1.1.1 Constitutional Mechanisms of Personal Control

The threat of removing rulers has long been the most important check on governmental power and the primary means to combat bad governance. Today, the obvious mechanism for removing bad rulers is an election. As Karl Popper argued – and Schumpeter subsequently affirmedFootnote ⁷ – democracy ensures that voters can remove the government without bloodshed, that is throw the rascals out.Footnote ⁸ Inefficient and incapable politicians will pay the cost at the ballot box and will simply be voted out of office, or so goes the theory.

Yet, we know that reality does not always neatly correspond to this idealized template. Elections provide some measure of governmental accountability, but they provide no comprehensive protection against politicians and administrators who fail to do their jobs conscientiously. Elections and election results are dependent on a host of factors, and existing failures are likely to be only one argument among many in the competition for votes. Different political regimes also come with different degrees of accountability. The possibility of holding politicians accountable depends on many things, including on how clearly responsibility can be attributed in the first place.Footnote ⁹ Coordination problems and capacity issues may lead to inefficiency and sometimes failure, which in turn may be hard to attribute to a particular government and harder still to remedy. And in some regimes, democratic processes themselves are largely dysfunctional, and thus elections provide only very limited means to change governmental policy and improve administrative capacity.Footnote ¹⁰

Of course, elections are not the only means of controlling governments and removing incompetent or unwilling officials. In parliamentary systems, parliaments hold the power to force governments, and sometimes individual ministers, to resign. This is not true for presidential systems, but here too, legislatures typically have some means to remove heads of government or state and sometimes other officials by impeaching them. Not all of these mechanisms are easy to operationalize in practice, however. It is rare for parliaments to force the resignation of governments as this requires the majority party to act against its own party members. Where parliaments take that step, it will typically be in situations where a political coalition has come apart – but we would expect a governmental failure of some magnitude before this would be the response to it.

Something similar is true for impeachment provisions. Whether they are suited to dealing with cases of dysfunction or failure will depend both on their concrete framing and interpretation by the relevant actors as well as political will. As Tom Ginsburg, Aziz Huq and David Landau argue, impeachment today often serves as a means not just to remove officials in cases of criminal offences but to address broader political crises where incumbents have lost support, which may involve dysfunction.Footnote ¹¹ But while some constitutions explicitly include broader clauses, allowing for impeachment in light of persistent governmental dysfunctions, not all do. As a result, we sometimes find arguments for an expansive reading of impeachment provisions.Footnote ¹² But impeachment often comes not just with high legal but also practical hurdles, given the influence and interests of political parties. The US example is a case in pointFootnote ¹³ and by no means exceptional, if we consider, for example, the South African ANC’s refusal to impeach then-President Zuma for corruption. Political parties often have little incentive to impeach their own leaders. Given that impeachment can be abused, the high hurdles that come with it are also not unreasonable, especially considering that (in presidential systems) it undermines the democratic choices of voters by removing directly elected officials and so should be a means of last resort. As a matter of addressing failure, impeachment may also not be the ideal tool because failure will often be less a matter of bad intent than a mix of incapacity, neglect and coordination problems. Add to this that both no-confidence votes and impeachment of central figures are nuclear weapons in politics, triggering major changes.Footnote ¹⁴ As such, they will likely be deployed only to deal with problems of a certain magnitude rather than as a tool to address discrete cases of failure and dysfunction.

Within the administration, too, there are means for removing officeholders. Administrative oversight typically allows superiors and sometimes ministers to replace inefficient or corrupt officials with other more suitable candidates, though not all officials can be removed easily or without specific grounds for doing so. Typically, political appointees who ‘serve at the pleasure of the President’ may be dismissed more easily, whereas career civil servants enjoy stronger protections, both against being moved to different offices and even more so against dismissal. This would suggest that systems with a higher number of political appointees – such as the United States – overall do better than those operating largely with a career civil service, like Britain. Often, however, we think the reverse is true, and the more recent political science literature offers support for this intuition. Thus, for example, David Lewis shows how politicization in the United States enables and drives patronage appointments as well as leads to higher turnover in staff, both of which often hurt performance in the longer run.Footnote ¹⁵ And there are a range of factors that influence whether and how politicians might get involved in such oversight in the first place, including the broader political context, the nature of the problem, time pressures, etc.Footnote ¹⁶ Not all decisions on personnel are made by politicians, of course, but overall it is often difficult to dismiss career civil servants without a specific cause, which may be hard to establish. And moving them to other offices, if possible, may create new problems. More importantly still, an exchange of personnel will not always suffice if the problem at hand is structural and thus not tied to any particular officeholder. Broader legislative or administrative remedies will then be required.

1.1.2 Issue-Specific Mechanisms of Control

Another way of grappling with institutional failure is not by exchanging personnel but by intervening directly in the issue in question. There are a range of constitutional and administrative mechanisms providing for such interventions. Sometimes, they merely involve the rights to publicly question officials about their and their departments’ actions, as parliaments usually enjoy. Sometimes they provide the means to actually change existing policies, directly or indirectly.

The standard mechanisms for intervening directly on specific issues are judicial review and administrative oversight. Administrative oversight comes in different forms and may empower both politicians and administrative superiors to intervene at lower levels. Yet, there are many reasons why such mechanisms might not work: Bureaucrats may be overworked or may not care; bureaucracies may be too big to control effectively or be captured by certain lobby groups; close personal ties and personal loyalties may make intervention difficult; and tight budgets and coordination problems with other actors may get in the way of efficient oversight, even in reasonably well-functioning systems. And politicians, too, may not get involved for all these reasons and more. As Rosalind Dixon points out, even in well-functioning democratic systems there are routine malfunctions that may inhibit governmental responsiveness.Footnote ¹⁷ Current legislators have to consider a great deal of issues, often involving difficult and technical questions. Blind spots result from these pressures for time and attention within the ordinary legislative process, which may lead to some problems simply falling through the cracks. And previously declared priorities or the necessity for coalition building may prompt legislative inertia, with the result that some issues and rights violations may be ignored or sidelined, even in cases where they would arguably enjoy majority support.Footnote ¹⁸

Judicial review, too, grants courts powers to intervene, subject to certain conditions. However, courts need support from the other institutions routinely tasked to implement their decisions – and such support may or may not be forthcoming. (As we will see in the second part of this chapter, it is precisely when this assumption of support no longer holds that we see courts intervene more expansively, on the basis of governmental failure.) Judicial review is also limited by the fact that it typically happens ex post and in response to individual cases – and this in turn can make it a less than ideal tool to address broader structural and systemic problems, as we will see in Chapters 4 and 5.

Finally, there is a range of more indirect and subtle means of intervention, both within the legislative and administrative branches. Such mechanisms are often specific to individual legal systems, working to establish accountability within the bureaucracy and thus dealing with failure when it arises.Footnote ¹⁹ Within the legislative branch, too, there are indirect powers of accountability, for example through budgetary control over government spending, which parliaments may use to increase leverage over the executive branch. Here, too, though, the scope and impact of such powers will depend on the specific legal and constitutional context.Footnote ²⁰ And as we have seen, some constitutions explicitly provide federal governments with powers to intervene in individual states under certain circumstances or contain emergency clauses that shift powers to the executive branch in some situations, which I address briefly in Section 1.2.3.

