86 And then, third, the debates over reasonable balances of relevant public values, taking place in that “tolerant” space, can make their own contribution toward consolidations of evolving public views of the central cores.87 Perhaps the currently pending American debates and decisions in our cases of the wedding-supply purveyors may provide a test for that proposition.

4.4 Summing Up

As predicted, an approach to judicial restraint from the standpoint of justification-by-constitution leads toward a dis-bundling of dimensions of quiescence, weak-form-ism, and tolerance that a sole focus on democracy or majority rule tends to run together. According to our deductions here, once potentially divisive disagreements over constitutional-essential applications have broken out in public, a Rawlsian supreme court will need, in order to fulfill its role in the constitution-centered program of justification, to be activist and strong-form, but also noticeably tolerant—if only up to a point. “Dis-bundled” does not mean, though, that any of these deductions occurs with no attention paid to the others. We have just been noticing, after all, how judicial tolerance for reasonable disagreements over constitutionality can serve as an emollient for the counter-majoritarian thrust of judicial-supremacist activism, which might otherwise be found unacceptable in a democracy.


  1. 1.

    Compare Michelman 2015, pp. 184–185 (differentiating a “normative” from a “legitimation” function for constitutional law); Michelman 2018b, pp. 77–79 (differentiating a “regulatory” from a “proceduralizing” function). In what follows, I will be drawing freely from these and other papers in the series, as found in the References and cited below.

  2. 2.

    See, e.g., Hirschl 2004, pp. 43–47 (advancing a “hegemonic preservation” thesis to explain constitutionalization of substantive rights); Grimm 2005, pp. 193, 195 (advancing the thesis of an affectively “integrative” function for substantive constitutional law).

  3. 3.

    See Michelman 2017, pp. 44–50.

  4. 4.

    See Michelman 2018a.

  5. 5.

    See Michelman 2018b.

  6. 6.

    See Michelman 2018d.

  7. 7.

    See Michelman 2008, 2012, 2015.

  8. 8.

    See Michelman 2015, pp. 196–97.

  9. 9.

    See ibid., pp. 197–200.

  10. 10.

    See Michelman 2016, pp. 646–647; Michelman 2018c, pp. 758–762.

  11. 11.

    See Michelman 2018d, pp. 126–28.

  12. 12.

    See Michelman 2016, pp. 647–648. I here follow Alessandro Ferrara, who has named as “hyperpluralist” a social condition increasingly found in the constitutional-democratic west, of a visible and insistent presence of populations who do not at all or do not fully include themselves or their ways within the historical tradition of constitutional democracy, with resultant pressure on those countries to seek out “a form of democracy that does not amount to a Westernization.” Ferrara 2014, p. 17.

  13. 13.

    See Friedmann 2016, pp. 54–56 (summarizing aspects of a perceived activism of Israel’s supreme court in the latter part of the twentieth century).

  14. 14.

    See Tushnet 2006, pp. 2, 10. Tushnet takes as one leading example courts under the UK Human Rights Act, 1998, which authorizes judicial declarations of a statute’s non-compatibility with stated human rights norms while also authorizing parliamentary declination to abide by such declarations. See id., p. 7.

  15. 15.

    See Tushnet 2006, p. 1 (equating strong-form review with “judicial supremacy ... in which the courts have the final and unrevisable word on what the constitution means”).

  16. 16.

    Tushnet 2006.

  17. 17.

    Here, I may depart a bit from Tushnet, who has classed such long-term developments as a variation on weak-form review. See Tushnet 2006, pp. 16–20.

  18. 18.

    Bickel 1962, pp. 16–23 (on “the counter-majoritarian difficulty”).

  19. 19.

    Ibid., p. 18.

  20. 20.

    For a thorough and incisive recapitulation of the debates, see Sultany 2012.

  21. 21.

    Rawls 1993, p. 233.

  22. 22.

    Ibid., p. 231.

  23. 23.

    For my own thoughts in that direction, see Michelman (2018d).

  24. 24.

    Rawls (1993), p. xx (emphasis supplied).

  25. 25.

    Ibid. xviii-xix.

  26. 26.

    See Rawls 1993, pp. 136–137, 216–217; Rawls 2001, pp. 40–41 (all posing the same question in essentially identical terms).

  27. 27.

    Rawls 1993, p. 217. See also ibid., p. 237; Rawls 2001, p. 41.

  28. 28.

    We need not agonize here over exactly what is signified by “reasonable.” It is enough to say that Rawlsian reasonable citizens all share perceptions, first, of the very great moral and practical benefits to everyone of having some decent system of law effectively in force; second, of the persisting facts of conflicts of interests and moral disagreements that might be humanly understandable on all sides; and then third, of the commanding moral logic of a reciprocity of respect for each person’s quest for a life lived in dignity, according to aims and values that a person affirms for herself or himself in conditions of freedom.

  29. 29.

    The texts so plainly indicate. See Michelman 2018a, pp. 386–387.

  30. 30.

    As to whether (or in what sense) the constitution must be a “written” one, see Michelman 2018a.

