Transformative process theory

Abstract

Transformative constitutionalism and process theory are generally seen as worlds apart. But they may be more compatible than we think. A transformative understanding of process is very broad, but it represents a natural extension of the line already being taken by contemporary process accounts intent on expanding the theory to fit global practice. It can help us to understand why an expansion based on and including a wider set of justiciable process concerns has proved difficult to limit. Conversely, transformative constitutionalism badly needs a better account of judicial restraint to balance its preoccupation with judicial boldness. Since it shares with process theory a deep concern with democracy, it can naturally draw on process accounts to understand its own limits. Democracy-seeking review, in aiming to build as well as protect democratic capacity, needs to be as concerned with restraint as intervention, depending on the context. Working out a transformative process theory is therefore at least an exercise instructive to either side, and it can offer a way to overcome divides that hamper global engagement with these core constitutional issues.

Introduction

To propose the idea of a transformative process theory is to suggest the kind of marriage that might worry the best friends of the prospective couple.

Of course, process theory and transformative constitutionalism have some obvious things in common. In particular, both are deeply interested in democracy. That is the defining concern of process theory, while in Karl Klare’s original definition, the goal of transformative constitutionalism is ‘transforming a country’s political and social institutions in a democratic, participatory and egalitarian direction’.¹ Both theories also place a special focus on the role of the judiciary in serving these democratic goals and want to understand the judicial role in terms of them. Even their first date, however, would reveal important differences. There are three main areas of disagreement, although they are interrelated and will need some disentangling.

First, process theory typically is interested in limiting judicial power. It wants to identify the situations where courts should intervene so as to make it clearer where they should not, in order to reconcile judging with democracy. Transformative constitutionalism will have no objection to a democratic separation of powers, in principle. But it thinks transformative constitutions require us to reimagine that idea, and it finds too much talk about what courts should not do to be suspiciously reactionary.² It is much more interested in pushing these limits than in policing them: it wants to be on the progressive cutting edge, not trying to find middle ground as Ely was.³ This is a widely noted point of incompatibility between the two theories. As the convenors of a recent symposium put it, ‘most scholars would certainly understand transformative constitutionalism to encompass a more robust role for judicial review that goes beyond even a broad reading of Democracy and Distrust’.⁴

Second, transformative constitutionalism is a proudly value-based account. It sees itself as a theory for constitutional texts such as South Africa’s, which are full of open-ended value terms. It views judicial engagement with values as a necessary condition for judicial enforcement of these texts.⁵ But at least when Ely is holding the pen, the whole point of process theory is to avoid judges making value judgements.⁶ This is widely noted as an obvious reason why process theory does not fit transformative constitutionalism or transformative texts.⁷

The third problem, reflected by the other two, is that the two theories come from different worlds. Process theory looks most at home in its native United States or in systems that adhere closely to a tradition of firmly limited judicial review, such as New Zealand.⁸ By contrast, transformative constitutionalism often defines itself in terms of not following approaches such as the US model or the Westminster tradition, and is instead primarily associated with its South African birthplace and with other Global South jurisdictions.⁹ The idea of a transformative process theory may therefore seem an obvious non-starter. But the divide between the theories is not nearly as unbridgeable as it may seem.

To start with, post-Elyian¹⁰ moves in process theory have already eroded some of the starker differences. Ely’s bid to avoid value judgements has been recognized as dubious and abandoned.¹¹ The idea that judging involves value judgements is no longer a point in dispute between the theories. These recent accounts have also self-consciously sought to broaden the scope of a process-based account of judicial review beyond what Ely cashed out in his US context. This broader global perspective weakens the sense of a sharp geographical divide between an old world of process theory and a new transformative domain.¹² Today’s fears of democratic decay in even the most established democracies have had a similar effect. The situation has challenged the complacent conceit that the kind of limited process theory that suffices in established democracies must be very different from what is needed in the more fragile democratic conditions further South.¹³

For its part, transformative constitutionalism has made fewer moves in the direction of process theory. It has too often remained simply a rallying cry for bold judicial creativity.¹⁴ At its core, however, transformative constitutionalism is a deeply, richly democratic theory. While it may have no patience with the traditional constraints and citations of conventional US-style liberalism, the overlapping concerns it shares with process theory go right to its heart.

The marriage implied by a transformative process theory is thus a more serious proposition than we have tended to assume. Working it out is at least an instructive exercise, and it just might lay the groundwork for a genuinely closer, richer relationship.

A proposal of marriage: The basic line of argument

The idea of a transformative process theory asks each side – transformative constitutionalists and process theorists – to make a basic concession to the other. Each will likely find these concessions uncomfortable, but they are less troubling than they may appear. Working out the details is the task of this article. But it will be helpful to start with a core statement of the two basic concessions that can bring the accounts closer together.

Central to how we think about process theory is that it has a limited account of process – a limited account of what counts as a ‘political process’ and what counts as a problematic blockage or distortion of these processes. This is what limits the judicial role when text does not – either exclusively, as in Ely, or in significant part, as in contemporary accounts.¹⁵

The concession I ask of process theory is to give up on this limit – not to give up all limits, of course, just this one. A unifying transformative process theory needs to start by conceiving of political processes, and their obstacles and distortions, in the way that transformative constitutionalists will think about those things: very broadly.