1.1.3 Summary and Consequences

This brief overview shows that there is a range of formal and informal mechanisms to respond to failure by exchanging personnel or by intervening directly through mechanisms of oversight and control. Yet, it is also easy to see how some cases of failure or dysfunction might slip through the cracks. Not all available mechanisms will be adequate to all kinds of failure, and some may not themselves be working, for pragmatic or structural reasons. This is especially true for the kind of failures that are not likely to be remedied by exchanging officeholders but are structural or systemic in nature. Addressing these often requires significant efforts, including coordination between multiple branches. Not surprisingly therefore, administrators and politicians have sometimes sought to take over issues and deal with them outside of those legally prescribed channels. Doing so, of course, then implies a certain degree of legal flexibility.

1.2.1 Arguments from Failure

A prototypical example for a rare legalized argument from failure is rules on federal coercion, such as Art. 356 of the Indian Constitution. India is not an outlier in this regard. Rules for federal coercion exist in a host of states, sometimes framed in more explicit, sometimes in less explicit terms, allowing federal states to intervene in cases where sub-national units such as states fail to fulfil certain duties.Footnote ²¹ My own home country, Germany, provides for this possibility in Art. 37 of the German Basic Law:

This rule has a considerable tradition in German constitutional history, going back to the German Empire.Footnote ²² The takeover of the Prussian government by the federal German government under Chancellor Franz von Papen in 1932, the so-called coup or putsch in Prussia (Preussenschlag), represents an infamous example for the exercise of that power. It led to a prolonged legal battle in court (Staatsgerichtshof), which ultimately upheld the measure, albeit with some modifications.Footnote ²³ The measure was based on the second paragraph of Art. 48 of the Weimar Constitution, which concerned cases of the disturbance of public order and security, and thus somewhat differs from Art. 37 of the Basic Law. But Art. 48 allowed intervention also in cases where states did not fulfil their duties towards the Reich, and, perhaps surprisingly, this provision survived National Socialism. Both provisions essentially constitutionalized arguments from state failure to the benefit of the respective federal governments. However, unlike in Spain, where the federal government made use of the similarly phrased – and German-inspired – Art. 155 of the Spanish constitution to take over the Catalan government in the wake of its secessionist actions in 2017,Footnote ²⁴ in Germany, this provision has never been used in the post-war era, perhaps due to its abuse in the Weimar Republic.Footnote ²⁵ India’s Art. 356, by contrast, has been relied on in many controversial cases, with the Indian Supreme Court declaring some instances unconstitutional, in spite of claiming only (limited) powers of review.Footnote ²⁶

Beyond cases of federal coercion, where we typically have a legal basis for intervention, there are many further examples for arguments from failure in a number of different contexts. They range from arguments for why courts need to undertake structural reform and intervene in expansive ways in failing institutions to examples for triggering intervention by different international bodies and organizations, among many other contexts and areas.

Both the reasons for failure and its magnitude, however, may be very different. Sometimes, the original design may have been faulty from the start. Sometimes, institutional inertia blocks necessary adaptation, or institutions are captured by influential groups. Sometimes there may be coordination problems between different offices that need to work together. Failure can be limited to concrete instances of regulatory oversight, or it can be of longer duration. In the latter case, it often makes more sense to speak of dysfunction, though I use both terms interchangeably here. Sometimes, it arises because legislators or administrators actively want to harm certain groups, or because they do not care, since those groups lack political power; sometimes because they are unsure what to do to address a problem or feel they do not have the means to address it. Arguments from failure or dysfunction appear in all of these different circumstances, as a rhetorical device to justify expansions of authority – ‘power grabs’ – in certain exceptional situations. And while they are often blurred with arguments about implied powers or emergencies, they are not identical with either, nor should they be treated as if they were.

1.2.2 Teleological Interpretation and Implied Powers

Arguments about implied powers are, strictly speaking, one important example of a purposive or teleological approach in interpreting institutional competences. As such, they share the broader problems associated with purposive or teleological interpretation. Those problems are exacerbated when dealing with the interpretation, not of rights, but of competences. Teleological interpretation entails an interpretation of law in light of the key purpose or telos of the specific rule, which is typically deduced not from the historical drafter’s original intent, though such intent may play a role, but from a broader reading of the respective norm in its legal and political context. Thus, while teleological interpretation seeks to realize the primary aim or goal associated with a specific norm, there is always a risk of such an interpretation going considerably beyond the legal text. For example, the European Court of Justice has used the doctrine of the so-called effet utile to justify a ‘functionalist’ readingFootnote ²⁷ of many Treaty provisions in ways that have dramatically strengthened individuals’ economic rights under the Treaty, as well as the scope of Union law more broadly. This has sometimes occurred in a context of perceived political failure where the Court has felt it necessary to step in.

While such teleological arguments are primarily or ostensibly about the realization of certain rights (under a constitution or treaty), they often have significant and sometimes unintended consequences for our understanding of institutional roles and competences, which can be far-reaching. They may also blur into more specific arguments about the roles of particular institutions, for example in the context of structural reform litigation, where courts are called upon to ensure the effective protection of rights.

In other cases, the argument is more explicitly about the scope of the competences or jurisdiction of specific institutions. This complicates matters insofar as such competences are meant from the outset to be limited by the constitutional framework.Footnote ²⁸ Nevertheless, public lawyers have often adopted broad approaches to the interpretation of competences as well, where this has seemed necessary to exercise governmental functions, for example, by developing the doctrine of implied powers. Today perhaps most commonly used in the realm of international law, the doctrine was originally developed by the US Supreme Court in McCulloch v. Maryland, an early decision discussing whether the federal government had the power to establish a (second) federal bank for the United States, in the context of a state’s efforts to undermine this effort by taxing the bank.Footnote ²⁹ The US Supreme Court famously argued that while the power to establish a bank was not a power specifically enumerated in the US Constitution, it was implied by others that were. The general clause entitling the federal government to make ‘all laws which shall be necessary and proper for carrying into execution the foregoing powers’ therefore provided it with such power.Footnote ³⁰ Similarly, albeit in slightly different terms, the German Constitutional Court has developed the concept of federal competence in those cases where the issue to be regulated is per se a federal concern (Kompetenz qua Natur der Sache) – and many other courts have come up with similar ideas.Footnote ³¹ In international and European law, too, ideas of implied or inherent powers play an important role.Footnote ³²

Arguments about implied powers are sometimes blurred with emergency arguments (which I will discuss below). This is regrettable because the two are ultimately based on different rationales. Not infrequently, such a blurring of argumentative frames indicates that we are really dealing with a case where an argument from failure would be most appropriate. This becomes clear once we understand the purpose of implied powers arguments. That purpose is to read competences into an institution’s governing framework, ‘not in order to modify it or add to the members’ burdens, but in order to give effect to what they agreed by becoming parties to the constitutional treaty’.Footnote ³³ Giving effect to what the parties agreed on may imply filling in some gaps, but only where this is possible in light of the overarching framework in place rather than providing a new mechanism to deal with unforeseen events, such as emergencies. In other words, in cases of emergencies and indeed when we encounter arguments from failure, the special powers sought are typically not already provided for by the existing legal materials. This is different from arguments about implied powers, which are geared towards giving effect to the original purpose of the existing text, albeit by interpreting it in a broad manner.