  31. 31.

    Rawls’s “essential” terms for a justification-supportive constitution are of “two kinds,” comprising provisions both for “the ... structure of government and the political process” and for “equal basic rights and liberties of citizenship that legislative majorities are to respect.” Rawls 1993, p. 227.

  32. 32.

    Rawls 1971, pp. 197–198. For more extended treatment of the procedural character of the LPOL solution to the problem of political liberalism, see Michelman (2018a), pp. 384–386.

  33. 33.

    See Rawls 1993, p. 232 (“The principled expression of higher law is to be widely supported,” and so “it is best not to burden it with many details and qualifications.”).

  34. 34.

    Rawls 1993, p. 295.

  35. 35.

    Ibid., pp. 291, 333.

  36. 36.

    Rawls 1993, p. 295.

  37. 37.

    Ibid., pp. 297–298.

  38. 38.

    Ibid., p. 296.

  39. 39.

    See Rawls 1997, pp. 804–805 (“[C]onflicts arising from the burdens of judgment always ... limit the extent of possible agreement.”).

  40. 40.

    For examples, see Sections 3.4 and 4.3.

  41. 41.

    See Rawls 1993, p. 237.

  42. 42.

    See ibid., p. 234.

  43. 43.

    See Langvatn 2016, for more extended discussion.

  44. 44.

    Rawls 1993, p. lviii.

  45. 45.

    Ibid., pp. lvix–lvx.

  46. 46.

    See ibid., p. lvix. See also, on these same points, Rawls 1997, pp. 773–775.

  47. 47.

    See Section 3.2.

  48. 48.

    See Rawls 1993, pp. 297–298.

  49. 49.

    See Section 3.2.

  50. 50.

    See Rawls 1993, pp. 293, 297.

  51. 51.

    Ibid., p. 298.

  52. 52.

    See ibid. See generally Edmundson (2017).

  53. 53.

    See Rawls 1993, pp. 18–19, 29, 34; Rawls 1997, pp. 799–800; Michelman 1994, pp. 1815–1816.

  54. 54.

    See Rawls (1996), xlviii-xlix.

  55. 55.

    See Section 3.3.

  56. 56.

    See, e.g., Dworkin 1986, pp. 265–266.

  57. 57.

    See Fallon 2006.

  58. 58.

    This expression designates a calculated shaping of electoral districts for local, state, and federal legislative bodies so as to give artificial advantage to a favored political party, allowing that party although outvoted nevertheless to gain and keep control of the legislative body in question.

  59. 59.

    See Vieth v. Jubelirer, 541 US 267, 305–306 (2004) (plurality opinion).

  60. 60.

    See Gill v. Whitford, 585 US __ (2018).

  61. 61.

    See Rawls 1971, pp. 76, 83; Rawls 2001, pp. 58–59.

  62. 62.

    See Rawls 1971, p. 73, explaining fair equality of opportunity as follows: “In all sectors of society there should be roughly equal prospects of culture and achievement for everyone similarly motivated and endowed. The expectations of those with the same abilities and aspirations should not be affected by their social class.”

  63. 63.

    See Michelman 2015, pp. 194, 197–198, for full discussion.

  64. 64.

    See Rawls 1991, pp. 227–230; Michelman 2012, pp. 1016–1017.

  65. 65.

    Shelby 2007, pp. 145–146.

  66. 66.

    See Michelman 2012 (examining Rawls’s exclusion from the constitutional essentials of a principle of fair equality of opportunity); Shelby 2007, pp. 145–151 (finding such an exclusion unacceptable for a justification-bearing liberal constitution).

  67. 67.

    See Section 2.1.3.

  68. 68.

    See Dixon 2007.

  69. 69.

    See Gerstenberg 2012.

  70. 70.

    See Young 2012, pp. 167–191 (2012).

  71. 71.

    See Michelman 2015 for the argument in extended form. For general confirmation, see Tushnet 2008, pp. 18–42.

  72. 72.

    See Section 3.3.

  73. 73.

    See Section 2.1.3.

  74. 74.

    Abraham Lincoln famously took this view. See Lincoln 1857.

  75. 75.

    See Dixon 2007; Young 2012.

  76. 76.

    See Section 3.4.

  77. 77.

    See Section 3.2.

  78. 78.

    See Rawls (1997), p. 770. While citizens when casting votes are expected to vote in favor of orderings of political values they judge the most reasonable, see ibid., 773, 797, they are also expected to respect outcomes that they can find at least reasonable, see ibid. 770.

  79. 79.

    See Section 2.1

  80. 80.

    Rawls (1993), p. 253.

  81. 81.


  82. 82.

    See Michelman 2015 for affirmation of this possibility.

  83. 83.

    See Rawls 1993, p. 236 (explaining that courts engaged in constitutional interpretation “must appeal to ... values that they believe in good faith ... that all citizens as reasonable and rational might reasonably be expected to endorse”).

  84. 84.

    Mark Tushnet’s views appear similar. See Tushnet 2006.

  85. 85.

    See Rawls 1993, pp. 8, 14, 186; Langvatn (2016), p. 143.

