I. Institutionally-hard cases

Cases that are easy or straightforward as a matter of ideal constitutional principle may still be ‘hard’ due to non-ideal social and political circumstances. While this predicament is widely familiar, it is insufficiently understood. In particular, when liberal constitutional courts face strong opposition from other branches of government or significant segments of the public, adjudication gives rise to a fundamental dilemma of constitutionalism, one that the main lines of existing constitutional theory do not adequately diagnose, much less address.

Consider for example the situation facing the Federal Constitutional Court of Germany in 1995, when it had to rule on the constitutionality of the state requiring crucifixes to be displayed in public schools.^{Footnote 1} From the vantage point of constitutional principle, this was arguably a simple case of religious freedom protected by the Basic Law, requiring the invalidation of the regulation.^{Footnote 2} However, the judges must have known that a decision to that effect would be highly unpopular with those whom it directly affected: public school boards in culturally-Catholic Bavaria. The Court nevertheless went ahead and declared void the relevant portion of the Bavarian elementary school regulation, and lived to face the consequences: mounting anger, politicians openly defying the Court, and the promulgation of new regulations and practices in blatant disregard of the decision. It is often said that there are more crucifixes in Bavarian schools now than there were before Crucifix was decided.^{Footnote 3}

Constitutional judges care about how their decisions are received and enforced, their social impact, and their effect on the court’s own status and perceived legitimacy. The Chief Justice of the German constitutional court in Crucifix was evidently concerned enough to write an article in a daily newspaper, explaining to the public ‘why a judicial ruling deserves respect’.^{Footnote 4} What made this case hard for the court – and interesting for us – was not a difficulty in interpreting or elaborating constitutional norms. Rather, it was the social and political circumstances – circumstances we may say were non-ideal – that made this decision institutionally hard. In ideal circumstances, a constitutional court would see its decisions enforced, respected, and lived by – even if they are not universally thought to be ‘right’.

An unpopular decision should not, in and of itself, translate into a concern with legitimacy and non-compliance. Non-ideal circumstances pertain to the actual position that constitutional values and institutions occupy in the social and political landscape of a particular polity, and to the role and status of the court as a ‘forum of principle’^{Footnote 5} for that polity, relations that are often more fragile and volatile than they appear in ideal images and frameworks of how constitutional institutions should act and interact. If constitutional judges lose sleep over cases such as Crucifix, this is not – or not primarily – due to difficulties in getting to the right answer^{Footnote 6} as a matter of constitutional principle, but to the foreboding sense that a right-but-unpopular decision could be detrimental to the Court’s status, to the constitutional system, and to the normative force that liberal constitutional values have within the actual practices of social life. The dilemma of institutionally-hard cases arises, then, where there is a significant tension or conflict between what the court would hold to be right constitutionally (in ideal circumstances) and what seems wise or prudent institutionally, given the actually existing non-ideal circumstances.

The first objection to this characterisation is that this should not be seen as a real dilemma: a Herculean judge should stand by what is constitutionally right, and not lose any sleep over the institutional implications of her decisions. And isn’t this obvious? Is it not the very point of constitutional adjudication, the very essence of the rule of law, that the court be independent from all forms of public and political pressure, and defend constitutional rights against such interests and sentiments? On this familiar understanding of constitutional legality, the German court was not faced with a dilemma, but did the right thing, plain and simple. A constitutional court will, by definition, not always be popular; so what? Herculean integrity should not be tainted by approval-seeking and savvy public relations.