In contrast, the less we can draw on text and standard interpretative methods, the more we move into the realm of failure and emergency arguments, where we will typically encounter efforts to distinguish the situation from other, more ‘normal’ ones and appeals to necessity. Moving in that direction means moving into a grey area of legality. Whereas implied powers and teleological interpretation are today widely relied upon and accepted as ‘legal’ methods/doctrines, the same is not true for arguments from failure and indeed emergency arguments, unless they are explicitly provided for in the legal text. As we have seen in the introduction, failure in particular seems too vague a notion, too politically and ideologically polarizing to count as a legal concept as opposed to a convenient rhetorical trope.

1.2.3 Emergency Arguments

Perhaps we should thus simply understand arguments from failure as a specific case of emergency arguments and defend them in those terms.

Contemporary constitutions today routinely include emergency clauses. These typically transfer powers to the executive and often allow for the abridgement of certain constitutional rights in cases of emergency. The database of the Constitute Project lists emergency provisions in 180 constitutions.Footnote ³⁴ Indeed, the US Constitution is today an outlier in not providing a specific regime of emergency powers, but here too a set of constitutional doctrines and statutory provisions lay down specific rules to apply in different kinds of emergencies. Many modern constitutions include emergency provisions in the text itself. Consider, for example, the Indian Constitution (Art. 352 ff.), the South African Constitution (Art. 37) and the Colombian Constitution (Art. 212 ff.), all of which transfer certain competences to institutions that do not normally hold them. The Colombian Constitution stands out insofar as it endows the president with emergency powers not only in the standard cases of war or internal violence but also in situations that ‘disrupt or threaten to disrupt in serious or imminent manner the economic, social, or ecological order of the country or which constitute a grave public calamity’ (Art. 215).

But even where such explicit rules are lacking or indeed unclear, we may encounter emergency arguments seeking to defend acts that would otherwise be illegal in terms of necessity.

Indeed, several authors defend a specific category of failure arguments, namely arguments for certain kinds of structural judicial intervention, in terms of a response to an emergency. Thus, in his work on the separation of powers, after criticizing courts that initiate judicial proceedings by themselves, as happens in some cases of public interest litigation, Christoph Möllers argues:

Möllers’ framing suggests that we may be moving outside of the realm of constitutional law in such cases and instead be operating within the realm of morality, albeit that it is unclear in what kind of cases this verdict might apply and whether it is confined to cases where judges initiate proceedings on their own.

Möllers is not alone in invoking an emergency argument in this context, though his qualification of the practice as extra-legal is not widely shared. Thus, in his work on the separation of powers and the role of courts, Dimitrios Kyritsis approaches failure as a case of emergency, but one that can be accommodated as legal under certain conditions:

These arguments are not surprising. Arguments from failure share an argumentative structure with emergencies and can accordingly be hard to separate from the latter. As Christian Kreuder-Sonnen argues in his book on international emergency powers, ‘the state of emergency can be understood as a legal institution regulating the suspension of (certain provisions of) the normally applying constitutional order – the state of normalcy’.Footnote ³⁷ Kreuder-Sonnen argues that exceptional cases in this sense presuppose two things: First, there needs to be a ‘constitutionally deviant expansion of political power for executive actors increasing the reach and/or the intrusiveness of authority (i.e. the adoption of emergency powers)’; and second, this expansion of powers will be justified in terms of ‘necessity’.Footnote ³⁸ Moreover, emergency clauses typically respond to anticipated failure or at least inefficiency of the institution ordinarily responsible by delegating power to another institution.

This closely parallels the kind of arguments we encounter in cases of institutional failure, and it is therefore no surprise that the two are often blurred and confused. However, there are several reasons why we should treat arguments from failure as distinct from emergency arguments.

First, unlike emergency arguments, arguments from failure are in practice not always or necessarily geared towards increasing executive power. Indeed, quite a few of the cases discussed in this book involve shifting powers away from the executive branch. To be fair, this difference would not seem decisive if we are trying to justify failure as an emergency insofar as shifts to other branches may be normatively less problematic. In other words, we might think that if we accept unwritten emergency powers for the executive branch, we should have even less of a problem with granting emergency powers to other, less dangerous branches and institutions.

Emergencies also often involve the suspension, or at least significant limitation, of individual rights, whereas arguments from failure are often directed towards better protecting individual rights, as we will see in later chapters. Once again, however, this would presumably be an argument for rather than against defending arguments from failure. More importantly though, while emergencies are notoriously difficult to define in advance, they are typically characterized by two qualities: scale and urgency.Footnote ³⁹ Urgency, however – that is, a rapidly escalating situation where things will get significantly worse unless speedy action is taken – will not always and not necessarily be present in cases where we encounter arguments from failure. As we will see in the course of this book, arguments from failure are often arguments about governmental omissions over a certain time period. Accordingly, the situation is often much less urgent, in the sense of not arising suddenly and requiring an immediate response to ward off further damage, leaving no time to reflect.

Moreover, the existence of unwritten emergency powers as such is already a highly problematic proposition and their legality is heavily contested in public law.Footnote ⁴⁰ While there are some who assert the need for such powers and defend them in terms of unwritten rules or indeed on the basis of a non-legal concept of emergency powers, following Carl Schmitt,Footnote ⁴¹ many scholars are rightly sceptical, given the high risks of abuse.Footnote ⁴² A range of theorists have emphasized how the toggle-switch paradigm of the emergency as an exception as opposed to a state of normalcy neither fits the very different realities of the cases that arise (ranging from natural disasters to terrorist attacks to cases of war and martial law) nor indeed provides a normatively convincing template to deal with them in terms of public law.Footnote ⁴³ What we need, therefore, are differentiated responses to different problems, and this entails recognizing arguments from failure as a category of their own, as opposed to treating them in terms of a dichotomous emergency framework.

