Beware these Dworkinian Wolves (in ‘Neo-Elyian’ Clothing)

Introduction

Judge Hercules is still living rent-free in the heads of constitutional theorists. Ronald Dworkin’s fantasy appears in the opening paragraph of The Collaborative Constitution.¹ The task of protecting rights should not be seen as the ‘solitary domain of a Herculean super-judge’, says Aileen Kavanagh.² We must dispense with this image of judges as having a ‘pipeline to truth’,³ or as ‘heroes in a “forum of principle” valiantly defending our most basic liberties’.⁴ And there he is again in the opening chapter of Responsive Judicial Review.⁵ Rosalind Dixon rejects the ‘everything view’ of courts and of judicial capacity, associating it ‘with Dworkin’s hypothetical judge “Hercules”’.⁶ And she sees the attendant thinking as assuming ‘a heroic conception of individual judicial skill and capacity, and an even more ambitious view of what courts can achieve as institutions’.⁷

Now as I read Ronald Dworkin, he never supposed that judges could divine the content of moral-political justice or, even if they could, that it might be acceptable for them to impose it upon the polity at large. But Kavanagh and Dixon seem to deploy an exaggerated form of his argument to frame and illuminate their own. They want to sweep the debates in constitutional theory beyond the impasse brought about by those blows landed on the case for judicial review by Jeremy Waldron at the turn of the century. Each pitches her argument as a kind of middle way: Kavanagh presents hers as a corrective to the ‘Manichean narrative’ that has prevailed since Dworkin and Waldron had emerged as ‘theoretical Titans’ on either side,⁸ while Dixon elaborates her preferred ‘sometimes view’ of judges and courts with reference to the ‘binary everything/nothing view’.⁹ And each wants to emphasise the ostensibly non-Herculean credentials of her argument. They are both attuned to the democratic concerns attending the imposition by judges of their own substantive views on moral-political questions. They both set much store by the notion of judges ‘calibrating’ the intensity of their review, holding off, for instance, where an impugned legislative provision is the outcome of ‘reasoned and recent’ parliamentary debate.¹⁰ And each is cognisant of the potentially deleterious consequences of robust judicial intervention.

But the arguments of these two books end up landing in a broadly Dworkinian space (or so I shall argue). Kavanagh and Dixon seem to me to underappreciate the full implications of the most basic theoretical argument made by the political constitutionalists. This is not that any judicial power to invalidate legislation is necessarily illegitimate. Rather, it is that institutional questions of the kind that Kavanagh and Dixon address are to be addressed through the lens of the concept of legitimacy rather than that of justice. ‘Given that we lack a process all can acknowledge as offering a means for grounding the truth of moral, legal and political decisions’, says Richard Bellamy in Political Constitutionalism, ‘our reasons for adopting any process will have more to do with its legitimating than its epistemic properties. Even when we dispute the decision, we will need to feel that we should accept it all the same.’¹¹ And I would add what I think is an important wrinkle to that basic argument. I would say that in the case of judicial decision-making, specifically, questions pertaining to it are to be approached less in the light of ideas around how we might get to the right kinds of outcomes on constitutional-democratic matters, and more in the light of ideas around how it might be that citizens will come over time to accept judge-imposed constitutional-democratic outcomes even when they disagree with those outcomes in substantive moral-political terms.¹² That is, that judges cannot approach their decision-making on the basis of what they might think is the most morally attractive result – that this outcome might more readily protect ‘democracy’, for instance, or that that one might more robustly guard against ‘irreversible risks to dignity?’¹³ But rather, that that legitimacy-based lens might suggest to judges something like that idea insisted upon by Herbert Wechsler: that what distinguishes the judicial process – and what might give some prospect that judge-imposed outcomes would come to be accepted by winners and losers alike – is that it must ‘rest with respect to every step that is involved in reaching judgment on analysis and reasons quite transcending the result that is achieved’.¹⁴

I know that Kavanagh and Dixon would both bristle at my claim. ‘I talk about “legitimacy” all the way through my book’, they would each insist. And it is true. Two of Dixon’s chapters are substantially devoted to the topic. And Kavanagh is very much interested in process-considerations of the legal-judicial kind (i.e. of the kind tending to be emphasised in Wechsler’s work, or in that of Lawrence Solum, or of Judge Donal O’Donnell, for instance).¹⁵ But these constitutional theorists – like most, I venture, in this post-Rawlsian age – are unduly influenced in their thinking on institutional questions by ideas pertaining to the concept of justice. Kavanagh seems to give the game away in chapter 3 of her book. She refers to the ultimate ‘goal’ of the collaborative constitution as that of securing ‘just government’, where ‘by “just government”, I simply mean government which acts justly and fairly on behalf of the community, informed by key constitutional principles including democracy, the rule of law, justice and the protection of rights’.¹⁶ And this line corresponds with what Kavanagh herself described as the ‘central claim’ of her famous ‘Reply to Jeremy Waldron’ back in 2003.¹⁷ This was that the right to democratic participation upon which Waldron had rested his defence of majoritarianism, though it was ‘undeniably valuable’, was not so valuable as to ‘displace the central importance of the “instrumental condition of good government”’.¹⁸ And that ‘political decision-making mechanisms should be chosen (primarily) on the basis of their conduciveness to good results’.¹⁹ Indeed, Kavanagh went on in that ‘Reply’ to clarify that by this she meant that ‘the justice of the outcomes of political decisions is the fundamental criterion for judging political institutions and it determines what political procedures we choose’.²⁰ Which of course corresponds with what one Ronald Dworkin had to say on the subject. He supposed that ‘the best institutional structure is the one best calculated to produce the best answers to the essentially moral question of what the democratic conditions actually are, and to secure stable compliance with those conditions’.²¹ And he could see ‘no alternative but to use a result-driven rather than a procedure-driven standard for deciding [institutional questions]’.²²

I believe that this charge that I level against these books is important. (Let us think of it as the ‘Justice, not Legitimacy’ charge, or perhaps as the ‘Outcome, not Process’ charge). But I should clarify some things at this early stage. First, I should say that I regard these as deeply insightful, sophisticated, and original books. In fact, I see them as monumental scholarly accomplishments, and as ones that will register as reference points in the fields of constitutional theory and comparative constitutional law for a long time to come.²³ Second, I should acknowledge that I have not been able to be fully comprehensive here in respect of either book. I ignore almost everything Kavanagh says about the role of the executive in the collaborative constitution, for instance, and everything Dixon says about judicial ‘tone’ and ‘framing’ as they relate to her ideas around responsiveness.²⁴ I have selected elements of the respective arguments – almost all of them relating to the judicial role. But I feel I have managed to capture the essence of each argument, and I hope I relate them fairly to the theoretical charge that I lay. Third, I should point out that the arguments of these books are actually very different from one another, and that they are different in ways that relate to this charge. Indeed, as should become clear towards the end of this review, I feel that one of the books is guilty in a relatively serious way – and affords what looks to me like a deeply uncomfortable (almost extra-legal) role to the constitutional judge; whereas the other, though guilty to a degree, affords a more modest and law-governed role to that judge.