But against this uncompromising commitment to constitutional principle, which undergirds official discourse in both judicial opinions and constitutional theory, an opposing ideal often comes to the fore in more informal discourse, one that accepts the importance of judges taking seriously the social and political circumstances in which they make their pronouncements. This is the notion that good judges are good statesmen, their role calling for delicate, situational judgments concerning the timing and scope of their pronouncements of principle. Resonating Bickel’s avoidance tactics,^{Footnote 7} this undercurrent is perhaps given clearest expression in cases where a decision is said to ‘mature’. Consider the celebrated same-sex marriage decision of the US Supreme Court in 2015, Obergefell, where the Court recognised a constitutional right to marriage equality.^{Footnote 8} The ruling came after many state and federal courts had already held that same-sex marriage was guaranteed by the Fourteenth Amendment’s Equal Protection clause, decisions supported by a steady shift in public attitudes.^{Footnote 9} Indeed, the line of Supreme Court cases preceding Obergefell has been praised as masterful ‘judicial statesmanship’ precisely for its intelligent timing vis-à-vis public opinion, which carefully avoided the right outcome until it was ‘due’.^{Footnote 10} Does this mean it would have been wrong for a federal court to issue such a decision earlier, say 20 years ago, because it would have been unpopular then?^{Footnote 11} Judge Posner clearly thought so when, in 1997, while in favour of marriage equality on the merits, he argued for letting ‘the matter simmer for a while before the heavy artillery of constitutional rights making is trundled out’.^{Footnote 12}

Here, then, is the issue. Had a principled but ‘premature’ decision imposed from the bench a constitutional norm that most would have found unacceptable, this could have ensued in turmoil and deeper entrenchment in currently-held positions – effects that would have harmed not only the Court but also the cause of marriage equality.^{Footnote 13} But how can a ‘forum of principle’ plausibly be affected by conservative public attitudes to delay and deny constitutionally-protected fundamental rights? The ideal of constitutional legality cannot tolerate a judicial practice of selective or strategic restraint, where constitutional rights are made to hinge on contingent majoritarian sentiment. And yet, for a constitutional system to be sustainable, it is evidently undesirable that judges be utterly disconnected from context and impervious to circumstances that could affect its institutional status and social impact.

Indeed, a recent ‘realist turn’^{Footnote 14} in political science literature on judicial behaviour contends that judges not only consider context, including the attitudes of other actors, but in fact that judges strategise, in the rational-choice sense of the term.^{Footnote 15} That is, judicial outcomes reflect judges’ calculation of expected results, with a view to maximising their preferred policy outcomes. This approach is diametrically opposed to ideal constitutional theory: sceptical of the normative force of principled constitutional reasoning, it sees judicial opinions as ex post facto rationalisations of the strategically best decision. Yet, like ideal constitutional theory, the idea that judges strategise also fails to recognise the dilemma of institutionally-hard cases, now for the mirror-opposite reason. Since judicial strategy is a rational calculation, it is free from normative grey areas. It simply takes into account the various possible outcomes and their effects on the relevant agents, perhaps aided by some game-theoretical sophistication, to maximise the preferred outcome. In the absence of principled normative constraints, there is no dirty-hands insomnia.

We have, then, three competing approaches to institutionally-hard cases, two of which deny the dilemma and a third that recognises it but fails to address it. Ideal constitutional theory ignores the non-ideal or, as we will see, idealises it, folding it into generalised principles and doctrines. Judicial strategy, propelled in part by ideal theory’s disconnect from social and judicial realities, does the inverse: it neutralises the force of the ideal, leaving judges free to calculate. Finally, judicial statesmanship recognises institutionally-hard cases as truly hard and praises those judges that handle them wisely. But this approach offers no standards by which to guide and assess institutionally-hard decisions, treating their adjudication as an art form that reflects the personality and knack of individual judges. Thinking in terms of judicial statesmanship leaves us with intuitive appreciation or dismay: any one decision may be considered brave or imprudent, cowardly or cautious, depending on the eyes of the beholder. The combined upshot of these three distinct approaches is a lack of critical yardstick by which to evaluate decisions in institutionally-hard cases, as better or worse decisions in terms of explicit constitutional principle and reflective practice.^{Footnote 16}