Moreover, the general risks associated with emergency powers are heightened if we rely on them to address institutional failure. Crisis and failure are, as we know, hard to measure objectively. Blurring failure and emergency arguments plays into the hands of populists, emphasizing the ‘out of the ordinary “character”’ of events in order to justify radical breaks with existing rules. Famous examples include the fire of the German parliament, allowing a further clamp-down on political opponents by the German National Socialists in the 1930s, as well as the more recent military coup and its suppression by the Turkish government and its role in the installation of a new, more authoritarian constitution.Footnote ⁴⁴ Crises and their performances are important instruments in the emergence and strengthening of populist movements, as Moffitt shows drawing on Laclau’s work.Footnote ⁴⁵ Consider in this context also Bhuwania’s troubling account of Indian Public Interest Litigation, which he argues has enabled a form of judicial populism (see Chapter 4).Footnote ⁴⁶

Such risks of abuse loom especially large because the central justificatory framework when it comes to emergency arguments is necessity: Something truly dreadful will happen if the executive does not seize extraordinary powers to act. This also implies that the greater the danger, the more can be justified – opening up a dangerous slippery slope that may justify the adoption of authoritarian measures whenever a transgression of legal competences seems warranted as a matter of output. If we are seeking to ‘abuse-proof’ legal doctrines, adopting an emergency framework is therefore clearly the wrong way to go.

Finally, the emphasis placed on necessity in emergency arguments only partly captures what is going on in arguments from failure, which are, inter alia, concerned with inter-institutional relationships or at the very least should be understood in these terms (see more below). Thus, while considerations of necessity are hardly irrelevant to arguments from failure, they should not be understood as their animating spirit. The term failure – which is irrelevant to a straightforward emergency argument – emphasizes that.

2.2.1 Rethinking the Separation of Powers

The more recent literature on the separation of powers takes an approach that is more flexible than both the traditional model – which emphasizes that each branch of the government must exercise a distinct task – and Christoph Möller’s autonomy-based conception. More recent writing often starts from a critique of the traditional model, which has indeed never prevailed in theory or practice in a pure form.Footnote ¹⁶ It places emphasis on the need for the separation of powers principle not just to constrain power and secure individual liberty by creating checks and balances, but to bring about a rational and efficient ordering of the state.Footnote ¹⁷ Indeed, this ‘effectiveness rationale’ itself is nothing new in many ways, but in fact goes back a long way, and is discussed in such canonical sources as the Federalist Papers.Footnote ¹⁸ Placing emphasis on considerations of effectiveness opens the door to theories of inter-institutional comparison, which ask which institutions are best able to fulfil certain functions as a basis for allocating legal authority. This in turn opens the door to arguments from failure – albeit without presupposing the need for them. It may be the case, after all, that more rigid conceptions will provide a better basis for an effective government. However, as soon as we accept effectiveness as a central consideration, the debate changes and starts being about degrees of flexibility and rigidity, rather than about clashes of principle.

2.2.2 Democratic Experimentalism

The democratic experimentalism literature culminated in a series of publications in the first decade of the 2000s. This work applied the concept of democratic experimentalism in a variety of different settings, but with a particular emphasis on structural reform litigation and administrative processes, as well as EU governance issues. It is related to the broader governance discourse,Footnote ⁴⁰ which set out to describe changes in the processes of how states governed, emphasizing public-private cooperation, decentralization and a shift away from a command-control mode, all against the background of the 1990s’ turn to neoliberalism.Footnote ⁴¹ With this literature, democratic experimentalists share the belief in the need for decentralization and collaboration among multiple – public and private – players, but take a more normative position as compared to most strands of the governance literature.

Democratic experimentalists often understand their approach to be in line with a pragmatist conception of democracy that emphasizes epistemic uncertainty. With that comes the need for constant learning and experimentation as a key democratic virtue. Simon Sabel and Joshua Cohen in particular have developed a concept of ‘directly-deliberative polyarchy’, which they describe as a ‘radical, participatory democracy with problem-solving capacities useful under current conditions and unavailable to representative systems’.Footnote ⁴² Apart from Dewey, intellectual support for such ideas is drawn from a wide variety of sources, including business modernization literature, for example, on the so-called Toyota model,Footnote ⁴³ as well as the work of Roberto Unger. In a famous paper titled ‘Destabilization Rights’, Simon and Sabel build on arguments developed by Roberto Unger in his 1987 book False NecessityFootnote ⁴⁴ to defend structural reform litigation.Footnote ⁴⁵ Unger put forward a broader argument for democratic experimentation in an attempt to find a path beyond Marxism with its rigid understanding of history and emphasis on structural conditions, on the one hand, and social democracy with its acceptance of capitalism and its resulting inequalities, on the other.

More recent authors writing along similar lines, such as Roberto Gargarella, have moreover drawn on work by Jürgen Habermas on deliberative democracy.Footnote ⁴⁶ Gargarella argues for the introduction of more participatory elements in certain political systems where elites block real democratic participation through representative processes and defends judicial intervention to that purpose, including in cases of structural reform litigation.Footnote ⁴⁷

In the context of structural reform litigation, the key point of this line of thought is that the legitimacy of structural intervention by judges depends less on the ‘who and why’-question than the ‘how’-question. In other words, what matters is not so much under what conditions judges intervene in administrative or legislative processes, but how they do it. Thus, as long as judges act in an inclusive and participatory manner that is open to later correction – as opposed to issuing commands from the bench – their justificatory burden for acting in the first place is significantly reduced. While failure therefore plays a role in some theories that adopt an experimentalist approach, such as Simon and Sabel’s concept of destabilization rights, this is often not where the emphasis lies.

But democratic experimentalism goes beyond defending a particular ‘experimentalist’ style of structural reform litigation. It also plays a role in conceptualizing (and ratifying) the interplay of different state and supranational actors as well as private actors in the context of policymaking in the European Union – in particular, the Commission’s White Paper on Governance of 2001Footnote ⁴⁸ and the so-called open Method of Coordination. These sources provided a framework for developing soft-law policies on particular issues and should thus be understood as part of a broader move towards more flexible understandings of competences in the European Union. Subsequently, the concept has been broadened to encompass the relationships between different courts, such as national constitutional courts and the CJEU. These are understood as a form of experimentalist governance, insofar as they involve a series of judgments where courts shape their jurisdiction in a mutually interdependent manner, leading to the creation of a polyarchic and pluralist structure where no one institution has the last word.Footnote ⁴⁹ More recently, the concept of experimentalism has also been employed by Gráinne de Búrca in a human rights context as a way of describing how different local and global bodies, both public and private, are involved in the process of creating new human rights through political struggles as well as implementing such rights on the ground.Footnote ⁵⁰

Such ideas can help accommodate and justify arguments from failure. They shift the key challenge not towards defending intervention in the first place, but rather towards evaluating and designing responses in a way that is sufficiently open and participatory. And they suggest that if the response to failure consists in opening up avenues for participation and experimentation, perhaps we should understand this as an improvement of democracy, rather than as a problem.

2.3.1 Separation of Powers and Transformative Constitutionalism

The literature on transformative constitutionalism and the new separation of powers, with its focus on effectiveness, comes with significant risks of enabling a slide towards authoritarian modes of governing. The more we focus on output rather than input as a standard of public legitimacy and legality, the more we enable governance unrestrained by legal and democratic safeguards.