There is a fourth item I should mention that is a little more complex. I have already gestured at the idea that Kavanagh and Dixon are not alone in approaching institutional questions through the lens of the concept of justice. I think the same is arguably true of the work of Mattias Kumm and Lawrence Sager, for example, and perhaps of the bulk of those judge-friendly-but-not-quite-Herculean constitutional scholars writing in English today.²⁵ This is not a particularly novel suggestion, I guess. It is largely in line with views articulated by Jeremy Waldron and Richard Bellamy, and it shares a family resemblance with arguments put forth in recent essays by John Finnis and Noel Malcolm, for instance.²⁶ But I have a slightly different take on the idea than Waldron’s and Bellamy’s. Waldron opens Law and Disagreement with that simple yet critical observation that ‘[t]here are many of us, and we disagree about justice’.²⁷ But he quickly turns to a broader point about the state of contemporary political theory. He says that John Rawls transformed it, insofar as no-one following in the wake of A Theory of Justice could think about anything else other than the concept of justice.²⁸ That every theorist through the late 20^th century came to focus on offering their own particular theory of justice to the world – here was how Justice as Fairness was deficient, and there was how it might be put right, or what about this liberal-feminist or radical-feminist or conservative alternative? – with each account presented by its Creator ‘as a candidate for moral and political hegemony’.²⁹ And that everyone had lost sight of the fact that here in the real world there would always be disagreement on the fundamentals of justice. Or, more to the point, that here in the real world ‘social decisions are reached, and institutions and frameworks established, which then purport to command loyalty even in the face of those disagreements’.³⁰

Waldron’s point was thus that we must take legitimacy seriously. (I should point out that when I refer to ‘legitimacy’ in this piece I have in mind much the same idea as Waldron has when he refers to ‘the circumstances of politics’.³¹ I would stress, however, that legitimacy has different kinds of process-related implications in different institutional contexts).³² But it was also that legitimacy had come to be forgotten in contemporary political theory: that this most basic concept had all but disappeared from the post-Rawlsian political theoretical stage. And that seems too sweeping to me.³³ The better view, I suggest, is offered by AJ Simmons, and it is endorsed by Philip Pettit, among others.³⁴ They argue that rather than having been forgotten, legitimacy has come to be conflated with justice in the minds of contemporary political and constitutional thinkers: that Rawls’s work, as Simmons put it, helped bring about ‘a very distinct narrowing of the argumentative grounds for claims of [justice] and legitimacy’.³⁵ Simmons refers to this as the ‘Kantian approach’ to political theory, and he sees it as having ‘effectively replaced’ his preferred ‘Lockean position’ – whereby these two concepts are understood as separate and distinct.³⁶ But he pins the conceptual error on the door of John Rawls. (‘The fundamental criterion for judging any procedure is the justice of its likely results’, says Rawls in A Theory of Justice).³⁷ And my suggestion is that we can see traces of that error in the rivers and lakes of contemporary constitutional theory, not least in the pages of these landmark books.

The remainder of this review article is in four parts. I give an overview of the core argument of each book in the section immediately below. In the third section, I look more closely at the conception of the role of the judge in The Collaborative Constitution. I turn in the fourth section to Responsive Judicial Review, exploring what Dixon has to say about how judges might approach the interpretation of a written constitution under her model. (I take a brief theoretical detour in that section too, exploring Rawls’s take on legitimacy. This is prompted, as we shall see, by Dixon’s explicitly relying on Rawls’s thinking on ‘political legitimacy’ in support of her recommendations to judges in respect of how they might approach the interpretation of a written constitution). I return to my broader charge in the Conclusion, exploring the significant differences between these two books as they might relate to it.

Overview of each book

Rosalind Dixon’s theory of ‘responsive judicial review’ is rooted in the idea that judges might protect the democratic ‘responsiveness’ of the polity. They should approach the exercise of their powers of review with one eye on the conventional legal materials at issue in the case, leaving the other to focus on the goal of protecting the capacity of the democratic system to respond to popular preferences in respect of rights-related policy choices. A judge might choose an interpretative option by virtue of its being the one more likely to guard against the emergence of an electoral monopoly, for instance (i.e. insofar as such a monopoly would diminish the responsiveness of the polity to popular preferences in respect of such choices). Or she might adopt a constitutional construction on the basis that it facilitates an outcome that best chimes with an emerging social consensus on the policy question engaged (insofar as such an outcome would itself be responsive to popular will). Judges must therefore have ‘the requisite mix of legal and political skills’, says Dixon.³⁸ She presents her book as ‘a roadmap for judges and scholars interested in building up the capacity for courts to engage in review of this kind’.³⁹ And she hopes that ‘with the benefit of this roadmap’ more democratic polities around the world might become ‘sufficiently responsive’ as to be ‘worthy of public trust’.⁴⁰

Dixon’s is thus a variation on John Hart Ely’s ‘representation-reinforcing’ approach to judicial review.⁴¹ She sees her own approach as of a piece with those of other ‘neo-Elyian’ constitutional theorists such as Stephen Gardbaum, Samuel Issacharoff and Michaela Hailbronner.⁴² But Dixon seeks to go ‘beyond Ely’.⁴³ The latter focused on two relatively unsophisticated forms of democratic dysfunction, both of them reflecting the particular case of the suppression of African-American voters in the US South in the mid-to-late 20^th century. The first of these, in Ely’s widely-cited words, was ‘when the ins are choking off the channels of political change to ensure that they will stay in and the outs will stay out’, and the second was when, ‘though no one is actually denied a voice or a vote, representatives beholden to an effective majority are systematically disadvantaging some minority out of simple hostility or prejudiced refusal to recognise commonalities of interest, and thereby denying that minority the protection afforded other groups by a representative system’.⁴⁴ Whereas Dixon rests her theory on more common and broader forms of dysfunction she identifies from around the democratic and barely-democratic world of the early-to-mid 21^st century. She is concerned with ‘antidemocratic monopoly power’, first of all.⁴⁵ And she is concerned with ‘democratic blind spots’ and with ‘democratic burdens of inertia’ as well.⁴⁶

That first form of dysfunction might be associated with seriously declining or fundamentally non-democratic polities.⁴⁷ In elaborating it, Dixon draws on the notion of a ‘democratic minimum core’ – an idea she and David Landau had advanced in earlier work.⁴⁸ The thought is that a democratic polity must at least comprise ‘a system of free and fair elections among multiple political parties, based on the accompanying protection of political rights and freedoms and a system of checks and balances’.⁴⁹ These then represent the ‘minimum core set of norms and institutions’ required to satisfy ‘electoral and institutional accountability’.⁵⁰ And these in turn correspond with the two variants of antidemocratic monopoly power Dixon identifies. ‘Electoral monopoly power’ is threatened where a dominant party or leader seeks to leverage their dominance so as to threaten the foundations of electoral competition.⁵¹ And ‘institutional monopoly power’ would involve such a party or leader seeking to diminish the independence or strength of courts or ‘fourth branch’ accountability institutions.⁵²