The transformative vision of process is a richly substantive vision of participatory, egalitarian democracy. It will consider any aspect of a society that stands between it and the achievement of this vision to be, in principle, a matter for judicial concern. If X has an impact on political processes and participation in them, transformative constitutionalists will refuse to accept that X lies outside the judicial role – whatever X is, and whatever its proximity to the conventional themes of process theory or the judicial role, whether it concerns free speech or poverty, gerrymandering by statute or gender discrimination produced by diffuse social habits. This approach is adopted in the full awareness that many things can be linked to political processes and their problems in this way – indeed, it is adopted precisely because of this awareness. The result is a process theory of huge scope, which does not seek to limit the judicial role in the theory by means of a limited definition of processes or obstacles. I will refer to this as the transformative understanding of process.

The sheer breadth of this understanding may unsettle, as I said. But this first impression is quickly softened. Adopting this broader account of process does not alter the fact that many constitutional systems just do not go this far, as a matter of local law and legal culture. A transformative process theory is an account of how to support political processes. It is not – indeed, if it wants to be a global theory, it cannot possibly be – a description of how much any one system provides for this account or acts on it. And the claim that so many things fall within the scope of a process theory is only a claim that these fall within judicial jurisdiction, not that this jurisdiction must be exercised in all cases, or exercised in a strong-form way.

This, in turn, is the subject of the countervailing concession that I ask of transformative constitutionalism. It must take the issue of when the courts should not act, or not act boldly, far more seriously than it usually does. It must do this on its own terms because it would be self-defeating for a transformative court not to adopt the best available means to its transformative goals, so this must be a constant (if sometimes hard) question. This point is sharpened in the context of this article because here the transformative goal in question involves democratic processes. If that is the goal, then success cannot be extensive judicial intervention informed by a rich understanding of democracy. Success is richly functioning democracy, which courts can as a result often leave substantially alone. That is the ideal outcome. Like all ideals, it is aspirational, and courts will often pursue it in conditions that are far from ideal. But if this is the ideal, then transformative constitutionalism must be strongly interested, where possible, in how it should limit judicial intervention.

Transformative constitutionalists have tended to be suspicious of too much talk of judicial limits. But the more we focus on democratic processes, rather than the US/liberal conventions with which process theory is customarily entangled, the more receptive transformative constitutionalism will be – or at least can and should be. And the limits I have just outlined come from within transformative logic, about what best serves to protect and promote that broad understanding of process. I will refer to this, drawing on a much less famous phrase of Karl Klare’s, as democracy-seeking review, to mark its greater breadth.¹⁶ But it also reminds transformative constitutionalists that what they are seeking is democracy, not judicial expansiveness as an end in itself. This is the way to become the fully rounded account of constitutionalism that the theory, by its name, holds itself out to be.

In what follows, I work out these arguments in more detail. In the next section, I consider the moves this line of argument asks process theorists to make towards transformative constitutionalism. In the following section, I reverse the direction. I hope to convince you that the commitments of process theory and the goals of transformative constitutionalism need not divide nearly as much as they are currently seen to do. And maybe, like the proverbial marriage of opposites that works, they can unite more than anyone thought possible.

A transformative understanding of process

In its most familiar form, process theory is a limited account of judicial responses to discrete deviations from a basically functioning democratic status quo. From a transformative perspective, this vision is too narrow and complacent. Transformative constitutionalism is broadly aspirational. It believes that the legitimacy of the constitutional order depends not just on preserving the status quo, but on striving constantly to improve it.¹⁷

This perspective is natural in emerging democratic systems, where the status quo may not be anywhere near comfortably democratic. There, the job at hand necessarily implicates building democratic capacity and not just protecting something already largely in place.¹⁸ A purely defensive account that omits judicial work to promote democratic processes will not suffice in this context. But the appeal of this perspective is not limited to emerging contexts.¹⁹ All democracies can be improved, and all are also in continuous need of renewal.

This argument has the potential to expand the judicial role greatly – but in this respect it is not out of step with today’s post-Elyian political process accounts.

Expanding process

Stephen Gardbaum’s seminal article on comparative political process review argues that if we step outside Ely’s US context circa 1980, we will find other kinds of process problems and other kinds of judicial responses to them. If we are to offer a global comparative account, we should include these interventions, not leave them out because they are not in Democracy and Distrust. ²⁰ The result is a broader account, more globally inclusive and globally relevant, but with a logic that leaves it open to further broadening still.²¹

For example, Gardbaum argues that if we look to systems such as Colombia, Israel and South Africa, we will find courts intervening in internal legislative processes to address obstacles to democratic accountability and deliberation. This kind of intervention was not Ely’s focus, but Gardbaum plausibly argues that since these interventions also contribute to important process goals, a global process theory should include them too.²²

This kind of logic is easy to take further. Gardbaum’s argument, if generalized, tells us that we should reject a traditional limit on the scope of process theory review if (1) in fact, process-relevant abuses and distortions are occurring on the other side of that limit and ( 2 in practice, we see that courts can do something meaningful to respond to those abuses because they are in fact doing so in some jurisdictions. If courts should engage internal legislative procedures because justiciable and process-relevant abuses sometimes occur there, should they engage all comparable abuses in any context in which they occur?²³