This risk strikes me as particularly prevalent in some of the more radical South African critiques that treat constitutionalism only as a value insofar as it furthers certain substantive aims, that is the eradication of poverty and the construction of a more equal society. Somewhat ironically, this kind of outcome-focused style of argument strongly recalls similar purposive arguments in an EU context, though deployed there to bring about the common market, rather than equality. This speaks not only to the ideological malleability of such arguments but their authoritarian underpinnings. We can see glimpses of that already in an EU context,Footnote ⁵¹ but for a different picture of where such an approach can lead, we might also consider India’s experience with Indira Gandhi’s emergency regime from 1975 to 1977, in which arguments from the failure to bring about a more equal society played a central role.Footnote ⁵² Indira Gandhi’s political campaign was based precisely on the eradication of poverty in India. She used this as an argument to resist the threat of the Indian judiciary and in particular the Supreme Court in blocking her political agenda.Footnote ⁵³ In particular, she blamed India’s Anglo-Saxon legal system for being insufficiently attuned to the needs of the ‘poor and weak’, which in turn fed into her broader attack on the judiciary as a guardian of this outdated concept of law – foreshadowing contemporary critiques of constitutionalism as insufficiently decolonialized.Footnote ⁵⁴ Framing this as a ‘conspiracy’ due to her introduction of ‘certain progressive measures of benefit to the common man and woman in India’, as a result of which ‘[d]uly elected governments have not been allowed to function’, she invoked the constitutional emergency clause to suspend a range of rights as well as democracy. Before it ended, 11 million Indians were sterilised (as part of the plan to eradicate poverty) and 110 000 people jailed.Footnote ⁵⁵ India’s story is not, of course, an outlier; authoritarian regimes have served different political guises, some furthering not only left-wing but also right political agendas.

That said, transformative constitutionalism as a concept is committed to legal and constitutional means of change. (And this is why those who adopt a more radical approach, are increasingly critical of transformative constitutionalism, which they view as an insufficiently decolonialized project.Footnote ⁵⁶) African comparative scholarship on the separation of powers, in contrast, often still takes a more traditional liberal approach.Footnote ⁵⁷ And even where constitutional realization approaches are popular, as in South Africa, they will not necessarily be dominant in judicial practice. The South African Constitutional Court has accepted the need for collaboration among the different branches, in particular with an eye to ‘fulfilling practical constitutional promises to the country’s most vulnerable’.Footnote ⁵⁸ At the same time, it has also often been more careful and cautious than many of its academic commentators would like it to be when it comes to involving the court in the details of policy-making.Footnote ⁵⁹

In Europe too, the call for flexible responses and collaboration comes with limitations. Thus, Nick Barber emphasizes the need to reform institutions rather than take over their powers in the long run.Footnote ⁶⁰ Aileen Kavanagh puts forward a more explicit argument about how courts should approach situations when other branches are unwilling to do their job. She argues that political deadlock or legislative inertia does not allow for judicial intervention in all cases, both because judges cannot actually fix all problems and because their doing so would not be legitimate.Footnote ⁶¹ Ultimately, for Kavanagh, whether judicial intervention in cases of legislative inaction is legitimate depends on a range of factors, including the

This is fairly close to the model I propose below, but some further qualifications are necessary. Courts cannot always wait for legislatures to act, and there are other factors we need to take into account. It also matters whether courts expand their powers to protect and fulfil rights and alleviate poverty or whether administrators do so in the interest of, for example, realizing the common market in the European Union, even though both discourses open up risks of a slide towards authoritarianism. Most importantly, however, this literature suggests the need for clearer legal categories that cash out limits to the role of effectiveness arguments in public law – and thus the need for the project taken up in this book.

2.3.2 Democratic Experimentalism

The democratic experimentalism literature suggests a way of mediating some of these issues. One way to respond to the inherent authoritarian drift of arguments from failure is by insisting on inclusiveness and participation. It is therefore important to think about how institutions should intervene, in particular in the context of structural reform litigation, which I discuss in Chapter 5. Yet, while the focus on inclusive procedures and experimentation is important, they are not a solution to all problems. Indeed, they also create some problems of their own.

Start with experimentation. Roberto Unger himself argued that in order to enable experimentation, citizens must be able to rely on the existence of a certain safety net, protected in the form of ‘immunity rights’.Footnote ⁶³ Immunity rights ‘protect the individual against oppression by concentrations of public or private power, against exclusion from the important collective decisions that influence his life, and against the extremes of economic and cultural deprivation’, giving him ‘the justified confidence of not being fundamentally endangered by the expanded conflicts of an empowered democracy’. Healthcare, housing and educationFootnote ⁶⁴ – all major fields of public law litigation in the United States as well as in the Global South – therefore all fall within the scope of immunity rights, whose realization Unger thinks must be guaranteed as a matter of course. Immunity rights are not predicated upon their effect on others and thus must not be balanced against other interests or rights. They are absolute, circumscribed in bright and clear lines. In all of these fields, including freedom from governmental and private(!) oppressionFootnote ⁶⁵, Unger therefore discouraged experimentation, because he thought that even within the empowered democracy citizens must be safe in the sphere circumscribed by immunity rights. In fairness to Simon and Sabel, Unger himself is not clear what precisely would be covered by immunity. He refers to judicial intervention in schools or electoral organization as an example of institutional destabilization, as Simon and Sabel rightly point out.Footnote ⁶⁶ Nevertheless, experimentalism should clearly not be without limits and these limits thus need to be spelt out.

Second, developing participatory practices and encouraging negotiation and collaboration may be misplaced in some cases. For example, as Katie Young and Sandy Liebenberg have emphasized in the context of socio-economic rights litigation, collaboration and engagement are all very well if all parties of the conversation hold roughly equal power. But they are likely to generate unfair outcomes where significant imbalances of power, often affected by wealth, exist.Footnote ⁶⁷ This argument would seem to apply beyond the realm of structural reform litigation in other contexts involving collaborative structures of governing.

Most importantly though, we should not mistake ‘stakeholder democracy’ for democracy. This applies especially where final decision-making powers do not lie with state institutions, such as courts. Nor do public hearings provide a substitute for parliamentary debate and decision-making. Even where courts or the EU Commission bring in new voices, forcing governmental officials and/or legislators to sit down with civil society representatives to set out a path for structural reform together, the group selected will necessarily be just that: selected and selective. It will not represent a majority of citizens, not in either a formal or an informal sense. In other words, participation does not mean democratic accountability, whatever other values, such as transparency, reason-giving and judicial scrutiny, it may serve. Democracy also includes the voters’ ability to throw the rascals out. Where policymaking is conducted in complex multilevel networks, including private and public actors, accountability can be difficult to achieve. No one can be voted out of office in the next elections. And whatever else democracy entails, this idea is at the heart of the prevailing conception of democracy in nearly all democratic states as well as part of Dixon and Landau’s concept of a democratic minimum core, building on a wide range of democratic scholarship and practice.Footnote ⁶⁸

Of course, representative democracy may itself not be all that we make it out to be. With its emphasis on the value of participation in local units and experimentation, democratic experimentalism connects to moves in the more recent radical democracy literature, seeking to go beyond representative democracy.Footnote ⁶⁹ For example, Helen Landemore has recently put forward proposals for greater openness in democracy, suggesting the creation of mini-publics consisting of individual citizens randomly selected by lottery to be involved in deliberating policy and in law-making over a limited time period.Footnote ⁷⁰ How much room there is for such arguments will depend on the broader constitutional framework in place; some constitutional systems like Germany are based on a firmly representative concept, while others are more open in this respect. Yet, regardless of these details, democratic accountability will remain a key concern as elections remain the central mechanism for securing it.Footnote ⁷¹ If democratic experimentalists shift power to fora and procedures that are lacking in this regard, this requires justification, even if those fora may be inclusive or participatory. As it stands, such justifications often involve claims of failure, as we have seen, whether this is a specific institutional failure or a wholesale democratic failure.