As for the second and third forms of dysfunction, they are more readily associated with reasonably well-functioning democratic states. Democratic ‘blind spots’ are where the governing party or legislature is unresponsive to a policy claim that would at once remove a serious limitation on the rights of a particular minority and be quite acceptable to, or perhaps even popular among, the democratic majority.⁵³ It may be that legislators have simply failed to anticipate the impact of a legislative scheme on a particular minority where, if they had foreseen it, they would happily have tweaked the scheme so as to accommodate the minority interest (‘blind spots of accommodation’).⁵⁴ Or it may be that an electoral or districting system makes for a deficit in the representativeness of the legislature overall, translating into legislative schemes tending to fail to reflect the full range of experiences and perspectives of those subject to them (‘blind spots of perspective’).⁵⁵ Burdens of inertia involve similar ‘blockages’ to the enactment of rights-based policy measures that would tend to enjoy some kind of general support in the broader society.⁵⁶ Here it may be that a rights-based claim being advanced by a minority has come over time to enjoy tacit support among the majority, but that legislators have not yet responded, perhaps for no other reason than that they have not yet managed to find the time in the hectic legislative calendar (this would be an example of ‘priority-driven burden of inertia’).

Dixon’s argument, to be clear, is not the simple one that courts should boldly intervene in these cases, stepping in heroically where political actors have attacked democratic structures or ignored democratic will. Rather – just as Ely had linked the intensity of judicial review to the particular forms of democratic dysfunction he identified – Dixon suggests that the presence or absence of her preferred forms might inform how judges ‘calibrate the scope and “strength” of their decisions’.⁵⁷ And this applies in respect of two related judicial practices. It bears upon constitutional interpretation, first of all. Judges can more readily depart from ‘“legalist” modes of interpretation’ when confronted by democratic blockages, relying instead upon ‘values-based modes of reasoning that embrace a concern for democratic protection and promotion’.⁵⁸ And it bears upon the application of doctrines used by courts to assess permissible limitations on constitutional guarantees. These ideas might represent ‘additional guideposts’ for judges, informing their approach to those unavoidably evaluative judgments as to how deferential they might be in respect of something like ‘narrow tailoring’ under Canadian-style proportionality.⁵⁹

Professor Kavanagh also embraces this notion of ‘political process review’, as we shall see. But The Collaborative Constitution is rooted in the idea that rights protection is a ‘shared responsibility between all three branches’, with each having a ‘distinct but complementary role to play’.⁶⁰ The book is an exploration of the differences between these institutions – the compositional, procedural and epistemic differences between courts and legislatures, for example – and of the relevance of these differences to questions concerning the role that each is to play. And it is an exploration of the relationships between them. Far from the ‘enemies at war’ dynamic evoked by the Dworkin-Waldron debates, these relationships are in reality shaped by the humdrum norms of ‘respect and restraint, fortitude and forbearance’.⁶¹ Political actors comply with court rulings, for example, and refrain from discussing matters that are the subject of judicial proceedings. Judges likewise ‘often leave space for democratic deliberation, and regularly defer to legislative decisions, out of respect for the competence, expertise and legitimacy of the democratically elected legislature’.⁶²

Those Dworkin-Waldron narratives play a more prominent role in Kavanagh’s argument than they do in Dixon’s. Their ‘Manichean’ thinking had tended to obscure these humdrum realities: ‘Why would a mythical demi-god with a pipeline to truth ever defer to a bunch of moral degenerates hell-bent on violating rights’, wonders Kavanagh?⁶³ The task is thus to identify a way of thinking about constitutionalism that is capable of capturing these realities while at the same time offering a normative ideal to which a system might aspire. To this end, Kavanagh rejects the ‘function-based’ conception of constitutionalism, which she associates with the ‘one branch – one function’ model (i.e. the so-called ‘pure doctrine’ of the separation of powers, where each branch operates as though in solitary confinement).⁶⁴ It fails the reality test, for one thing. It is ‘an open secret that all three branches perform all three functions to some extent’.⁶⁵ But it also fails the test of normative desirability. ‘We want the Executive to make delegated legislation’, just as we ‘need the courts to be able to develop the law even as they apply it, and sometimes to change it in significant, albeit interstitial, ways’.⁶⁶

Kavanagh thus prefers this ‘role-based’ constitutional model, where the various tasks are allocated to the branches in accordance with the suitability of the particular branch to the execution of the particular task. We might begin with a kind of outline sketch: that the Executive has the ‘energy and efficiency’ to initiate and propose new policies, for instance, and that Parliament has the deliberative capacities to scrutinise them, and the representativeness to enact them in the form of legislative frameworks.⁶⁷ But this is a ‘dynamic’ division of labour.⁶⁸ The legislature will often use vague terms in legislative enactments, for instance, thereby delegating to the courts the task of filling in the gaps in the context of the concrete cases that only they get to see in full colour.⁶⁹ And so the idea is that rather than looking blankly at the function and associating it with this or that institution, we look instead at the features of the institution – its ‘composition, decision-making process … and the skills and expertise of the officials who work within them …’ – and then try to figure whether it might be well-suited to the task at hand. In Kavanagh’s words:

By relating substantive tasks to different institutions based on their institutional characteristics, skills, competence and sources of legitimacy, we try to secure a good division of labour designed to enhance the likelihood of good decision-making overall. Thus, if an institution has epistemic or legitimacy-based strengths, it should be allocated tasks which speak to those strengths. Similarly, if it has epistemic shortcomings, it should not be assigned a task that requires the corresponding epistemic virtues. In short, a good separation of powers channels the multiplicity of decision-making tasks to the forum best placed to carry them out.⁷⁰

Kavanagh points to certain virtues of this model of constitutionalism. If the function-based approach places each branch in its hermetically sealed box, this one can account for the ‘multifunctionality’ of contemporary constitutional institutions. It can conceive of judicial law-making as a feature rather than a bug of the system, for instance. But it also accommodates that notion of ‘role obligation’ that is familiar to constitutional actors. When we talk about the role of a judge or a legislator, ‘we are not just making factual statements about what they do’ but are rather making ‘normative statements about what they ought to do qua judge or qua legislator’.⁷¹ We thus situate the tasks of the branches of government within a ‘constitutional role morality’, or within a ‘set of norms, principles, and standards which constitute the role and guide its exercise’.⁷² And so Kavanagh’s model can explain that relational nature of institutional power: that it is really a question of ‘relative authority’, with each branch ‘sharing powers and functions, whilst remaining mindful of the legitimate role of their constitutional partners in governance’.⁷³