Similarly, Rosalind Dixon’s recent work on responsive, representation-reinforcing judicial review focuses on particular defects – blind spots, burdens of inertia, monopolies – in the legislative process. She also recognizes that the legislature is not alone in having these defects, and that judicial review itself can pose comparable problems in the wrong hands or under the wrong conditions.²⁴ But if these things are problematic in both those contexts, shouldn’t a process theory engage them in any process-relevant context in which they arise? For instance, why not in private contexts?²⁵ After all, it is with non-state economic power that we primarily associate the word ‘monopoly’. Since entrenched social elites tend to be bad at seeing or engaging those who do less well under the status quo, it is natural to think of (functional equivalents of) blind spots and burdens of inertia here too. This kind of logic readily embraces actors like political parties or companies – the latter most obviously, though not only, of certain kinds (media companies, for example) or in relation to certain activities (lobbying, campaign contributions, attitudes to unions) with obvious political salience.

This kind of argument is natural to transformative constitutionalism, which rejects the public/private divide. It broadens the scope of process theory by broadening the definition of what counts as a ‘political process’. But from the transformative perspective, it makes little difference if we resist this and stipulate that ‘political process theory’ is about only the processes most classically associated with the theory. Broadening would still happen, and just as easily, because many things are relevant to political processes narrowly understood.²⁶

Consider, for example, the relationship between the non-traditional area of justiciable socio-economic interests and the classical issue of elections. We might think of narrow protective links, such as the abusive disruption of transport services or electricity or water in opposition areas on voting day. We might include broader suspicious patterns, such as generous distributions of government benefits at strategic moments to win over voters, or differences in the welfare services provided in loyalist areas as compared with opposition areas. We might address problems that would usually be seen as obstacles more than abuses, by requiring free public transport on voting days or allowances on election day akin to those often provided for jury service, or that mobile phones must provide free access to electoral content. We might go all the way to arguing that any judicial intervention protecting socio-economic interests is process-relevant because poverty generally restricts a person’s ability to participate in society’s political processes and socio-economic ‘haves’ hold incumbent advantages over the less resourced.²⁷ It is certainly true that, as the links get vaguer and more diffuse, the link may get less interesting. A process-based approach to socio-economic rights, for example, is more interesting where it makes us think about specific links between the two that we might not have seen, rather than just reminding us that poverty is broadly crippling. But that does not change the fact that these links can be very real.

From the transformative perspective, broadening is also natural in other ways. Transformative constitutionalists will reject any embrace of judicial work to protect processes that excludes judicial efforts to promote their functioning. They will understand the judicial role to include working to improve the quality of processes – promoting meaningful participation, for example – and not just to protect them against abuses.²⁸ Indeed, they will see these goals as ultimately inseparable. This, once again, is particularly obvious in the context of an emerging democracy where protecting processes is necessarily a matter of building democratic capacity, not just preserving the status quo. Once again, though, the argument applies everywhere because democratic capacity everywhere needs renewal. For example, Dixon’s argument that responsive judges should design interventions in ways that promote civil society capacity or encourage public acceptance among those on the losing side of an issue illustrates this natural blurring of protection and promotion.²⁹

Finally, something very similar would happen if we moved from processes to minority rights, another of Ely’s concerns that Gardbaum’s account, for instance, leaves to one side.³⁰ Ely’s account was focused on ‘discrete and insular minorities’, informed above all by the experience of Black people in the US Southern states. But it is easy to think of other kinds of structural obstacles to certain groups, including majorities.³¹ Given the scope of modern rights review, it is even more obvious than in the context of processes how much this idea can expand. Indeed, transformative constitutionalists will see rights as interdependent, so any interest in classical political rights such as free speech or the right to vote will be seen as linked to an array of other rights, and vice versa.³² So this argument easily expands too.

All this is suggestive, rather than a systematic account. But it will suffice to illustrate why contemporary process accounts, on their own logic, are very susceptible to wall-to-wall broadening if we look for it, as transformative constitutionalists strongly will be. Those who take process theory as a starting point, however, will usually not be, which is why these wider implications are not usually cashed out, or are raised as objections. The traditional constituency of process theory will be (at least) uncomfortable with this race to the horizon. They will frown to see how the idea of a court intervening to respond to discrete abuses seems to have become a wall-to-wall judicial mandate to seek a better society for all. This may well seem to cross lines that should not be crossed and – perhaps more worrying – to reveal a troubling indifference to drawing any lines at all.³³

Their dismay will deepen when transformative constitutionalism indeed shows signs of not caring very much about all this. It will be intent on addressing social problems and will want courts to spearhead that effort. It will, therefore, be perfectly happy with a very broad understanding of potential judicial jurisdiction, and will have little interest in narrowing it down because this might enable traditionalists or limit future judicial freedom of action. This is what is usually seen as grounds for an irresoluble stand-off. And the two sides can certainly have a stand-off here if they want one. But they also do not have to.