This is not to say that democratic experimentalism is not a useful descriptive and normative frame. It is, and indeed there is much to like about it. By focusing on enabling political participation of citizens in the policy-making that affects them, and drawing attention to institutional dysfunctionalities that target minorities, democratic experimentalism is not just another neoliberal governance theory. However, unless its proponents spell out more clearly its relationship with the rule of law and democracy, it risks becoming just that. Experimentation must take place within certain legal parameters if it is to respect the equal dignity and voice of participants. It must ensure the accountability of the central actors involved. Democratic experimentalism is also not the only problematic concept in this regard, of course. This applies more broadly to other theories that sideline central constitutional principles in favour of an emphasis on networksFootnote ⁷² and collaboration.

2.4.1 Constitutional Authority in Context

The first requirement for arguments from failure is that we must be operating within a political and legal context where expectations of mutual collaboration and responsiveness prevail. Arguments from failure are not just devices for self-help,Footnote ⁷⁴ but are rooted within a broader legal framework that envisages different institutions to work together.

Within domestic constitutional systems, this is usually the case. Most constitutional democracies typically subscribe to some version of the separation of powers. It is also clearly the context in which Kavanagh’s theory is meant to apply.Footnote ⁷⁵ That said, some legal systems are more open to this kind of flexible approach than others. It is not accidental that Kavanagh herself writes from a background in the United Kingdom, a stable democracy based on the principle of parliamentary supremacy, without a strict understanding of the separation of powers. Some other contemporary constitutions already come with a thick normative agenda obliging the state to pursue certain goals, such as a transformative constitution. That, too, can readily ground a constitutional imperative to collaborate to achieve the shared goal. Yet, even absent such more obviously legally friendly environments, most constitutional democracies should be open to arguments from failure in principle, albeit some in much more limited ways than others. This is because even in systems where more emphasis is placed on constraining state power and preserving individual liberty understood in a traditional way, we will find expectations of cooperation and of effective government. This is not to say, however, that arguments from failure can be defended in all circumstances.

Things become more difficult in the realm of international law, where such expectations will not always hold. There is typically no traditional separation of powers in this context, even within specific treaty regimes, and this begs the question how to assess arguments from failure in that context – a question I address in Chapter 7.

2.4.2 Legalizing Failure

To evaluate arguments from failure, it also matters how existing competences are formulated in the legal text – and in particular whether they are clearly framed in a broad and flexible manner or in a narrower manner and whether they set out a comprehensive framework or not. It also matters what other legal mechanisms exist to address failure. Finally, we need a clearer sense as to what constitutes failure – a question which I have bracketed so far. While it would be impossible to put forward a definition of failure to fit all circumstances where arguments from failure come into play, as I argued in the introduction, we can still give some more content to the idea to get a better grip on it.

2.4.3 Managing Risks

However, in light of our discussion above and the literature survey, there will be some cases where arguments from failure should not be considered permissible, in light of the grave risks to democracy they would entail.

3.2.1 Functionality

Functionality is a standard part of proportionality analysis nearly everywhere. It requires that the institution seeking to respond will be likely to resolve the problem at hand or at least contribute significantly to resolving it. Two related considerations are important in this regard. First, this involves thinking about what needs to be done to resolve the issue. Second, we need to assess the capacities of the institution seeking to respond.

What needs to be done is often not an easy question to answer. It depends on why state institutions are not acting to address the issue or indeed why they are causing it. Resistance in particular can be hard to overcome, particularly if important institutional actors are unwilling to do something. The same can be true for problems stemming from a mix of coordination problems, capacity issues and other structural factors. Overcoming what will often be entrenched dysfunction, due, for example, to corruption and capture of agencies, can be exceedingly difficult and sometimes impossible. In at least some cases, therefore, functionality will mean that arguments from failure cannot justify intervention by other institutions simply because there is no realistic chance of change. At the same time, the fact that it is not immediately clear what needs to be done is not a conclusive argument for doing nothing. But the remedy may not always lie in other institutions expanding their powers to take over. In some cases, the best strategy may lie in constructing so-called institutional bypasses that allow citizens to sidestep failing institutions by creating new channels and institutions rather than fixing the old ones. The idea – explored in more detail by Mariana Mota Prado and Michael Trebilcock in their pioneering work in this area – is that this can either exercise pressure on the old institution to reform or make the old ones superfluous in the long run.Footnote ²⁴ And where new institutions are created, this is of course not necessarily an exercise of power that requires special justification.

In our context, we need to consider if the institution seeking to take over is itself institutionally suited to doing so. In many cases, this will be doubtful as institutions created for certain purposes will not have the necessary expertise or capacities to act on other issues. Some challenges in this regard will be more easily remedied than others. Thus, if expertise is the problem, the institution in question may bring in outside experts on specific questions. However, if the problem is of a strongly polycentric nature and we are dealing with a court considering to step in, it may not be amenable to resolution within the traditional bipolar structure of judicial trials, and courts then will have to decide whether to develop new procedures – as they often do in practice – or step back from the issue entirely. This is also where the literature on democratic experimentalism may show us the way to address structural problems in a creative and experimental manner.

3.2.2 Necessity and Costs

Necessity in our context must be understood both in terms of the legitimacy of the institution acting and the one seen to be failing – requiring us to conduct a no-more democratic means test – and as a qualifier when it comes to the problem that purportedly needs fixing.

On the first point, it bears emphasizing that it is not just the institutions that matter but the process too. Here, too, the democratic experimentalism literature offers important insights when it comes to favouring more flexible as well as participatory approaches over final dictates by individual judges. We must also address the question if the problem that has been left unaddressed due to the institutional failure is significant enough to warrant the intervention of an institution that ordinarily has no powers to address the issue in the way needed. What factors should matter in this context?