  86. 86.

    See Tushnet 2006), p. 1.

  87. 87.

    See Rawls 1997, p. 799.


  1. Bickel A (1962) The least dangerous branch: the supreme court at the bar of politics. Bobbs-Merrill, IndianapolisGoogle Scholar
  2. Dixon R (2007) Creating dialogue about socioeconomic rights: strong-form versus weak-form judicial review reconsidered. I-CON 5:391–418Google Scholar
  3. Dworkin R (1986) Law’s empire. Harvard, CambridgeGoogle Scholar
  4. Edmundson W (2017) John Rawls: reticent socialist. Cambridge, CambridgeCrossRefGoogle Scholar
  5. Fallon R (2006) Judicially manageable standards and constitutional meaning. Harv LR 119(2006):1274Google Scholar
  6. Ferrara A (2014) The democratic horizon: hyperpluralism and the renewal of political liberalism. Cambridge, CambridgeCrossRefGoogle Scholar
  7. Friedmann D (2016) The purse and the sword: the trials of Israel’s legal revolution. Oxford, OxfordCrossRefGoogle Scholar
  8. Gerstenberg O (2012) Negative/positive constitutionalism, ‘fair balance,’ and the problem of justiciability. I-CON 10:904–925Google Scholar
  9. Grimm D (2005) Integration by constitution. I-CON 3:193–208Google Scholar
  10. Hirschl R (2004) Towards Juristocracy: the origins and consequences of the new constitutionalism. Harvard, CambridgeGoogle Scholar
  11. Langvatn S (2016) Legitimate but unjust; just, but illegitimate: Rawls on political legitimacy. Phil & Soc Crit 42:132–153CrossRefGoogle Scholar
  12. Lincoln A (1857) Speech on the Dred Scott decision at Springfield, Illinois. In: Fehrenbacher D (ed) (1989)Abraham Lincoln Speeches and Writings 1832-1858. Library of America, New York, pp 390–403Google Scholar
  13. Michelman F (1994) The subject of liberalism. Stanford LR 46:1807–1833CrossRefGoogle Scholar
  14. Michelman F (2008) Socioeconomic rights in constitutional law: explaining America away. I-CON 6:6633–6686Google Scholar
  15. Michelman F (2012) Poverty in liberalism: a comment on the constitutional essentials. Drake LR 60:1001–1021Google Scholar
  16. Michelman F (2015) Legitimacy, the social turn and constitutional review: what political liberalism suggests. Krit-V 98(3):183–205CrossRefGoogle Scholar
  17. Michelman F (2016) A constitutional horizon. Phil & Soc Crit 42:640–648CrossRefGoogle Scholar
  18. Michelman F (2017) Proportionality outside the courts with special reference to popular and political constitutionalism. In: Jackson V, Tushnet M (eds) Proportionality: new frontiers, new challenges. Cambridge, CambridgeGoogle Scholar
  19. Michelman F (2018a) “Constitution (written or unwritten)”: legitimacy and legality in the thought of John Rawls. Ratio Juris 31:379–394CrossRefGoogle Scholar
  20. Michelman F (2018b) Human rights and constitutional rights: a proceduralizing function for substantive constitutional law. In: Voeneky S, Neuman G (eds) Human rights, democracy, and legitimacy in a world of disorder. Cambridge, Cambridge, pp 72–96Google Scholar
  21. Michelman F (2018c) Israel’s constitutional revolution: a thought from political liberalism. Theor Inq Law 19:745–766CrossRefGoogle Scholar
  22. Michelman F (2018d) Book Review: The Framers' Coup. Const Comm 33:109–28Google Scholar
  23. Rawls J (1971) A theory of justice. Harvard, CambridgeGoogle Scholar
  24. Rawls J (1993) Political liberalism. Columbia, New YorkGoogle Scholar
  25. Rawls J (1996) Political liberalism. Columbia, New YorkGoogle Scholar
  26. Rawls J (1997) The idea of public reason revisited. U Chi LR 64:765–807CrossRefGoogle Scholar
  27. Rawls J (2001) Justice as fairness: a restatement. Ed. In: Erin Kelly. Harvard, CambridgeGoogle Scholar
  28. Shelby T (2007) Justice, deviance, and the dark ghetto. Phil & Pub Aff 35:126–160CrossRefGoogle Scholar
  29. Sultany N (2012) The state of progressive constitutional theory: the paradox of constitutional democracy and the project of political justification. Harv Civ R-Civ L LR 47:471–455Google Scholar
  30. Tushnet M (2006) Weak-form review and civil liberties. Harv Civ R-Civ Lib Law Rev 41(2006):1–22Google Scholar
  31. Tushnet M (2008) Weak courts, strong rights: judicial review and social welfare rights in comparative constitutional law. Princeton, PrincetonGoogle Scholar
  32. Young K (2012) Constituting economic and social rights. Oxford, OxfordCrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Robert Walmsley University, Emeritus, Harvard Law SchoolHarvard UniversityCambridgeUSA

Personalised recommendations