To appreciate the far-reaching significance of the problem, consider the predicament of liberal constitutional courts facing political threats from executive and legislative branches that seek to weaken them and strengthen majoritarian institutions.^{Footnote 17} Should a court that responds by toning down its decisions and adopting a strategy of restraint be praised for its shrewd pragmatism, or should we instead demand that it bravely stay the course, regardless of retaliatory consequences, even if these might amount to the demise of the constitutional system? One such example is Israel’s High Court of Justice, which has come under escalating threats of constitutional reform, and appears to have heeded the call.^{Footnote 18} Consider for instance its decision from April 2015, upholding the Boycott Law enacted by the Knesset in 2011.^{Footnote 19} The law imposes civil liability on anyone who publically calls to boycott Israel or the Settlements, be it economically, academically or culturally.^{Footnote 20} As a matter of constitutional law, it would have seemed an easy decision to declare the law a violation of protected political speech. And, indeed, to defend the opposite outcome, the HCJ’s opinion relies on strikingly unpersuasive stretches of constitutional reasoning, including the claim that public calls in support of a boycott do not fall within the core area of protected speech,^{Footnote 21} and a comparison of such advocacy with false cries of ‘fire’ in a packed theatre.^{Footnote 22} The question before us, however, is not whether the decision was tenable in terms of principle but rather, first, whether it is useful to think of this as an institutionally-hard case – a case raising a dilemma between what is constitutionally right and what seems to be institutionally wise or prudent – and, second, supposing that it is useful to see the case as institutionally-hard, whether we can talk meaningfully about the way in which non-ideal circumstances may or may not permissibly affect the court’s decision.

By contrast with the descriptive orientation of the ‘judicial strategy’ school, this article is concerned with a normative question: how should judges decide institutionally-hard cases? Granted, the significance of the inquiry does rest on an empirical premise, but a rather modest one: namely, that, at least in some actual cases, courts do shy away from rendering the best possible constitutional decision – one that they would provide in ideal circumstances – due to non-ideal institutional pressure. Whether or not such non-ideal considerations did in fact play a role in any one case, such as Boycott, is of less importance here. What is required for this inquiry to be valuable is a reasonable likelihood that such considerations play some role in some decisions, although (indeed, especially where) this is not explicitly articulated in the written opinions, or even in internal deliberations. Thus, the court in Boycott said nothing of the escalating threats from members of Israel’s coalition government, citing instead standard legal-constitutional justifications, a gamut of authorities, an ethos of ‘restraint’, and a balancing test that found the chilling effects on political speech to be proportionately outweighed by a legitimate state interest. Yet astute observers of the HCJ in the last few years have lamented its ‘cowardice’ and its scaling-down of constitutionally protected rights.^{Footnote 23} It is easy to see, but unnecessary to prove, that non-ideal circumstances have had a chilling effect on the HCJ.

To properly address institutionally-hard cases, we need, first, to recognise that they do present a dilemma, and that constitutional courts hold a dual position: on the one hand, they are a forum of principle – I will argue for this being their primary role. On the other hand, they are and should be concerned with the meaningful and sustainable realisation of constitutional ideals. Second, we should then extend constitutional theory to consider the dilemma of institutionally-hard cases as amenable to normative reflection, and seek to prescribe what kinds of non-ideal circumstances should or should not affect constitutional decisions, and how. If the right answer as a matter of principle is not always the right answer all things considered, constitutional theory should subject this intuition to reasoned reflection. This would be the only way to maintain the commitment to principle while taking reality seriously. In other words, to take seriously the reality of non-ideal circumstances, we need to develop a non-ideal constitutional theory. This would be the constitutional branch of the Rawlsian notion of non-ideal theory. Rawls spoke of non-ideal theory as charting a legitimate and effective path for shifting from a current – unjust – society, to one where justice is realised.^{Footnote 24} In the context of constitutional adjudication, a non-ideal theory should chart the legitimate and effective path toward a meaningful and sustainable realisation of constitutional ideals. The concern with perceived legitimacy and the status of the court is then not seen as a will to power, nor as the maximisation of preferred outcomes, but rather as an instrument for realising constitutional ideals – the pronouncement and development of which is the court’s primary commitment. The last part of this article proposes a first sketch for such a non-ideal theory. Preceding it is a closer examination of the alternative approaches already mentioned: ideal constitutional theory, judicial strategy, and judicial statesmanship. Each of these, though unsatisfactory, is crucial for developing the inquiry, informing the proposed theory, and evaluating it critically.