Recall the introduction to this chapter and the attempt by the UN Commission on Human Rights to carve out an exception to the standard practice of non-intervention for situations in which they encountered a consistent pattern of gross human rights violations. Spelling out what this meant was not an easy task. The Sub-Commission later asked a Russian expert, Dr. Stanislav Chernichenko, to prepare a paper inter alia to define what cases qualified.Footnote ²⁵ This study was subsequently submitted to the Human Rights Committee. Chernichenko emphasized the difficulty of establishing clear criteria when it came to distinguishing ‘gross’, ‘massive’ or ‘systematic’ human rights violations both from each other and from other cases:

Chernichenko also offered examples. According to his account, gross violations referred to a specific kind of human rights violation, for example cases of torture or arbitrary and protracted detention. They often led to ‘massive’ violations (understood in terms of range) such as genocide, disappearances, slavery-like practices, and mass and arbitrary executions, which also counted as gross violations at the same time. And though in principle open to measuring rights violations in some way, Chernichenko argued that the three factors to consider offered in the literature, namely ‘the scope of violations, i.e. the degree of seriousness; their intensity, i.e. the frequency of their occurrence over a given period of time; and their range, i.e. the size of the population affected’, were ultimately insufficient. In particular, when it came to economic, social and cultural rights, the problem was often going to be that governments simply did not have the means of improving things, and such cases thus typically should not be considered gross violations under the definition.

Though the efforts to define a consistent pattern were ultimately abandoned at the UN, Chernichenko was not the only one to attempt to define the term.Footnote ²⁷ Similar to the three-factor test Chernichenko had discussed and dismissed, the Austrian scholar Ermacora suggested in 1974 that three elements had to be present: first, a time element in the sense of some continuity of the violations; second, a quality element that required violations to have a similar scope as apartheid and racial discrimination; finally, there should be a quantity element comprising a substantial number of cases and an expression of a broader policy or ongoing practice where no effective remedies existed within the state concerned.Footnote ²⁸ These criteria lingered on in a later attempt at definition by US commentator Maxime Tardu. He added that there needed to be an additional element of planning or of sustained will on the part of the perpetrators, and that in considering the scope of violations, many thought that a degree of inhuman or degrading treatment should be required.Footnote ²⁹

The most comprehensive and convincing attempt at defining the term, however, was presented by Medina in her doctoral dissertation.Footnote ³⁰ Based on her analysis of the UN developments and her reading of Ermacora and Tardu, Medina proposed a four-factor test for rights violations to count as gross, requiring (1) sufficient quantity, (2) a temporal element, (3) quality and (4) an element of governmental planning. While the first elements are fairly self-explanatory, it is less clear how to assess the quality of rights violations. Medina argued for an integrated approach which would consider the quality of the right or prohibition in question, the kind of violation, and the status of the victim. Regarding the quality of the right in question, she referred to the category of ius cogens offences, the categorization as international crimes or the fact that only some rights tend to be non-derogable in human rights treaties. The fourth and final element of planning is unfortunately not fleshed out much by Medina. She argued that violations cannot be committed ‘at random’ but must typically serve a specific goal and thus be part of a broader plan or policy. (This understanding, as we will see later, would make it very hard to capture socio-economic rights violations, today a major area of expansive action by domestic courts.) Summed up in one sentence by Medina:

This analysis focusing on the quantity, duration and quality of rights violations has strong parallels to the doctrines we see in some courts today to deal with structural reform cases. This includes the Colombian Constitutional Court with its ‘unconstitutional state of affairs doctrine,’Footnote ³² which I discuss in Chapter 4, as well as the European Court of Human Rights’ pilot judgment procedure under Art. 61 of the Court’s Rules of Procedure. In both courts, considerations of quantity and time are central to analyzing whether the cases they confronted warrant the application of the unconstitutional state of affairs doctrine or the pilot procedure – and thus an exceptional structural remedy.

Intent, on the other hand, usually plays a lesser role. In my view, it should not be taken as a requirement for intervention, which may appear necessary also in cases of incapacity or incompetence that do not include intent to violate, as the literature on structural reform litigation demonstrates. That said, intent can play some role as an additional argument for certain kinds of interventions.

Finally – costs. The costs of the respective intervention should ideally be understood to include both immediate costs as well as potential long-term consequences, which will often be harder to assess. They can accrue in relation to the institution failing to act as well as the one seeking to intervene or the broader public. Thus, it may be that an intervention will both remedy a pressing problem and help restore trust in the state and its institutions. However, the intervention may also contribute to fostering an abdication of responsibility by the institution in question and draw attention to dysfunctionalities in a way that damages public trust in institutions more permanently. Finally, as we know from the emergency literature, what starts as an intervention in an exceptional case may become normalized and entrenched in the longer run (the so-called ratchet effect),Footnote ³³ extending the power of the institution to act in circumstances beyond the original case. In some situations, we may be able to take mitigating measures to lower the risks of certain negative consequences, but in others this may not be possible. Unfortunately, assessing which of these consequences is most likely will often be difficult and involve prognoses with significant uncertainty.

3.4.1 Background

The events leading up to the passage of this legislation constitute one of Britain’s major political scandals in the last ten years. They have been the subject of much public discussion as well as the release of a mini-series, titled Mr. Bates vs the Post Office.Footnote ³⁴ In brief: From 1996 onwards, the British Post Office had piloted a new software called Horizon, designed by the Japanese company Fujitsu, to streamline accounting in its post offices. However, that software had serious flaws that were not public knowledge at the time. These led to the appearance of shortfalls in a range of post offices in Britain and subsequently to post office employees being investigated for a series of offences. As a result, many of those employees lost their jobs and their savings on account of having to make up for alleged shortfalls out of their own pockets. Four people killed themselves and hundreds ended up criminally convicted, with some serving time in prison. As the scandal slowly unravelled, the Post Office was shown to have ignored early warning signs about the software. More than that, the Post Office, which enjoys the legal power to conduct its own investigations instead of involving the British police, had not always conducted these investigations with the necessary care. It had sometimes only passed on parts of the evidence to the state prosecutors, who subsequently brought the criminal charges against the employees. Efforts by the postmasters to uncover the truth were hampered for a long time by the Post Office’s lack of cooperation, until they ultimately succeeded in the courts. In the landmark Bates v. The Post Office judgment, computer flaws were finally determined to be the cause for the alleged shortfalls.Footnote ³⁵ Subsequently, a number of postmasters involved with the litigation successfully challenged their criminal convictions in the appeals courts, leading to a finding that not only had the defendants previously not received a fair trial, but also that the Post Office had acted to subvert the integrity of the justice system and undermine public confidence.Footnote ³⁶

However, it also became clear that there were hundreds of wrongful convictions still remaining on the books, leading to the question of how to deal with this situation. The normal route to challenging such convictions depends on where the original cases were tried and whether the individuals in question had pleaded guilty – which had often been the case, as many postmasters took deals to avoid prison time. Nevertheless, ultimately overturning convictions could only be done by the appeals courts, in some cases (especially after guilty pleas), only after a referral by the criminal cases review commission. However, given that some of the criminal convictions went back 20 years and evidence was not always still available, the process of overturning the convictions in the courts was neither straightforward nor fast, with each case requiring separate consideration. In other words, this route was quite burdensome and problematic. In response, the British government chose to draft legislation instead, quashing all convictions of post-office employees for certain offences falling within the prescribed time frame.