Start with the fact that, from the transformative point of view, the kinds of expansion that we have just seen are not just strategically welcome. They represent consistency. If we say that we care about political processes, if we say that we want them to work well and fairly and we don’t want them blocked and distorted, then a consistent theory should care about obstacles and distortions wherever they arise. And if this is the judicial task, then judges should engage with it wherever the need arises. There is no difference, as a matter of normative process concern, between – for example – a maleficent government statute that places formal limits on women’s voting rights and a drifting public indifference to doing anything about social obstacles to women’s equal political participation. There are, of course, any number of practical differences between the two, which may affect how a court responds to them. But the one is not necessarily any less of a process problem than the other. This is what the transformative constitutionalist will insist upon our seeing. And if that is true, then it looks arbitrary to draw lines between such situations on process grounds and claim that courts should act only on one side of those lines. Social oppression can disenfranchise as effectively as a statutory ban, and poverty can dilute agency no less than gerrymandering, so courts should be concerned with the one no less than the other.

The transformative constitutionalist will therefore argue that if you are not open to this, if you want to resist and cabin the idea that courts should engage with process-relevant issues wherever they arise, then it is not because these problems are unrelated to process interests. It must be for some other reason, based on something else.

Traditionally, that ‘something else’ might well have been the ability of courts to engage with some of these kinds of issues.³⁴ The contention, in other words, would be that something like poverty might well have serious impacts on political participation, but poverty is not a problem for courts to deal with. But if we take a Gardbaum-like look around the world at what courts are in fact doing, we will see that courts are engaging many poverty-related issues. There are concerns and limitations, of course, but more than enough is going on globally to rebut any sense of an absolute bar to justiciability on this kind of basis.

If you resist these expansions in principle, then your principle must be something other than a process theory principle. Perhaps you are a species of traditional liberal. Perhaps you think that courts should not engage ‘policy issues’ or ‘political questions’ or socio-economic rights, or should not get involved in promoting as opposed to protecting rights, or should not cross some version of the public/private divide. Perhaps you mistrust all these sharp categories but still want to retain some sense of what they are trying to capture (in which case, you may find the approach to limits in the next section more appealing).³⁵ Or perhaps you experience limits like these not as a matter of personal conviction, but as a matter of the local constitutional law and legal culture of the system in which you exist. If judges in your system lack the power to invalidate legislation or will only engage socio-economic issues in very limited ways, you will not as a matter of professional legal practice be able to follow process arguments wherever the transformative understanding may take them.

For transformative constitutionalism, this is usually a reason to reject traditional US-style liberalism, chuck all the old distinctions like the public/private divide towards the dustbin of history, and ignore or disdain constitutions with more limited visions of the judicial role. But the conclusion I am driving at here is different and less confrontational. The point is that we can adequately express any disagreements here without having to take a normative stand on what counts as a ‘political process’ or what is relevant to the functioning of one. The transformative understanding of process is not something we have to put in dispute.

Instead, we can express disagreements as debates about judicial legitimacy and capacity – not as an absolute matter, but as a debate about what courts can usefully do in a given context, which includes what local constitutional law and culture accepts.³⁶ In systems with broad texts where legitimacy depends on a transformative approach to constitutional law, this may well pose no limits to the transformative understanding of process. In systems with more limited traditions, the room to act on the transformative understanding will be more limited. But it is simply more accurate and honest to describe these as systems that limit judicial engagement with parts of the field of process issues. If you decline to have your courts get into all the ways in which female political participation is restricted by social structures, then the basis for this is that decision – not that female political participation is not restricted in these ways or that this is not an important process issue.

All this naturally means that the scope of the transformative account of process will exceed even the potential scope of existing constitutional provisions in many systems, let alone current practice. But that ship has already sailed. Once we move to globalize process theory, it ceases of necessity to be an accurate description of any one system. It will simply become an account of process-supporting interventions, whose prescriptions different courts in different systems will follow to varying extents.³⁷ In expansive transformative texts, whose scope more or less matches the ambition of the transformative account of process, it will be a way to interpret the constitution. In more limited texts, it will either represent a resource for activists, offering guidance on where reform or constitutional (re-)interpretation should go, or it will simply represent an account of process issues in general – for which, in that context, solutions must be sought somewhere other than the courts.

Back to Ely

The effect of the argument just made is that the account of process itself no longer does the work of limiting the judicial role. That may seem to abandon the point of a process theory; however, this is in fact what is going on, unremarked, in Democracy and Distrust itself.

For example, Ely recognized that something like poverty can be an obstacle to political participation. He saw efforts to help poor people as a failed but ‘glittering crusade’ on the part of the Warren Court.³⁸ The reason socio-economic issues are marginal in his book is not because he thought them unconnected to process or because he thought courts could play no meaningful role in this context. They are marginal because the crusade failed: because these issues stayed marginal as topics of judicial work in the US federal system about which he was writing. What is doing the limiting is not the idea of political processes or what courts can conceivably do about them. It is the sense of what is accepted in the system.

Another illustration is the exception to the general absence from Ely’s book of the judicial intervention in internal legislative processes addressed by Gardbaum. The one example of this that Ely did consider was the review of legislative refusals to seat a representative elected by the people – because this is a rare example where the US Supreme Court has been willing to intervene in Congress’s internal decision-making on procedural grounds.³⁹ If Ely looks limited on this point, it is once again not because this context is of limited process relevance (as Gardbaum shows); it is because US practice is limited on this point.