The role of the judge in the collaborative constitution

Let us stick with The Collaborative Constitution for now, as we turn to consider the role of the judge and of courts. Kavanagh begins with those institutional features: their independence, for instance, and their legal expertise and individualised and concrete focus. And she proceeds to explore their attending strengths and weaknesses as they might pertain to the various tasks. Courts have that ‘more focused, pointillistic perspective’, for example, whereas governments are typically better-placed ‘to see the big picture on a broad policy canvas’.⁷⁴ This does suggest that the main job of the judge is to resolve disputes about what the law requires in this or that case. But it does not mean that hers is a purely mechanical role. Far from it, says Kavanagh, pitching the judge instead as a ‘partner’ in the joint enterprise of governing. And the judge takes an ‘active’ part in various aspects of governing – including in that process the making of law. She ‘fills in the gaps’ in legislative frameworks, ‘fleshing out the detail in an ongoing process of application and elaboration’.⁷⁵ She adapts provisions ‘to fit with changing circumstances and different social needs’.⁷⁶ And she ‘integrates disparate legislative measures into ‘the general constitutional and systemic fabric’ of the law, thus helping to ensure ‘coherence, stability and “normative harmony within the system”’.⁷⁷ Sure, the judge owes her legislative partners ‘a duty of respect’. But her ‘ultimate fealty and larger loyalty’ is ‘owed to the constitution’.⁷⁸

This all sounds positively Herculean, of course. But it is by no means the basis of my gripe with Kavanagh’s conception of the judicial role. (In fact she seems to capture here what judges routinely do in cases concerning statutory interpretation, and she provides a persuasive account as to why we might want them to do it).⁷⁹ Where I begin to get uneasy is with Kavanagh’s endorsement of Lawrence Sager’s comparison of the role of judge to that of the ‘quality-control inspector in an automobile plant’.⁸⁰ (This in a section of Kavanagh’s book entitled ‘Courts as Quality Control’). Now, her point here, and Sager’s, is that the political actors ‘get there first’ in the law-making process – in the manner of the engineers and those technicians down on the factory floor – and that judges have a subsequent and restricted role. But consider how it is that Sager articulates the ‘mission’ of the judge in this analogy (this in an excerpt from his work that Kavanagh includes with approval in her own text). The quality-control inspector ‘has only the job of assuring that the cars which leave her plant are well-built’, says Sager, and her role ‘comes on top of the efforts of the people who actually put the cars together’.⁸¹ Which is seen as analogous to the role of the constitutional judge. ‘Their mission is singular – to identify the fundamentals of political justice that are prominent and enduring in their constitutional regime and to measure legislation or other governmental acts by those standards.’⁸²

We must be careful, of course, about the implications we draw from the words people use. Just because Sager and Kavanagh deploy that phrase in that way – that the mission of the judge is to identify the fundamentals of political justice that are prominent and enduring in their constitutional regime and to measure legislation or other governmental acts by those standards – does not mean that their conception of the judicial role is necessarily rooted in any supposed epistemic properties of judicial review, or on its supposed facilitation of ‘just’ or ‘enlightened’ or ‘correct’ outcomes.⁸³ But the phrasing bears emphasis, I suggest, and it seems a portent of Kavanagh’s elaboration of the on-the-ground functioning of the constitutional judge through the key chapters of The Collaborative Constitution.⁸⁴

I say this because that on-the-ground functioning reads as though it is concerned as much with ‘coaxing’ legislators along the path of moral-political justice as it is with making authoritative determinations in respect of disputes between individuals as to their legal rights. The starkest example may be in chapter 10, which Kavanagh situates around the Nicklinson case on assisted dying.⁸⁵ Tony Nicklinson’s challenge to the blanket statutory ban was rejected by a majority of 7-2 of the UK Supreme Court, with two of the seven in the majority regarding it as an issue that only Parliament could resolve. But the other five, as Kavanagh explains it, were open to a finding of incompatibility – just not for the moment, and in part because Parliament was shortly due to debate the policy question at issue. Kavanagh likes that these judges ‘held fire’ in those circumstances, reflecting as a did a kind of ‘collaborative’ posture on their parts. She also likes that they had issued a warning to legislators: that the next time a Tony Nicklinson-type applicant came before them they might not be so indulgent. (The judgment was thus ‘not a judicial “no” for all time’ but was ‘more like a “no for now” but “never say never”’).⁸⁶ And she likes what she sees as their having ‘nudged’ legislators on policy detail. One of the five went as far as to offer a list of 18 criteria that legislators might consider would be appropriate to guide a High Court judge who would be tasked with assessing applications under a new statutory regime.⁸⁷ This was too much for Kavanagh, but she approves of the more general versions adopted by other judges in the majority, and indeed of the concept ‘Judge as Nudge’ in general. ‘If the courts can spot some deficiencies in the law’, she says, ‘it may be worthwhile to feed that information back to those on the factory floor, even offering some suggestions about what might fix the problem, so that the legislation can pass the quality-control inspection next time around’.⁸⁸

Now I hope it is clear from my account that Kavanagh respects the judicial obligation to operate ‘within existing legal structures’,⁸⁹ and that she worries about things like judicial ‘agenda control’ and ‘democratic distortion’.⁹⁰ And it should be acknowledged that judges across constitutional democracies do offer ‘hints’ or ‘prods’ of this kind – and often for eminently sensible reasons. But my concern is that Kavanagh gives these co-legislator style tasks such prominence in her elaboration of the judicial role. (I found myself wondering whether the metaphor she hit upon – ‘collaborative constitutionalism’ – tended to push her argument in that direction). Indeed, Kavanagh’s embrace of ‘political process review’ has judges similarly exploring all manner of extra-legal considerations (i.e. and then ‘calibrating’ the intensity of their review accordingly). She has them looking to whether an impugned legislative provision has been the subject of ‘reasoned and closely-considered judgment’⁹¹ on the part of legislators, for instance, or if there had been a ‘“thoughtful and well-informed” consideration of the rights issues involved’.⁹² And she even has them assessing ‘whether “all opposing views were … fully represented” in parliamentary debate, and whether the legislative framework was “the result of substantial research and intensive consultation with a wide range of interested and expert groups and individuals”’.⁹³

I have the similar quibbles with her idea of ‘Proportionality in Partnership’ – which Kavanagh sets out in chapter 9 of her book. She conceives of Oakes-style proportionality as a kind of judge-fashioned ‘filtering device’ that – much in the manner proposed by Mattias Kumm and Cristina Lafont – ‘screens out inadequate justifications for rights-infringing measures’ and accordingly provides for a ‘democracy-enhancing shared discourse of justification for action claimed to limit rights’.⁹⁴ (Which in turn promotes ‘the shared project of achieving good government under the constitution’. Again, notice that result-based orientation).⁹⁵ She acknowledges the objections proffered by the likes of John Finnis and Timothy Endicott: that proportionality requires judges to ‘weigh the unweighable’ and thus licences what are essentially political interventions.⁹⁶ But she counters on the basis that the bearing it has on judicial control of state acts ‘is not determined by the structure of the test but by the degree of judicial restraint practised in applying it’.⁹⁷ That is, that while judges could apply the proportionality framework with intensity they generally should not, and on the basis of general democratic considerations around institutional competence and legitimacy.