However, this legislation faced controversy among British public lawyers and judges. Some considered it an inappropriate intervention by the government into the judicial domain that interfered with the rule of law, the independence of the judiciary and, ultimately, the separation of powers. There have been more specific concerns about how to identify the people whose convictions would be quashed. Concerns about the legislation’s legality have also been raised by the Chair of the Justice Select Committee in the House of Commons, Robert Neill, and by the United Kingdom’s Chief Justice, who has challenged the claims that the justice system had not been able to process the appeals sufficiently, efficiently and swiftly.Footnote ³⁷ That said, others defended the legislation as an appropriate use of parliamentary powers to address legitimate grievances within the British system of government, with its particular emphasis on parliamentary sovereignty. In this view, the legislation contributed to upholding the rule of law. It ensured that justice was done in a situation where many criminal convictions had been based on incomplete evidence.Footnote ³⁸ In addition, the fact that the legislation targeted specific convictions meant, some argued, that it did not impinge on the rule of law or the separation of powers. Rather, it constituted an instance of parliament addressing a structural problem with corporate culture in the Post Office at the time and served to remind institutions with their own powers of investigation, such as the Post Office, of their duty to exercise these powers in a careful and impartial manner.Footnote ³⁹

While it, therefore, remains contested whether the separation of powers indeed allows for the passage of a statute quashing convictions, it seems fair to state that its passage involved an unconventional use of parliamentary powers. Thus, the official explanatory notes to the statute itself state under ‘legal background’ that ‘[t]his is an unprecedented and wholly exceptional legal solution to a miscarriage of justice of unparalleled scale and impact (…) The approach to quashing convictions in the Act does not set any constitutional precedent’.Footnote ⁴⁰

It seems to me that the postmasters statute is best understood in terms of an argument from failure, with the failure in question being the previous miscarriages of justice leading to hundreds of wrongful convictions and thus rights violations. Understanding the legislation in those terms makes clear that it does not involve an either-or-case, in the sense of parliament either having powers to quash convictions by legislative fiat or not. Rather, it is a case where such a power might be defended only in terms of the need to react to an exceptional situation by unconventional means and was not intended to set a precedent for the future.

As an argument from failure, we can then ask our set of questions. The first of these concerns the existence of a framework where there is a legal expectation of mutual collaboration and support. This is a hurdle easily cleared if we recall the broad basis for a collaborative understanding of constitutionalism in the United Kingdom, as argued convincingly by Aileen Kavanagh.Footnote ⁴¹

The same applies to the requirement of a breach of law, which in this case was the miscarriage of justice involving criminal convictions on the basis of flawed evidence, as set out in the judgment of Bates v. The Post Office.Footnote ⁴²

The third concerns the question if we can indeed identify something like a gap in the law, in the sense of there being no comprehensive framework to deal with issues such as the one that has arisen. Here, things start getting trickier, given that there obviously is a process for overturning criminal convictions in place. The question then is to what degree we should understand this framework to be comprehensive. More particularly, we might ask if it is a framework that is geared to operate or can operate not just to deal with isolated wrongful convictions but rather with structural problems leading to hundreds of wrongful convictions, as in this case. This is where different understandings of the rule of law and the role of the British parliament start to matter. Given the principle of parliamentary supremacy and the traditional role of parliament in addressing grievances, it seems to me that there must be at least some room in principle for a parliamentary role in dealing with structural problems in the justice system as opposed to an isolated individual wrongful conviction – and this suggests that we should not view the existing judicial mechanisms as fully comprehensive. However, we should be mindful here that such an argument might be more difficult to make in other constitutional democracies where a stricter understanding of the separation of powers prevails.

Fourth, we will have to ask about the functionality, necessity and costs involved in quashing the convictions by legislative action. We, thus, need to answer the question if the legislation will actually remedy the problem. The main challenge in this regard will be whether it is possible to identify all the relevant post office employees. Some have argued that the framing of the statute is vague in places, thus making it difficult to identify with certainty all cases where criminal convictions should be overturned.Footnote ⁴³ Nor does the statute address the question of financial compensation, which is being addressed through additional legislation.

More difficult still is the question of necessity in the case at hand. For while it is clear that the Post Office scandal indeed involved a structural problem leading to miscarriages of justice in hundreds of cases with severe personal consequences, it is less clear that legislating to overturn previous convictions was necessary to restore justice. This question is perhaps the one that has loomed largest in the British debate, albeit in a rather freewheeling way that speaks to how we understand necessity to be a core requirement whenever we confront arguments from failure. The key question has been in particular whether the existing judicial process was indeed as slow as it was suggested, and if so, whether the best way to increase speed might not have been to increase funding for the Criminal Cases Review Commission instead. Like many British institutions, the Commission had suffered from severe cutbacks in the last fifteen years.Footnote ⁴⁴ Others have, however, questioned if funding is the real or only problem. They have traced the very low success rates of appeals to the Criminal Cases Review Commission to the high statutory threshold for referring cases back to the appeals court, which requires a ‘real possibility’ of having the case overturned.Footnote ⁴⁵ But whether funding or the current statutory framework is slowing things down, both could be addressed either by increasing funding or by changing the legislative framework.Footnote ⁴⁶ It has also been pointed out that, currently, none of the people convicted are still serving prison sentences, and accordingly, the need for urgent action is less dramatic than it would have been some years ago. That said, previous criminal convictions of course continue to impact the lives of those concerned negatively in other ways. Overall, however, it seems that there is considerable scepticism within the British public law community as to whether legislation was necessary to deal with the situation. If the sceptics are right, the argument from failure would have to be rejected at this stage.

Lastly, when it comes to costs, additional problems surface. The legislation will likely exonerate some people whose criminal convictions were not based on flawed evidence. It will also lead to awkward results insofar as it excepts people who previously tried to challenge their convictions in the appeals court but were unsuccessful on the basis that their convictions were likely not based on the flawed Horizon evidence. Yet, this leads to a situation where people who would now have had their convictions purged under the more lenient statutory approach are excluded merely because they previously sought to use the ordinary appeals process to challenge their convictions. Most importantly, however, there is the risk of the legislation creating a precedent for other cases in which parliament might feel the need to overturn sets of judgments it considers problematic, albeit that the statute itself disavows that idea. On the other hand, there is a pressing need to bring about justice for hundreds of people whose lives had previously been turned upside down by wrongful convictions. All things considered, it would seem that the costs in this case are perhaps not the main challenge, but that it hinges on the question of necessity.

British public lawyers are far better qualified than I am to resolve that question, and I, accordingly, do not seek to provide a full answer here. My interest in this case is as an illustration of how treating the legislation quashing the convictions as an argument from failure provides a way to deal with the question of parliamentary overreach in a principled manner. As we have seen, evaluating the legislation in those terms means addressing many of the considerations already present in current British debates, but it does so within a broader theoretical and legal framework. That framework offers no straightforward answers. It also leaves plenty of room for discussion, as we have just seen. Context and the specifics of the local legal systems will have to be taken into account to come up with an answer. But the same is true for other approaches used in the existing debates, framed in terms of the scope of parliamentary sovereignty, or the somewhat free-floating references to necessity. Treating the legislation as an argument from failure at least means asking the right questions and thus provides an important step for ultimately developing a convincing answer.