This is a crucial and unremarked part of the explanation for the pattern we see: that once process theory is considered globally, it tends to expand in ways it finds hard to control theoretically. The reason is that normative ideas of process were never doing the work to limit the scope of Ely’s account. When process theory is deliberately stripped of its pre-1980 US context, in order to apply it globally, it is being stripped of what was doing that work.

Put another way, Ely’s book does not, in fact, produce an account of interventions that safeguard political processes, and then use that account to define the scope of the judicial role. Instead, he is taking the fairly limited conventional understanding of judicial review accepted in the US federal system and arguing that all of it can be explained and justified in process terms. He is showing that a process-based explanation can cover all the aspects of judicial review widely accepted in the United States. He is not, in fact, showing that it implies only these things, or that only these interventions can be justified in process terms.

This is why Ely’s theory is appealing, more broadly, to the intuitions of those who hold a fairly conventional liberal understanding of judicial review. That fairly conventional liberal understanding is, in fact, not the outcome of that theory, but its key moving part. It also makes sense in light of Ely’s aim of minimizing judicial discretion. That is achieved if judges stay close to the text or if, when the text is open-ended and so opens up potential space for discretion, they stay close to settled understandings in their system. So it makes sense that what is doing the real work in Ely is either the text or that more settled understanding.

If process theory is to come of age and travel the world, it needs limits of its own – not limits that actually come from US legal culture or its conventional liberal consensus. We have seen how process theory can naturally expand in the service of its own process logic. It needs to do the same, in the service of the same logic, to develop its own account of its limits. And since this is something that transformative constitutionalism also badly needs to do, this shared need is one more factor bringing them together. As in so many romances, in the crisis the couple will find what they are seeking by looking inside themselves.

Democracy-seeking review

Transformative constitutionalism was born with a genetic resistance to traditional limits on the role of courts. This resistance creates unnecessary distance between transformative constitutionalism and contemporary scholarship on restrained judicial review – work on which transformative constitutionalism, for transformative reasons, should be drawing.

Transformative constitutionalism began as an act of resistance against the Westminster tradition of judicial deference.⁴⁰ It was no less suspicious of conventional US constitutional liberalism. Klare, a US Critical Legal Studies scholar, labelled his account ‘post-liberal’ and repudiated an array of canonical Anglo-American authors, Ely among them.⁴¹ Process theory’s enmeshment with US-style liberalism had made transformative constitutionalism allergic to it. In the previous section, I unpicked these connections and argued that our understanding of process can come apart from liberal convention. That will already start making life easier for the transformative constitutionalist. But it is only a first step.

Transformative constitutionalism has also spent a lot of time rebutting general arguments that capacity and legitimacy constraints preclude judges from doing non-traditional things such as enforcing socio-economic rights. These rebuttals are correct, and have made space for valuable creativity.⁴² But the focus on rebutting scepticism has led transformative constitutionalism to elide the dynamic, contextual question of when judges should intervene.⁴³ Instead, as I have written elsewhere, it has tended to put judges on a bold standard.⁴⁴ By this I mean that expansive judicial creativity in the service of transformative ideals becomes the sole measure of judicial value and the sole test of whether a judge is committed to transformative constitutionalism. Too much talk of limits and restraint gets coded as the mark of the reactionary, and so is a danger sign, something to be resisted.

This is understandable as a matter of strategy. Transformative constitutionalists do not like to talk too much about limits in case this enables lazy traditionalism and undermines the effort to win acceptance for wider transformative judicial activity. But it has impoverished transformative constitutionalism. It has turned a vision of social transformation into an exercise in transforming courts, neglecting other institutions.⁴⁵

As a theory, transformative constitutionalism insists that people be open-minded about traditional limits, and about the possibilities of expansive judicial creativity. It needs to be similarly open-minded about arguments for judicial restraint, made not out of traditional habit but instead on the basis of what will best serve transformative ends. Ironically, given how intently and productively transformative constitutionalism subjects old habits to critical scrutiny, it has itself become court centric by habit rather than transformative logic. The concept happened to enter the world in a talk to judges, and the original article therefore focused on adjudication (while clearly seeing beyond that context). But that coincidental focus has remained, ill-fitting the theory’s much broader democratic commitments.⁴⁶

Those commitments also run deep. Transformative constitutions, which typically arise in post-colonial contexts and are deeply concerned with structural inequalities, place great value on democratic ideas. That includes a foundational respect for political agency, driving a project to build richly egalitarian and participatory democracy. India is an early and archetypal example.⁴⁷ The Indian Constitution of 1949 introduced universal democracy more or less overnight. The idea of recognizing Indians as democratic agents, in the wake of colonial rule predicated on denying that agency, must rank high on its list of transformative priorities. And while there is much that judges can and should do to contribute to this project, it would be self-defeating if democratic capacity were only supported and not, in the end, exercised.

The same applies, by extension, to elected institutions, and most obviously the legislature. Respecting that institutional authority is part of vindicating the agency of the voters. Doing so also gives effect to the constitution because any transformative text includes democratic institutions of government, and impliedly expects judicial authority to be constructed on a basis compatible with those institutions. This is a part of what judges must work to build.