Now again, there is no suggestion that this is at odds with actual judicial practice. And I might add – for what it is worth – that I find it plausible that the institution of judicial review may indeed help to force a kind of cleansing of public deliberation along the lines suggested.⁹⁸ But I do wonder whether this increasingly popular approach to justifying judicial review might tend to engender a kind of homogenisation of judicial practice across constitutional democracies, and an accompanying drift from attentiveness to legal particularities and indeed from traditional court-based methodologies (e.g. of the kind emphasised by Herbert Wechsler or Judge Donal O’Donnell).⁹⁹ That is, that constitutional review might become less of a specifically legal or law-based institution – one focused on what the particular laws (or constitution) of the particular jurisdiction might be understood by trained lawyers to say or indeed not say.¹⁰⁰ And that it would become more like ‘a forum of policy review analogous to auditing’ – as Martin Loughlin has put the idea (disapprovingly), or like a generic ‘quality control inspection’ of the kind I think is implied (not always intentionally) in some of the contemporary literature.¹⁰¹

There is a further wrinkle in Kavanagh’s thinking here that I might finally mention. And this is the extent to which she relies on that notion of ‘self-restraint’ in fending off those challenges from sceptics.¹⁰² (What I am getting at here is that she relies on notions of human-judicial restraint as distinct from legal-constitutional constraint – again in the manner explored in Wechsler’s work).¹⁰³ Now this may be attributable to the fact that she is working from the UK constitutional system, and thus from one in which there is no master-text constitution from which it might be possible to attempt to deduce such constraints.¹⁰⁴ And I grant that the distinction may appear to be a fine one. Legal Realists would dismiss its significance, as I think might those not intimately familiar with the workings of a constitutional system in which judges reason with reference to the text and structure of a constitution that is codified, relatively modern, and relatively easy to amend. But to me the distinction is meaningful, and yet it is rarely emphasised in our field. And when Kavanagh speaks of the various ‘calibrating factors’¹⁰⁵ she suggests might drive judges to intervene or to abstain, they seem to me to be generic and detached rather than to be anchored in a given constitutional text or structure (i.e. in the sense of having a legal ‘root of title’ in that text or structure).¹⁰⁶ It is as if they emanate from somewhere within the democratic/collaborative hearts of individual judges (i.e. as if those judges have a broad discretion to abstain or intervene), or, as John Finnis might put it, from ‘over there in a haze of “global law”, made how or by whom no-one can really say…’¹⁰⁷

The role of the judge in responsive judicial review

We saw Professor Dixon’s proposal that judges would be guided in the exercise of their powers by the presence or absence of those democratic dysfunctions, and that they could more readily depart from ‘legalist modes’ of constitutional construction when confronted by them, preferring ‘values-based modes’ instead.¹⁰⁸ Well, in elaborating her thinking in this regard in chapter 4 of Responsive Judicial Review, Dixon addresses the concept of legitimacy directly, distinguishing between ‘legal legitimacy’ on the one hand – which she takes to refer to ‘the degree to which judicial decisions conform with existing legal norms or constraints’ – and ‘moral legitimacy’ on the other.¹⁰⁹ (She sometimes refers to the latter as ‘political legitimacy’). Dixon does not explore moral/political legitimacy in a great deal of depth in the book – despite its playing a significant role in informing her take on judicial practice. But what she does say is notable. She says that it refers to ‘the degree to which a legal decision is “morally justifiable or respect-worthy”’, and that she takes a ‘Rawlsian’ line in respect of it.¹¹⁰ That is, that she prefers a ‘more minimal, political conception’ of moral justifiability than the ‘relatively demanding’ and ‘substantive’ conception she gestures at by way of alternative. (She presumably has in mind a conception relying on what Rawls would think of as a comprehensive doctrine). And that her preferred conception would be satisfied by ‘what is required in order to ensure the (hypothetical) consent of rational and reasonable citizens to a constitutional system’.¹¹¹

The most basic conclusion Dixon reaches from these ideas is that judges can draw upon either source of legitimacy in justifying their decisions – that is, on the basis of considerations pertaining to legal legitimacy or, in the alternative, on the basis of those pertaining to moral/political legitimacy. This is a critical aspect of her argument, and I shall get to it momentarily. But first – by way of a short detour – I want to consider a theoretical point that seems to me to be even more fundamental. And it concerns that relationship between legitimacy and justice in Rawls’s work, and by extension in Dixon’s.

Rawls (and Dixon) on ‘political legitimacy’

We saw in the Introduction the suggestion that Rawls had conflated legitimacy and justice, or that he saw the former as parasitic upon the latter. And we can see how that is so from several aspects of his main body of work including from his discussion of political obligation in A Theory of Justice. The story begins in the earlier chapters with the ‘original position’, and with the two principles of justice Rawls supposed would be chosen by the hypothetical parties that he places there. (These are the two principles of which Rawls’s particular theory of justice, ‘Justice as Fairness’, is comprised). But consider where it goes from there. Once they have chosen those principles they ‘move to a constitutional convention’, where – as ‘delegates’ to that convention – the parties are to choose ‘the most effective just constitution, the constitution that satisfies the principles of justice and is best calculated to lead to just and effective legislation’.¹¹² And it is this constitution that plays the foundational role in his analysis of political obligation (‘…we normally have a duty to comply with unjust laws in virtue of our duty to support a just constitution’).¹¹³ Indeed it plays the foundational role in the definition of legitimacy that Rawls sets out two decades later in Political Liberalism (1993):

Our exercise of political power is fully proper only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason. This is the liberal principle of legitimacy.¹¹⁴

Dixon is therefore quite right in associating Rawls with an approach to legitimacy focusing on, as she put it, ‘the (hypothetical) consent of rational and reasonable citizens to a constitutional system’ (i.e. those hypothetical parties in the original position/constitutional convention are doing all of the heavy lifting). But consider the apparent implication. It is that the justifiability of the exercise of coercion in circumstances of disagreement about justice is thought to depend upon agreement about justice! That is, that the question of whether a given law is legitimate – whether it gives rise to a presumptive obligation of compliance etc. – depends upon its compatibility with this ‘just constitution’, with that constitution’s credentials as ‘just’ determined with reference to the particular conception of justice Rawls supposes would be chosen in the original position. Which might all work in a world in which everyone subscribed to that particular (broadly social democratic) conception of justice. But – as Waldron in particular has insisted – it can hardly be thought to work here in the real world, where ‘full-blooded disagreement about justice remains the most striking condition of our politics’.¹¹⁵ It can hardly be thought to work in a world characterised by deep disagreement at the level of principle between socialists and libertarians, for example, or between radical feminists and conservatives.