The problem of institutionally-hard cases is a general one, emerging from the very structure of constitutional democracies and, therefore, not limited to one jurisdiction or another. To highlight this, I will continue to consider cases from different courts: the US Supreme Court, the German Constitutional Court, Israel’s High Court of Justice, and the European Court of Human Rights. But, first, what is wrong with idealisation?

II. Ideals and idealisation

The prevailing constitutional discourse of both judges and scholars is inherently idealised, and its approach to non-ideal circumstances is to reject their relevance, deny it and/or subsume it under ideal doctrines and principles. While this judicial commitment appears necessary for defending the substance of constitutional principles and the court as a principled and independent institution, the categorical denial of the non-ideal comes at a heavy price: as non-ideal decisions are idealised, they create diluted or eroded ideal precedent as a matter of both doctrine (first-order) and theory (second-order). This kind of idealisation should be troubling to any committed constitutionalist, much more so than a simple concern with a lack of ‘realism’ or ‘transparency’ in judicial opinions, in the sense of their failure to reveal the ‘real’ reasons behind them.

III. Judicial strategy, or: What’s wrong with ‘realism’

One response to the problem of institutionally-hard cases is that judges should be strategic about their decisions. What might this entail, and is it desirable? An increasingly influential approach in political science literature contends that judges already are – descriptively – strategic actors, in the rational-choice sense of the term.^{Footnote 34} In sharp opposition to ideal constitutional theory, scholars that take the ‘judicial strategy’ approach examine judicial ‘behaviour’ and take pride in the ‘realism’ of their findings. Such realism is seen as the antidote to constitutional theory which, for the most part, ‘grossly misrepresents reality’, for it ‘sticks to an idealist notion of constitutional law as a politically unencumbered sovereign virtue’ and ‘unreservedly portrays constitutional courts as a ‘‘forum of principle’’’.^{Footnote 35} However, important as it is to demystify courts and adjudication, the notion of strategy cannot provide a satisfactory response to the normative problem identified here. This is because, at least in its technical version, this notion denies that the problem exists.^{Footnote 36} Premised on a radical scepticism of legal reasoning, ‘judicial strategy’ implies abandoning the premise of the constitutional courts being a forum of principle and, with it, the notion of institutionally-hard cases. If, as its proponents claim, judges act with the sole aim of maximising their given policy preferences, the cases that we have considered simply exemplify complex strategic scenarios, rather than a dilemma of constitutionalism.

Constitutional theory does have to reckon with the bold claims of the ‘judicial strategy’ school which threaten to undercut the very foundations of constitutionalism. Rather than contest their empirical accuracy as such, however, here I will discuss the theory’s implausible jurisprudential premises and its unacceptable normative implications. Yet it is also important to see how, despite these shortcomings, the frameworks developed by judicial behaviour scholarship should inform the project of non-ideal theory, for they highlight the kinds of considerations – other than constitutional principle – that scholars and judges take to be significant in constitutional adjudication, and examine closely the institutional dynamics that courts participate in. If these considerations and dynamics could be made to assume their proper non-ideal place in constitutional theory, normative legal scholarship would become more connected with, and relevant to, constitutional practice.