None of this, of course, represents an absolute argument. It is not an argument for judicial restraint in general, nor a suggestion that this must be some kind of default. The claim is only that this is why, for the transformative constitutionalist, the default also cannot simply be expansive judicial action. The bold standard is a very imperfect measure of transformative value.⁴⁸

This is why all transformative constitutionalists must be interested in – among other things – the concerns of process theory. They must be interested, on transformative grounds, in the value of judges leaving political processes alone or intervening only lightly.⁴⁹ Democracy can be in need of protection and promotion, but it also needs to be allowed room to work and establish itself.⁵⁰ Put another way, if you seek democracy, one of the possibilities you must be prepared for is that you will find it.⁵¹

Process theory is not, however, always presented in a way that makes it easy for transformative constitutionalists to draw upon it. For example, the last sentence of Democracy and Distrust says that ‘constitutional law appropriately exists for those situations where representative government cannot be trusted, not those where it can’.⁵² This resonates with a classical idea in which rights and constitutions cover only certain limited things seen as needing special entrenched protection. But this is hard to fit with more expansive modern texts, such as those of India and Germany, among many others. In systems like these, it is not at all easy to think of a publicly significant but non-constitutional issue: the constitution is understood to apply to more or less anything important. This is, of course, the world in which transformative constitutionalism feels most at home.⁵³

However, this kind of obstacle is easily avoided, if we try. It is easy to express the arguments both for and against judicial review in constitutional terms, and then to balance them. That is old news to lawyers. But it can be a salutary exercise for transformative constitutionalists. Ely is easily read as telling lawyers to look out for democratic problems, which judges should then address. Such problems are easy to find in emerging systems – certainly if one is guided by the broad transformative understanding of process. So it is easy to conclude that judges have a lot to do. An explicit balancing exercise compels transformative constitutionalists to consider and weigh the transformative reasons against judicial intervention. It nudges a theory on the bold standard towards a more finely calibrated judgment. But doing so requires more work to cash out all the elements in the equation.

Distrust and trust

We have seen how the transformative understanding of process has a nitrous tendency to broaden the scope of judicial review. It expands, from situations where political institutions should be distrusted, to cover any justiciable, process-relevant problem. But a return to thinking about distrust can assist the transformative constitutionalist to make decisions about how courts should intervene. Weak-form review comes from systems transformative constitutionalists tend not to look at, but there are important things to learn here.⁵⁴

Sophisticated weak-form accounts point out that motives matter, as they did for Ely. A rights violation stemming from malice is more likely to require judicial intervention than the same factual violation stemming from more innocent causes. A busy legislature may simply not have noticed that a statute it was passing had an unconstitutional effect. Or it may have seen the problem but not seen an easy way to avoid it without impairing the larger aims of the statute.⁵⁵ This kind of mistake is perhaps particularly likely in a typical transformative setting, where the ratio of pressing tasks to institutional capacity can be especially high.

In these scenarios, there is often no reason to think that the legislature wanted to limit the right in question. If that is true, judges might need to do no more than bring to the legislature’s attention the problem or the workaround that it missed.⁵⁶ And by doing so, if the assumptions hold, the court will strike an optimal constitutional balance. It will identify a rights violation, take plausibly effective steps to have it remedied, and support the elected legislature’s constitutional status as a key rights-defending body. This is a good illustration of why weak-form arguments, which pay more attention to judicial deference even where rights have been violated, can help courts to better serve transformative goals.

Transformative constitutionalists tend to be alive to the importance of designing remedies flexibly to fit each situation.⁵⁷ But the point goes beyond remedial design. It also affects whether a court intervenes at all: it encompasses decisions by a court to review a complaint, find at least some evidence of a problem, and nevertheless issue no remedy.

This possibility matters, perhaps especially in emerging systems, because there will so often be some reason to think there is a problem or that the institution to which the court defers will not respond adequately. Emerging systems are characterized by greater social problems and weaker institutions. This may also more often be the perception there.⁵⁸ For example, influential accounts of the role of courts in emerging democracies have a tendency to focus on large judicial interventions to respond to large political threats. Judges in these accounts oversee critical election rules, party bans and lustration laws; adjudicate dramatic election disputes and high-level separation of powers clashes; and deploy basic structure doctrines against dominant parties.⁵⁹ Transformative constitutional scholarship, in turn, typically is focused on the lack of adequate transformation in societies that badly need change, with the clear implication that the powers that be are not doing enough and so courts need to do more.⁶⁰ One also suspects that, often, transformative constitutionalism is court centric for the simple reason that the legal scholars who write about it tend to trust courts more.⁶¹

Not that this is necessarily wrong, of course. Sometimes we do need to look to courts to engage big problems. And some of the situations where this is not the case are also often easy to spot – sometimes, for example, courts themselves cannot be trusted because they are part of the abusive pattern.⁶² But we can adjust for this. The problem is the less clear cases, where we are uncertain about problems or how other institutions will respond to them, and so we are uncertain about what courts should do and how interventionist they need to be.