Now it may be that this sketch of Rawls’s approach to legitimacy is too simple. I know Rawls himself would regard it as such.¹¹⁶ And I would certainly see it as at odds with the approach that we see emerging in his later writing. We see Rawls distinguish explicitly between legitimacy and justice in his ‘Reply to Habermas’ (1995), and we see his definition of legitimacy evolve in accordance with that in ‘Introduction to the Paperback Edition’ (1996) – and again, more dramatically this time – in ‘The Idea of Public Reason Revisited’ (1997).¹¹⁷ We also see the Rawls of this period coming to insist that he had only ever intended Justice as Fairness as ‘one example’ of a liberal political conception of justice – with the implication that he had recognised all along that there would always be disagreement on justice in any democratic society, or certainly that he recognised that at this stage in his writing.¹¹⁸ And we see the Rawls of this period rein in his references to the original position, and to place actual people and what they actually think (i.e. as distinct from hypothetical people and what they might reasonably be expected to think) at the heart of his theory.¹¹⁹

I cannot get into the weeds of these shifts here, having done so in other work.¹²⁰ But I will say that I see them as significant, and that I read them as evidence of Rawls having come to appreciate the conceptual difficulties with his earlier approach to legitimacy. Indeed, I see them as evidence of his having come to appreciate the conceptual distinctiveness of legitimacy and justice, and the logical priority of the former in respect of institutional questions. I would add a further suggestion though. And this is that these shifts in Rawls’s thinking seem to have barely registered in the field of constitutional theory.¹²¹ That is, that his critics in our field seem to me to criticise him on the basis of his work as it stood in 1993: Waldron and Bellamy depict Rawls as having almost entirely overlooked disagreement on justice, and as having insisted upon a rigidly fixed or effectively permanent set of constitutional essentials.¹²² And that his disciples seem to buy in to his thinking as it stood at that juncture too. The likes of Dixon and Sager and Kavanagh might not say it quite directly: Rawls is barely mentioned in the pages of The Collaborative Constitution, for instance, and Dixon does not get into any depth in her analysis of his ideas in Responsive Judicial Review. But ‘justice’ seems to me to be the ‘first virtue’, or the overriding goal, in both cases, and legitimacy is seen in its shadow.¹²³

Constitutional ‘constructions’ in responsive judicial review

Let us return to the main route then and consider what Dixon’s prescriptions are for judicial practice? What conclusions does she draw in this respect from this thinking around legitimacy? Well, she refers throughout her book to the notion of constitutional ‘implications’, apparently in reference to interpretations/interpolations of a written constitution that do not enjoy strong support in the text or structure of that constitution, or in a court’s prior case law. And she has the following to say about when such ‘implications’ might be justified:

A responsive approach suggests that actual judicial legitimacy should be understood as an amalgam of legal and political sources of legitimacy, or as a cumulative concept that includes elements of both legal and political legitimacy. A judicial decision may be legitimate, in this view, if it has a strong basis in formal legal sources – that is, the text, history, and structure of a constitution, or a court’s prior case law. Or it may be legitimate because it gives effect to powerful political commitments, or has some (lesser) degree of both legal and political legitimacy. A responsive approach to judicial review further suggests a quite particularized focus for judgments about political legitimacy: it suggests there will be powerful democratic arguments for courts adopting a range of constitutional implications in order to protect the minimum core of democracy, and persuasive political reasons for courts to adopt implications aimed at countering legislative blind spots or burdens of inertia.¹²⁴

Dixon later introduces the idea of ‘irreversibility’ in reference to those ‘failures to protect human dignity [that] may literally be impossible to reverse’.¹²⁵ (A failure to provide housing or healthcare to children would presumably count, insofar as the damage it might tend to cause would often be difficult to reverse? A lack of access to abortion in cases of rape or incest might also qualify?). She talks about ‘forms of democratic inertia’ that may be ‘so widespread and systemic that there is little chance that they will be overcome, and basic citizen needs and demands met, without judicial intervention’.¹²⁶ This, she says, will ‘provide additional arguments for the political legitimacy of implications designed to help counter such failures’. And that paves the way for the three broad principles she elaborates for ‘governing the legitimacy of judicial implications under a written constitution’. Dixon spells them out as follows:

1. (1) Implications that have limited legal support, and no real political justification, will be presumptively illegitimate: by definition, the idea of an ‘implication’ suggests a limited degree of textual support for such a doctrine, and absent any real political justification, this should be sufficient to encourage a court to exercise restraint.

2. (2) Implications designed to protect the ‘minimum core’ of democracy will generally be legitimate, regardless of the degree of existing legal support for such an implication (or support in the text, history, and structure of a constitution, or a court’s prior case law). They will draw their legitimacy from political arguments in favour of courts seeking to respond to an urgent and systemic risk to constitutional commitments to democracy and democratic responsiveness.

3. (3) Implications designed to counter blind spots or burdens of inertia may also be politically legitimate, but only where they enjoy some meaningful degree of legal support, or are designed to counter serious and irreversible risk to human dignity, or systemic forms of inertia or state failure.¹²⁷

Let us dwell for a moment on that second principle. The suggestion appears to be that judges, when responding to what they see as threats to basic political rights and freedoms, can ignore legal considerations and be moved by moral/political considerations alone (‘regardless of the degree of existing legal support’). It would seem to allow for the judicial creation from thin air of a ‘basic structure doctrine’, for instance. Dixon approves of the construction of such a doctrine by the Supreme Court of India in its Kesavananda Bharati judgment. But the critical point is that she would have approved even if there had not been any foundation for it in recognised legal sources (i.e. any arguable such foundation). She notes its having had such an arguable foundation in the wording of Art. 368 of the Indian Constitution, as well as its having been ‘prefigured’ by an earlier judgment of the Court.¹²⁸ But she thinks of that as a kind of bonus, insisting that there would be no difficulty under a ‘responsive’ approach with judges constructing such a doctrine in the absence of any such legal foundations. A doctrine of unconstitutional constitutional amendments, she says, ‘can still be legitimate where it is developed afresh by a court, without prior doctrinal support – providing it responds to an urgent and systemic threat to democracy in the form of a credible risk of electoral or institutional monopoly’.¹²⁹

As for her third principle, it would seem to allow for extensive governing by judges of the most contested moral questions in contemporary political societies – and again on political rather than on legal grounds. Dixon elaborates with reference to various examples including abortion. She approves of those pro-choice judgments of the US Supreme Court in the late 20^th century on the basis that they ‘helped counter inertia in the recognition of rights of access to abortion in many states’.¹³⁰ But she prefers the position adopted in Planned Parenthood v Casey to that in Roe v Wade. And her preference appears to be rooted not in anything pertaining to legal material or argument but rather in the apparently superior judicial reading of the social and political tea leaves.¹³¹ That is, it is rooted in the notion that the particular outcome in Casey – on the evidence of polling data from the period – was more closely aligned with majority sentiment on abortion.¹³² The judges had gone too far too fast in the earlier case: Justice Blackmun had provided for an insufficiently restricted right of access in the first trimester when considered in the light of contemporaneous polling data in respect of early-stage abortion. Whereas in Casey they had narrowed the right so as to allow for legislative regulation prior to a fetus becoming independently viable – which for Dixon was in line with the data indicating that a ‘clear majority’ of American voters believed that abortion should be ‘safe, legal but rare’.¹³³

This ‘reading the room’-style thinking is similarly at play in Dixon’s recommendations to judges as to how they might stave off concerns around ‘reverse burdens of inertia’ and ‘democratic backlash’. The first idea is illustrated by the judgment in Roe v Wade: that while it struck down an overly broad prohibition on abortion it replaced it with a framework that was overly broad in turn (i.e. Justice Blackmun’s trimester framework prevented state legislatures from regulating abortion in the first trimester in a manner that not only would have had strong majority support but that would also have counted as ‘reasonable’).¹³⁴ And Dixon’s proposal is that judges might opt for a ‘weak-strong’ approach in this kind of scenario: that they might reason narrowly rather than broadly, for instance – thus ‘leaving open broad scope for legislators to respond to court decisions, including in ways that express disagreement with aspects of a court’s reasoning or logic’ (i.e. as per the reasoning in Casey rather than in Roe).¹³⁵ Or that they may opt for a declaratory rather than a coercive remedy, or a ‘suspended’ rather than an immediately operational declaration of invalidity.