By contrast with theories of ‘adjudication’, theories of ‘judicial behaviour’ take a self-conscious distance both from legal or constitutional doctrine and from normative prescriptions.^{Footnote 37} At the same time, however, the emergence of this field in political science rests on developments in twentieth-century legal thought: it was the insights of legal realism, as these political scientists perceive it, that rendered formalism obsolete as both a descriptive and a prescriptive enterprise, and which made possible non-legal and non-normative, but rather empirical, social-scientific, inquiry into judicial decision patterns. When political scientists in the 1960s and 1970s recognised that judges’ behaviour could not plausibly be described in terms merely of discovering and applying existing laws, they began developing what is now described as the ‘attitudinal’ or ‘social-psychological’ approach to judicial behaviour, one that sees judges as single-mindedly pursuing their own individual policy goals.^{Footnote 38}

The ‘strategic’ approach to judicial behaviour has more recently emerged as a development that, to some extent, supersedes the ‘attitudinal’ approach.^{Footnote 39} The judicial strategy line of research applies assumptions of rational behaviour and game-theoretical models, to argue that judges do not single-mindedly pursue the dictates of their own policy views, but rather, taking into account the positions and expected behaviour of other players (within the court and outside of it), engage in conduct calculated to maximise the attainment of their goals over the longer run, given these constraints and expectations. Thus, a judge in favour of a particular policy – say, allowing religious symbols in public classrooms – might not decide in direct correspondence with this policy-orientation in a particular case, if she expects this to be bad strategy for maximising this and other preferred outcomes in the longer run. The calculation would include her prediction of how her colleagues on the bench and other branches of government, as well as the public, would react. This rational-choice approach to public law adjudication claims to carry significant predictive and explanatory force of judicial behaviour, and has gained significant influence in the last few decades, although some continue to argue for the ‘attitudinal’ model as being better borne out by the data.^{Footnote 40}

For our purposes, it is the ‘judicial strategy’ model that is of greater interest and relevance, for it claims precisely to take into account how courts respond to various actors. The increasing influence of this model has spurred various refinements that seek to modify and improve its basic tenets, often by considering particular courts and their practices, identifying the end-goals of particular judicial strategic behaviour, and fine-tuning the detailed dynamics of legal rationalisation and strategising. Some works depart from the technical meaning of ‘strategy’, while others seek to elaborate it.^{Footnote 41} Our primary focus here will be the basic tenets of the rational-choice approach, postponing until later certain refinements, which inch notions of strategy closer either to what is discussed below as ‘judicial statesmanship’ or to the ‘principled strategy’ version of non-ideal theory.^{Footnote 42}

From the point of view of legal and constitutional theory, the first difficulty with the rational-choice approach to adjudication is its provocative claim that legal principle is normatively meaningless in the judicial decision. The judges’ agenda are their pre-given and fixed policy ‘preferences’. This problem is, to some extent, discussed in the literature, and some authors have sought to fold in legal normativity into the model, by bringing to the fore the prudence of courts giving the impression that they decide according to legal norms. In other words, since there is an expectation that judges interpret the law, and this expectation is the basis for the acceptance of the court’s authority, these models posit a strategising judge that maintains this perception and, with it, the legitimacy of her decisions, by creating a façade of interpreting ‘the law’.^{Footnote 43} That is, the concern – as understood by these scholars – is to explain why judges talk about legal norms as significant for deciding the case, while in fact they are not significant.

This is highly unpersuasive from a jurisprudential point of view. Brandishing a critical-cum-cynical perspective on legal norms, the ‘strategy’ approach rests on a crude version of legal realism, and thereby fails to account for the complex phenomenology of the process of normative reasoning and the interaction between legal materials, deliberation and reason-giving.^{Footnote 44} While some judges may have a clear policy preference going into a legal decision, other judges may seek to ‘find out’ what the legal materials most plausibly ‘require’. Even in the former case, a strategically-minded judge hardly ever experiences the field of legal materials as a free-for-all playground that can be walked through to get to the other end. Political scientists that are wedded to the ‘strategic’ model take the lesson of legal realism to mean the radical indeterminacy of the legal materials which, in turn, they take to mean that judges can extract from the legal materials any outcome they wish, in every case, effortlessly and costlessly.^{Footnote 45}