A crucial difficulty here is that central to what makes a democracy less established is that people have less faith in it. The risk of a distrust-based account in such a system is that its scope will tend to expand as the level of distrust does, and distrust lies partly in the eye of the beholder, especially at moments of uncertainty. That risk becomes a liability because such a democracy will not get more established unless the level of trust and faith increases – something that the distrust expressed or implied by judicial intervention can impair. This is a key paradox for a distrust-based trigger for judicial process intervention in these systems.

To better respond to this, a transformative process theory needs a positive account of trust. It should not just specify distrust-situations and leave the rest as trust-situations by implication, as Ely did; instead, it should add an account of the constitutional value of trust as an element of its own in the balancing equation.⁶³ This will address the fact that some reason for distrust will often be, or perceived to be, present, but this should not always be decisive because it may be outweighed by trust-based considerations. It will better allow for the situation where distrust and trust considerations are both present simultaneously, something a distrust-based account will struggle to capture.

An illustrative example I have studied closely arose in the 1999 elections in South Africa.⁶⁴ The government had announced a new voter ID requirement that many voters, a year out from the election, did not meet. There were claims that this would disproportionately exclude opposition voters, so there were typical process theory-type reasons for distrust. There were also less abusive concerns: that voters would not understand the new requirement, or that they would be unable to obtain the new ID documents in time from the rather troubled government department responsible for issuing them. However, at the time of the case, all the concerns were unclear or speculative: the situation the Constitutional Court had to confront was chiefly characterized by uncertainty and risk.

A distrust-focused account will certainly identify some reasons for concern here. In this case, why not intervene on the basis of better safe than sorry? Surely we should be especially cautious with something as important as voting rights?⁶⁵ South African scholars have overwhelmingly taken this view. But the problem is not nearly so easy. For one thing, South Africa really did need a better voter ID system, but the new one was launched rather late and there were, as so often, implementation concerns. The situation called for weighing up the risks that the new ID system would exclude voters against the risks to the credibility and efficiency of the election if the reform were not implemented.

For another thing – and here trust considerations are particularly manifest – the usual case for intervention under-weights the effect of the court intervening in the management of an election. Elections depend on trust. Judicial intervention implies that the government’s management of the election cannot be trusted. That is the signal that it sends, however carefully the court expresses itself.⁶⁶ If a government indeed clearly cannot be trusted, then courts may have no choice but to say so. But if the situation is more uncertain, and if a case seems more about implementation problems than abuse, then it is less clear that the value of intervention outweighs the costs of sending that signal. Framed more positively, defaulting to judicial intervention means missing opportunities to build trust by expressing it. By signalling that the government can be trusted to hold fair elections, the court can support the project – elusive and vital – of generating public faith in a fragile democracy.

Admittedly, we could understand this situation simply as one where distrust-reasons should prompt a court to give the situation a searching look – not that this look necessarily has to lead to intervention. Ely’s account can accommodate my example in this way. But his account is rather ambiguous on this point.⁶⁷ It does not bring out the positive trust-based considerations in the way needed to discipline transformative constitutionalists who are intent, for their own reasons, on defaulting to judicial intervention. I elsewhere defend the conclusion that judicial non-intervention – making no order – was the best transformative response to the 1999 case. That means being open to the possibility that the institutions primarily responsible for elections could address the problem themselves, and would be the stronger for doing so. The argument is impossible to defend unless one takes account of how non-intervention is not the same as doing nothing when it is an act of constructive trust. It is a way to build political processes by affirming that they can be relied upon, such that a court does not need to intervene despite the presence of problems. It will not always be realistic for courts even to try for this and, even when it is realistic, trying will involve risk. But a political process equation should not fail to include a term for this value.

Thinking along these lines can also assist with remedial calculations where the court does intervene in some way. The story of same-sex marriage recognition in systems like Colombia, South Africa and Taiwan poses a somewhat revised version of the scenario where the judge is deciding whether simply to refer a problem to the legislature, or fix it.

Each system saw a same-sex marriage judgment in which the judges referred the issue to the legislature. But since there was some doubt about how the legislatures would react, each court referred the issue subject to guidance and under a deadline, with a default remedy that would come into effect if the legislature did not act adequately in time.⁶⁸ This approach involves the risk that the legislature will not do so, meaning the only effect will be to delay the remedy and create legal uncertainty. That was the Colombian experience.⁶⁹ The pay-off, if the legislature does respond adequately as it did in South African and Taiwan, is multiple. It affirms the idea of the legislature, and not just the court, as a constitution- and rights-defending body. It builds comity between the branches.⁷⁰ It also obtains the stronger protection and symbolic imprimatur of legislative, not only judicial, recognition of same-sex marriage rights. This last point shows that taking trust into account in relation to political processes is not necessarily something to be weighed against rights protection.⁷¹

These examples – the 1999 South African election and the same-sex marriage cases – have been selected as favourable illustrations of situations where courts could emphasize trust amidst some reasons for distrust, and where there was value in doing so. But the implications of the argument matter in cases of all kinds: trust arguments do not just matter on the less-interventionist side of the spectrum. On the opposite side of that spectrum lie cases where other institutions fail systematically.⁷² Even here, where there may be no other functioning institution to trust, trust-building arguments still matter. For example, they counsel a court to intervene, from the beginning, with the goal of trying to reach a point where the issue can be handed back to the ordinary political processes.⁷³ That includes intervening, where possible, in a spirit of collaborative assistance rather than high-handed suspicion.⁷⁴ Judges can also act here in ways that engage rather than exclude the general public – for example, by holding open public hearings.⁷⁵ The argument for showing trust is also closely related to the concern, found even in the most established systems, that too much judicial intervention will lead to democratic debilitation. To remain robust, democracy needs to be used and relied upon, whatever its age and pedigree.⁷⁶