As for ‘democratic backlash’, this is where the disagreement with a court decision might be so ‘widespread and deeply felt’ as to prompt a project of ‘democratic retaliation’ more generally, or ‘an attack on the courts as institutions’.¹³⁶ And so here the concern is less about ‘reasonable disagreement’ of the kind we might associate with those dignity-type questions such as abortion or assisted dying, and more about matters going to that ‘minimum core’ of democracy (i.e. those threats of electoral or institutional monopoly that we considered earlier). She again proposes that judges might proceed responsively. But in this instance they would be responding to these serious threats and so their approach will be ‘wholly pragmatic rather than principled in nature’.¹³⁷ Their motivating goal will be to do as much to uphold democratic-constitutional norms as possible while at the same time managing to themselves ‘live to fight another day’.¹³⁸ And so Dixon recommends a ‘strong-weak’ approach in this kind of scenario – as she suggests had been adopted in her example of the Colombian judges in the two Re-Election cases in the early 2000s.¹³⁹ In the First Re-Election case the Constitutional Court upheld the validity of President Uribe’s proposed amendment, thereby allowing him to seek another term in office. (The suggestion is that the President’s popularity meant that any more robust approach could have prompted a backlash). But they made sure to lay down certain doctrinal markers that would tend to counteract any future attempt to extend term limits even further. And it was to these doctrinal markers that the judges turned some years later in the Second Re-Election case – thereby blocking President Uribe in his attempt to repeat the trick.

Dworkinian wolves?

It should be clear by now that my argument chimes in important respects with arguments put forward by Jeremy Waldron and Richard Bellamy. Certainly, I share their view that there has been a tendency to underappreciate the nature and the extent of the disagreement on justice prevailing in contemporary democratic societies (i.e. that disagreement goes to the core of the principles of justice, rather than being concerned merely with the application of those principles to concrete questions). Or – perhaps more to the point – that there has been a tendency to miss the logical implication of this disagreement, which is that we just cannot approach institutional questions using ‘the justice of the outcomes’ as our ‘fundamental criterion’ for assessment.¹⁴⁰ Or we cannot do so while expecting that people in democratic societies might continue over time to accept the outcomes that emerge, or that they might continue to regard the institutions as support and respect-worthy (Exhibit A: the US Supreme Court, and its standing in the eyes of the American public over time).¹⁴¹

But this does not mean that I am with Waldron and Bellamy on everything. I have already mentioned one subject on which I take a different view. I think they mischaracterise the work of John Rawls in respect of legitimacy and justice, or certainly that they fail to give due account to the shifts in Rawls’s thinking on these questions in his later writings. (And this is no small matter insofar as Rawls seems to represent the Big Bad Wolf for Waldron and Bellamy: the intellectual godfather of the ‘legal constitutionalism’ to which they so deeply object). I further suggest – as Paulo Sandro has too – that they mischaracterise the idea of ‘strong form’ judicial review, or that they use a particular version of it – the US version – as a stick to beat the idea in general.¹⁴² (This is no small matter either. The US version is about as difficult a version to defend on legitimacy grounds as it might be possible to imagine).¹⁴³ But there is a third quibble that I wish to underline, which I think maps on to the general argument I am making in this article. And this is that Waldron and Bellamy speak of judicial review of legislation as though it were just an alternative forum for the resolution of questions about justice, or as though constitutional courts, when exercising powers of review of legislation, were engaged in essentially the same enterprise as that engaged in by legislatures when exercising the legislative power. They speak of judicial review as if it were an enterprise that was fundamentally political in nature. (They object to it on that basis, of course. They say that elected legislatures have stronger democratic credentials than courts have, and that they accordingly have stronger democratic authority to resolve these essentially political questions).

Now it may that this is a function of their fixation on the US system: that it would be hard to avoid such a view of the enterprise if it is looked at through that lens. But my suggestion is that Waldron and Bellamy – and not only them, but also many of those judge-friendly-but-not-quite-Herculean scholars in the ‘legal constitutionalist’ camp – are not fundamentally interested in mundane legal material or doctrinal argument, or in the complex legal mechanics of particular constitutional systems.¹⁴⁴ Or that there is a tendency to give insufficient consideration to such questions, at least (a function of scholarly pressures for internationalisation?), and to questions of judicial method and process in particular. I have already mentioned Herbert Wechsler’s views on these questions – which I think would these days be thought of as quaint. I cannot explore them in depth here, of course. But I can say that Wechsler’s work is distinguished by the sophistication of its analysis around what the text and structure of a particular constitution might be understood to mean or imply in this or that context in respect of whether a judge is obliged to intervene or abstain.¹⁴⁵ And his point was that this, though it is often a very difficult question, is ultimately a question ‘of constitutional interpretation, to be made and judged by standards that should govern the interpretive process generally’ – and that this ‘is toto caelo different from a broad discretion to abstain or intervene’ (i.e. it is ‘by the whole extent of the heavens’ different).¹⁴⁶

Indeed, Wechsler insisted above all else – as we have seen – that what distinguished the judicial process was that it must ‘rest with respect to every step that is involved in reaching judgment on analysis and reasons quite transcending the result that is achieved’.¹⁴⁷ And what he meant was that it is perfectly valid for political actors to point to results in justification of their decision-making – they can point to a decision’s being one that is conducive to the dignity of a particular group, for instance, or to what a particular constituency of voters might desire, or indeed to the stability of core elements of their democratic regime. (These examples are influenced by Dixon’s ideas, of course; they were not spelled out by Wechsler in quite these terms). Whereas judges can not validly do so, at least not consistently with the obligations arising from the nature of their function.¹⁴⁸ Wechsler of course anticipated dissent on the point ‘from those more numerous among us who, vouching no philosophy to warranty, frankly or covertly make the test of virtue in interpretation whether its results in the immediate decision seems to hinder or advance the interests or the values they support’. But in response he had this to say:

The [judge or legal critic] who simply lets his judgment turn on the immediate result may not, however, realise that his position implies that the courts are free to function as a naked power organ, that it is an empty affirmation to regard them, as ambivalently he so often does, as courts of law. If he may know he disapproves of a decision when all he knows is that it has sustained a claim put forward by a labour union or a taxpayer, an [African-American] or a segregationist, a corporation or a Communist – he acquiesces in the proposition that a man of different sympathy but equal information may no less properly conclude that he approves.¹⁴⁹