Any practising lawyer or judge should not find this convincing. For our purposes, of seeking an evaluative and prescriptive take on the dilemma of institutionally-hard cases, the picture of the practice of adjudication should be truer to the experience of real judges, and to what they (and we) perceive as the operating demands, possibilities, and constraints of law. A much more persuasive account of the real practice of adjudication accepts the ‘strategising’ (in a non-technical sense) judge – a judge whose mission is to get out of the materials what he or she independently believes to be the best policy – and then elucidates how the work of strategic legal interpretation calls for ingenuity, dedication, time and resources, which are not present in equal measure across cases.^{Footnote 46} Such a phenomenology also speaks to how legal work – even with all resources amply present – does not always ‘yield’ the desired outcome. Finally – and crucially – this account needs also to acknowledge the ‘normative force of the field’, namely, how the legal materials regularly affect even the most strategically-minded judge in her perception of what is ‘right’, not only of what is ‘possible’.^{Footnote 47}

This descriptive/phenomenological failure also renders the ‘strategy’ model deficient as a basis for an evaluative project. The model does not ‘fit’ with the aspirations of the system, as participants see them and as they should indeed be seen: aspirations of giving content to our most fundamental political values through constitutional reasoning. Indeed, political scientists do not presume that judges should take and apply this model.^{Footnote 48} But while no one thinks to tell judges, as a matter of political morality, that they should strategise regardless of the position they are strategising for, any prescriptive theory would ultimately have to deal with the more persuasive of these political-scientific accounts.

IV. Judicial statesmanship: Anti-theory

While the rational-choice notion of judicial strategy is unlikely to win many adherents among normatively-committed constitutional theorists, a softer, informal ideal of judicial statesmanship lives in the shadows of ideal constitutional discourse and offers a respite to the adherents of its harsh demands. This common-sense position recognises that institutionally-hard cases are indeed (i.e., should be) hard, and affirms the desirability of judges looking behind the veil of ideal constitutional theory. In combination, this allows everyone to praise judges that seem to handle hard cases particularly well, all the while leaving both judges and scholars to idealise the non-ideal. The notion of ‘judicial statesmanship’ therefore remains unreflective and lacks critical bite. Considering the adjudication of institutionally-hard cases as an art form (we just know a good judge when we see one), the notion of judicial statesmanship may too easily turn (as with idealisation) into blanket Bickelian support for the exercise of ‘restraint’. In its more explicit variants, this position typically rests on conscious resistance to theorising the non-ideal. And while such resistance is understandable, it is also ultimately deeply unsatisfactory as a matter of constitutional aspiration, and our task is to overcome it.

The notion of judicial statesmanship thrives on the distinction between theory and practice. While law is, theoretically, about principles, matters of institutional, social and political prudence are a practical art. Great judges are master statesmen, for they ‘just know’ how to issue the institutionally-wise decision for the time and place in which they operate, sensing what they can get away with given the prevailing social and political climate. The decisions issued are, then, not just ‘right’ or ‘just’ as a matter of principle, but are good decisions all things considered. That this skill cannot be formulated into principles is indeed what makes institutionally-hard cases the kind of cases where truly great judges shine – judges like Chief Justice Marshall or Aharon Barak. The reason why these giants of constitutionalism shine is their grandness as statesmen, not simply as jurists. This recurring, deep-seated image in our political culture sustains, in its least critical version, a myth of judges as especially wise and benevolent. Like Aristotle’s golden mean, the image is of a virtuous and courageous statesman that masterfully draws the right line between recklessness and cowardice. By the same token, however, identifying which judges are good statesmen is also, alas, not amenable to well-reasoned justification. We only ‘know it when we see it’, and some judges are clearly better than others. Chief Justice Marshall was one such judge, Chief Justice Taney was not.^{Footnote 49}

A perfect early articulation of this view is offered by Tocqueville’s description of the justices of the US Supreme Court, in a famous passage brimming with heroic metaphor:

On a standard reading of this passage, it presents an antithesis to legal formalism. Historically, its image of judges as statesmen – indeed seamen^{Footnote 51} – neatly represents the period of what Morton Horwitz has called ‘grand-style’ jurisprudence, the predecessor to the ‘classical legal thought’ of the late nineteenth century, which avoided the latter’s delusions of judicial passivity in discovering and applying the law.^{Footnote 52} But more specifically for us, it represents the antithesis to ideal constitutional discourse, which denies the existence of non-ideal circumstances and institutional concerns that good judges do and should factor into their decisions. From the swirl of this passage, two images in particular stand out for their memorable capture of distinct, yet equally central, concerns posed by institutionally-hard cases. The first is that of ‘threatening currents’, vividly evoking concerns with legitimacy and survival – both of the court and the constitutional system as a whole. The ‘obedience due to the laws’ is precisely what Dieter Grimm evidently worried about when he published his newspaper piece in 1995.^{Footnote 53} Second, ‘the signs of the times’ captures those cases where judges are concerned with having decisions ‘mature’. Both images will provide crucial guideposts for the non-ideal theory proposed here.^{Footnote 54} But until such metaphors are subject to critical reflection, they reinforce the perception that judicial statesmen are master craftsmen, and that we should not try to pin down their choices by theoretical reasoning.

Currently, as commentators on constitutional courts have, in the main, gone beyond formalist assumptions in their various guises, the idea of judicial statesmanship is quite common, especially in more public, rather than academic, commentary on courts. A case in point is David Cole’s commentary on the Supreme Court’s line of recent decisions on same-sex marriage preceding Obergefell, which are interpreted to have promoted the cause slowly but surely. In Hollingsworth v Perry, the Court cited technical restrictions on ‘standing’ (the right to appeal to the Court), to refuse to decide whether it was constitutional for state laws to restrict same-sex marriage. The effect was to leave intact a district court’s invalidation of California’s Proposition 8 – which prohibited same-sex marriage – and thus, without delving into the substantive legal issue, the Court indirectly supported the legalisation of same-sex marriage in California. Cole hailed the decision as a ‘prudent if unusual act of judicial statesmanship’.^{Footnote 55} His observation is reflective of a widely-shared perception of this kind of situation:

While superficially appealing, this statement is typical of non-reflective adherence to ‘statesmanship’. After all, why should popular backlash, in and of itself, be even prima facie sufficient grounds to avoid issuing a principled constitutional decision on the merits? Cole’s analysis reflects a common-sense assumption that it should. The notion that it simply takes as given – namely, that courts should try to avoid controversy – is a recurring, fundamental, and fundamentally flawed dimension of the judicial statesmanship image.

Nowhere was the call to judicial ‘restraint’ in this sense more clearly articulated than in Alexander Bickel’s well-known case for the ‘passive virtues’.^{Footnote 57} One way to read Bickel is as a practical manual for the judicial statesman, where the art of practising judicial virtue is all about the decision not to decide.^{Footnote 58} Bickel provides a clear account of the various doctrinal techniques of avoidance that permit judges to stay out of important controversies: standing, ripeness, mootness, abstention and justiciability, are all methods that he advocates.^{Footnote 59} While Bickel’s notion of ‘restraint’ has been rejected as a matter of general constitutional theory on the appropriate role of the Supreme Court, there is arguably a lingering hold to his imperative of ‘not deciding’ when it comes to hot-button issues. Where the court acts carefully to avoid controversy, the image of statesmanship offers little basis for criticism. While some decisions may be hailed as ‘courageous’ or ‘wise’ in hindsight, few decisions would be described as ‘cowardly’.^{Footnote 60} In other words, the image of statesmanship, while structured as some form of balancing, lends in actual practice a cover for failures to issue principled constitutional decisions on account of non-ideal circumstances, ones raising institutional concerns of impact and perceived legitimacy. The ideal of statesmanship, that is, can too easily devolve into a general justification for privileging institutional prudence and perceived legitimacy over constitutional jurisprudence and normative legitimacy.