All these arguments reflect the simultaneous presence of both distrust- and trust-based arguments. I have spent more time here on the cases at the restrained end of the spectrum only because transformative discussion focuses on these less, and because the limitations of a distrust-focused account are more important and visible here. I entirely acknowledge that balancing distrust and trust considerations can entail hard decisions weighing uncertain things in tricky circumstances. But Ely’s efforts to minimize judicial discretion were never convincing, and it is no part of my argument that it is easy to be a judge. We do not make it any easier by underplaying in theory what makes it hard in practice.

Judicial processes as political processes

If we were looking down on a moving map of the political processes of a society, it would be hard to see most courts today as anything other than a substantial influence on the picture. That implies treating courts as potential subjects of process theory concerns, and not just as the guardians of the theory.

That may be particularly obvious when judges act abusively, including as agents of self-serving political incumbents.⁷⁷ They then manifestly represent process obstacles. But a judge does not need to be abusive to be distortionary. Judges can intervene in ways that do not work or are actively counter-productive from a process perspective.⁷⁸ This can also be true in an all-things-considered sense, where an intervention has a beneficent process impact considered narrowly, but not more broadly. For instance, a court may intervene to assist a group of litigants for good process reasons, but in ways that it does not or perhaps cannot extend to other similarly situated groups. That will create a process distortion.

The widened scope of the transformative understanding of process only reinforces this impression. The more a transformative account insists on a broader understanding of ‘political process’ or of what counts as a process-relevant problem, the harder it becomes to see how courts themselves can be excluded from this logic. Ely treated courts as outside the ‘political process’. But for a transformative process theory this starts to look like hypocrisy. If other traditional dogmas must give way under transformative logic, why not this one?

The conclusion is not that we must collapse the distinction between courts and the elected branches. Courts are meaningfully different institutions. But a transformative process theory, especially, must acknowledge how much its own arguments also apply to the judicial process. The wider the scope of judicial intervention, the more overall political success can depend on accessing judicial processes, and the more it will be affected by distortions there.

As with positive trust arguments, there is already plenty of scholarly work on the implications of this. Transformative constitutionalism, for example, has long been interested in issues of access. While the focus is sometimes just on making it easier for cases to get to court, more critical work worries about whether access to justice measures end up empowering more resourced actors who are better able to use courts.⁷⁹ Another example, from the traditional constituency of process theory, is the rich interest in dialogic approaches that reduce the finality of judicial interventions, offering the possibility to correct mistakes and reduce distortions by giving other voices more room to respond to judicial decisions.⁸⁰

Arguments such as these are well known. They offer many ways to understand how judicial processes themselves can pose process-relevant obstacles, and how judges and others might respond in order to manage these impacts. So thinking about subjecting judicial processes to the same kind of process-based scrutiny as everything else is not so much about breaking new ground as it is about reminding a transformative process theory why, by its own logic, it has to engage these issues. The traditional response to these types of concerns is simply judicial restraint.⁸¹ This is a response that transformative constitutionalism will reject. But with that rejection, and the advocacy of a much more sweeping judicial role, also goes the easy ability to treat courts as providers, but not subjects, of process-based scrutiny. An honest transformative process theory, which follows through on its own logic, also has to reject this comfortable idea. In doing so, it will allow for the expression, from within transformative logic, of valid concerns about the hazards of judicial intervention in democratic society that rightly matter to more traditional process theorists.

Popping the question

It may not only be this last point that is useful chiefly as a mental reminder or a thought experiment. Perhaps that is true for the whole notion of a transformative process theory. Perhaps this article is simply an exercise in bringing together two very different people, who will leave as two very different people, but not before discovering more common ground, and more reasons to meet and talk again, than they thought existed. It is certainly true that the account I have sketched leaves a great deal open to be debated, or decided differently by different legal systems. It may seem less a theory then a shared framework for discussion, achieved only by framing problems in broad and often rather vague terms.

But it may also be more than this. A process theory that expands beyond the United States, defines itself independently of old liberal habits, and opens itself to consider the many things that can distort political processes, is a changed process theory indeed. A transformative account will have come a similarly long way if it has learned to take a serious interest in non-intervention and weak-form review techniques, and to see courts as one institution among others. Each, in changing in ways implied by the other, acquires new capacities to apply itself, in sickness and in health, wherever democracy shall live. And each, in changing, may become truer to itself: truer to political processes and what it really takes to safeguard them, truer to transformation and what it really takes to produce it. They may have grown up in different traditions, speak different languages, have different friends. But when we see what they share, and how deep to the heart it runs, we might understand how an outwardly surprising marriage could be fulfilling to both partners – and how the union might also serve to bring a little closer together the different worlds from which they come. Although their marriage will surely have its conflicts, it might offer to each something that it has been missing all along.