Now I cannot help but imagine what those most immediate subjects of my criticism might think if they were to read these lines. Aileen Kavanagh would surely be thinking – exasperatedly, perhaps – that she is fundamentally interested in doctrinal legal material and in the complex legal mechanics of a particular constitutional system (indeed, in other work, she goes out of her way to stress this particular idea);¹⁵⁰ that she does give comprehensive attention to questions of judicial process; indeed, that she sets particular store by the procedural differences between the branches of government, proposing that we allocate this or that task to this or that branch in this very light (i.e. in the light of the suitability of the particular branch, given its procedural features, to the execution of the particular task). She would likewise point to her seeing simple dispute-resolution as the primary function of courts, and to her emphasis on the judicial obligation to operate always ‘within existing legal structures’.¹⁵¹ And she would no doubt object to any suggestion that she goes in for the kind of ‘ad hoc’ or result-based evaluation that Wechsler supposed ‘has always been the deepest problem of our constitutionalism’.¹⁵²

My response would be largely to concede these points, and to acknowledge that ‘legitimacy’ is an important theme in The Collaborative Constitution. Indeed, it has always been an important theme for Kavanagh. She is always attuned – as she herself put it in earlier work – to ‘general institutional considerations about the way in which legislatures make decisions in comparison to judges, the factors which influence their decision, and the ways in which individuals can bring their claims in either forum’¹⁵³ (i.e. to what I would see as how- rather than what-oriented considerations, or, as Bellamy might put it, to properties of this or that institution that might count as ‘legitimating’ rather than ‘epistemic’ in nature). And she could never be accused of engaging in any crude form of ‘ad hoc’ evaluation in Wechsler’s sense, or of coming at cases with the result primarily in mind.¹⁵⁴

But neither could Dworkin be accused of these things – that villain of Kavanagh’s piece. And my suggestion is that Kavanagh, much like her villain, is a ‘new Kantian’ in Simmons’s sense: that she tends to conflate legitimacy and justice, or to ‘narrow the argumentative grounds’ for claims of one and the other.¹⁵⁵ Indeed, at one point she seems almost to state it expressly. Late in her discussion of Nicklinson in chapter 10 – as it happens it is when she is considering Dixon’s ideas around legislative burdens of inertia and judges ‘stepping in where Parliament fears to tread’ – she approves of Lord Sales’s assertion that ‘in legitimacy terms, the courts know that public respect for the law depends in the long term to a considerable degree on the fairness and defensibility of the outcomes which law produces’.¹⁵⁶ (Which strikes me as a more subtle form of that results-based approach lamented by Wechsler). And this underlying idea shows its face in various concrete ways in her scheme. It shows up, for instance, in the extent of the emphasis placed on that notion of ‘Judge as Nudge’. (I found myself wondering what Tony Nicklinson was expected to think of the idea that the judges ‘held fire’ in his case while indicating that the next person in his situation who happened to bring a challenge might succeed. Isn’t this a court of law?). It shows up in her reliance on what I have suggested is an idea of restraint-not-constraint, and in her integration of such extra-legal considerations as ‘whether “all opposing views were … fully represented” in parliamentary debate’.¹⁵⁷ And it shows up – perhaps above all – in her overall conception of the ‘mission’ of the constitutional judge: as one directed at identifying ‘the fundamentals of political justice that are prominent and enduring in their constitutional regime’ and at ‘measuring legislation or other governmental acts by those standards’.¹⁵⁸

As for Rosalind Dixon, she would surely begin by insisting that her book is fundamentally concerned with ‘process’, and that ‘legitimacy’ is therefore its guiding ideal (i.e. where we might more readily associate ‘substance’ with ‘justice’). But I would counter more robustly in this case. First, I would point out that her process considerations pertain very much to the democratic-political rather than to the legal-judicial domain (i.e. she is fundamentally concerned with those ‘democratic malfunctions’, such as when legislators have failed to address some rights-based problem due to pressures on the legislative timetable, or where they have failed to perceive certain popular support for a particular right-based measure). This is not to suggest that it is intolerable that a judge might account for such considerations, as Justice Blackmun considered polling data in the course of writing his opinion in Roe, for instance.¹⁵⁹ But it seems to me – in Dixon’s case, if not in Ely’s – to come at an intolerable cost to process considerations of the kind applicable in the domain of judicial decision-making, specifically – namely process considerations of the legal-judicial kind (i.e. of the kind elaborated by Wechsler in particular). Dixon makes no bones about the scope and ambition of her thesis. It goes considerably beyond Ely’s, offering a normative and general theory as to how judges on apex courts – all across the constitutional democratic world – are to approach the process of ‘constructional choice’.¹⁶⁰ And right at the heart of that theory is a licence for judges to identify constitutional ‘implications’ that have no ‘existing legal support’ or support ‘in the text, history, and structure of a constitution, or a court’s prior caselaw’. Certainly, it is a provisional licence: it can only be used where the judge is satisfied that the ‘minimum core’ of democracy is at stake. But that is a contestable question in itself. And in any case it is accompanied by another licence, this one enabling judges to identify implications that have ‘some meaningful degree’ of legal support – just ‘some’ – or that are designed to counter ‘a serious and irreversible risk to human dignity’, which licence can be drawn upon in the more routine cases of ‘blind spots’ and ‘burdens of inertia’.¹⁶¹

In the end, Dixon’s theory seems to me to be driven by that American fixation on the ‘counter-majoritarian difficulty’ (or, to put it in contemporary terms, on that fixation with the work of Jeremy Waldron). It is as though the only possible route to justifying judicial review could be through its somehow protecting ‘democracy’, or through its securing of particular ‘democratic’ ends (i.e. bringing them about, by hook or by crook). But – as a prominent scholar has pointed out – the ‘key problem with “the counter-majoritarian difficulty” is that it narrows “the legitimacy register” to electoral credentials alone’.¹⁶² And my suggestion is that we constitutional theorists might look to process considerations of the legal-judicial kind in order to find the legitimating properties of judicial review. Whereas Waldron and Bellamy – and Dixon, among several others on that side of the aisle – seem to me to underappreciate such considerations. Indeed, Dixon’s theory seems almost to suppose that the most basic skill of the constitutional judge is to read the democratic tea leaves, or that the most basic function of the constitutional court is to make the representative process more representative (or indeed to step in to represent the people when the representative institutions fall short). Yes, she takes that ‘sometimes’ view of courts – that judges might often opt for restraint. But the gist of the argument is that that is largely a strategic call. That judges should fix democracy, or counteract blind spots, whenever they have sufficient political capital and can make it stick.

My suggestion, ultimately, is that Dixon’s constitutional judge might actually be more Herculean than Hercules himself. (Hercules J was constrained by the ‘criterion of fit’, after all). And that Dixon’s approach – if adopted by judges worldwide in the manner she recommends – may have the effect of jeopardising the standing of constitutional courts in general over the longer haul. That Dixon’s ‘roadmap’ would lead to a culture in which cases came to be determined more and more on the basis of political considerations – and on what might be thought the right democratic destinations. Is this outcome required in order to ‘keep open the channels of political change’? Does that one guard against ‘irreversible risks to human dignity’? Might the one over there be closer to ‘national majority opinion’? And that constitutional courts would come over time to be seen as – and indeed to be – something other than courts of law.