A distinct account, in contrast to the avoidance of ‘popular backlash’, is also offered by Cole in an earlier piece praising the same-sex marriage line of cases. Writing that Hollingsworth as well as Windsor (which invalidated parts of the Defense of Marriage Act)^{Footnote 61} should together be seen as a ‘consummate act of` judicial statesmanship’^{Footnote 62} Cole argues that, by refraining from imposing same-sex marriage on states that do not recognise it, the decisions thereby

The idea being advanced here – namely, allowing a social movement to take its course through the democratic process – is a very different one from concerns over a negative popular reaction, and mixing the two showcases the inability of the ‘judicial statesmanship’ idea to draw normatively-significant distinctions. By contrast with a concern with popularity, allowing progressive social change to take its own path presents a better claim to be a proper reason for rolling back an otherwise right and just constitutional decision. To be sure, it raises complications of its own. In particular, who is to decide what counts as social ‘readiness’? We often hear that certain decisions were ‘premature’, but, at the moment of decision, whence the confidence that the trajectory is a progressive one?

These problems indicate that, more importantly than its possible conservative tilt and Panglossian tendency, the heart of the problem with judicial statesmanship is that it offers no categories or standards with which to think through this issue normatively, an especially troubling deficit when the notion is being applied to the archetypal institution of principled normative reasoning. Why, when for the substantive aspects of adjudication we impose on judges stringent demands of reasoned argument, should we be satisfied with seeing them as artists when handling the fundamental dilemma of their institutional role? How can we take rights seriously if we informally allow them to be circumvented by intuitive, unreflective ‘statesmanship’? The position is especially wanting when we wish to evaluate a court’s behaviour: in some institutionally-hard cases, we need to be able to say that a decision was wrong or bad, and should have been otherwise, and that it was the court’s duty also to get it right on this non-ideal level, not only on the level of ideal constitutional principle. Should Israel’s HCJ have displayed greater audacity? Would it have been right for the US Supreme Court to have ruled on gay marriage much earlier? If not, did the German Court (as we imagine it) nevertheless do the right thing to act in disregard of a likely injury to its own position? What could make sense of the differences in how we think of these cases?

This is not to say that the dilemma has an easy or airtight normative answer – it may very well remain a ‘dirty hands’ situation. But at the very least we should attempt some account of the kind of considerations that should and should not count in working out the answer, and the respective weight that different legitimate considerations ought to have. The question persists also if we posit, on another interpretation of Bickelian ‘passive virtues’, that the crux is not for courts to be always and forever restrained, but rather for them to build up and preserve their legitimacy and credibility – the basis of their institutional power – in order to then ‘use it up’ only when the time is right. For this notion does not indicate what would be the parameters for assessing when such ‘accumulation’ is legitimate and when it is time to ‘cash in’. Thinking of such assessment as decisional ‘all the way down’, defying rationalisation and hanging on the intuitive act of more or less talented statesmen, is contrary to the most profound commitments of liberal legal thought. Consider, for example, a progressive court that shied away from controversy in order to build up its power, leaving aggrieved plaintiffs and other affected individuals with their rights violated, only to be later replaced by a less progressive court that is even less likely to recognise these violations. Is this to be simply understood – extending Toqueville’s metaphor – as the unfortunate wreckage of a naval expedition whose captain misjudged the currents?

V. Taking reality seriously: Non-ideal theory

In order properly to recognise and address the dilemma of institutionally-hard cases, a theory of constitutional adjudication is needed that sees courts as playing a complex role, entrusted both with the elaboration, articulation and pronouncement of ideal constitutional values, and with the meaningful and sustainable realisation of these values in actual social and political life. It is this realisation which is the ultimate aspiration of the constitutional system in which courts participate. Courts, it then becomes clear, should not be concerned with their status and popularity per se, or even with policy outcomes in a narrow sense, but they should be concerned with having an impact within the polity in which they operate, such that the principles that they propound will indeed be lived by. Every decision they make in an individual case, and in institutionally-hard cases in particular, should be seen through the lens of this role and purpose. This understanding of the institutional position of courts allows us to reorient the relationship between ideal aspirations and non-ideal circumstances, and to chart appropriate and workable principles and standards.