In August 2023, just an hour after the election results were announced in Gabon confirming previous president Ali Bongo Ondimba as the winner, a group of military leaders assembled to announce their take-over of power. An officer reading out a statement justified the coup as a response to the ‘serious institutional, political, economic and social crisis’ Gabon had been experiencing:
We are therefore forced to admit that the organisation of the general elections of 26 August 2023 did not meet the conditions for a transparent, credible and inclusive ballot so much hoped for by the people of Gabon (…) Added to this is irresponsible and unpredictable governance, resulting in a continuing deterioration in social cohesion, with the risk of leading the country into chaos (…) People of Gabon, we are finally on the road to happiness.Footnote 1
The Gabonese coup is only one among several recent coups in Africa. Here and elsewhere, military leaders have invoked the failure of elected governments to deal with ongoing crises, giving rise to a discussion in the literature if there might be such a thing as a ‘good coup’.Footnote 2 None of this is surprising. States and governments have long depended on their ability to achieve certain things – security first and foremost, but also, more broadly, the welfare of their populations. In his recent monograph All Is Well: Catastrophe and the Making of the Normal State, Saptarishi Bandopadhyay argues that disaster management became a core function of eighteenth-century states, whose governments sought to reframe disasters and emergencies as something to be managed, guarded against and administered rather than viewing them as natural events or indeed divine punishments.Footnote 3 By demonstrating their ability to grapple with disasters and restore normalcy, states and rulers asserted their legitimacy and to that purpose began building modern (‘rational’) bureaucracies, staffed with experts.
But it is one thing to seek to design the government to be effective and another to then argue for departures from the existing legal framework in cases of supposed or real failure. As we saw in Chapter 1, such arguments come in different forms and contexts. Sometimes they have a basis in law, but more often this is not the case. They often challenge key constitutional principles, such as the separation of powers, the rule of law, constitutionalism, federalism and, not least, democracy.
I start this chapter by spelling out these normative stakes more clearly, albeit briefly. Next, I turn to theoretical resources in the literature on the separation of powers, including in systems of ‘transformative constitutionalism’, and on democratic experimentalism, that offer a defence of arguments from failure. Some of these are more convincing than others, as we will see. These resources nevertheless suggest that a principled defence of arguments from failure as legal is possible in some cases. In other words, reiterating the introduction – the challenge is to make sure that arguments from failure are safe, legal and rare.
2.1 Normative Stakes
Not all arguments from failure come with the same normative stakes.
Whenever arguments from failure are not explicitly legalized – which means in most cases – they will raise questions about the rule of law and constitutionalism. State institutions exercising power over citizens must be established by law and only exercise their powers as circumscribed by the law. This core principle of constitutionalism, however, is put at risk if existing rules and competences can be simply expanded in the interest of better results. If institutions are free to fix problems arising in response to the failure of others, this may lead to what Alexander Somek labels in the context of the European Union ‘governance by situations’.Footnote 4 To Somek, the ‘normativity of distress demanding that something be done in the face of security hazards, pandemics, collapsing banks, volatile financial markets, high sovereign debt, or terrorist activity’ constitutes a threat to the traditional understanding of constitutionalism and legality, which presuppose that any exercise of public authority is based on law.Footnote 5 Arguments from failure may hence feed into an ultimately authoritarian paradigm. Within this paradigm, democracy, too, takes a backseat as legality is no longer understood in terms of acting in accordance with a power authorized by the people, but rather in the interest of bringing about certain results. This begs the question if arguments from failure ultimately pave the way for a triumph of utilitarianism, for what is ‘good’ over what is ‘right’, opening the door to any number of rights abuses in the name of collective welfare as associated with many totalitarian regimes.Footnote 6
Second: democracy. Whatever our understanding of democracy entails, democratic principles are at stake first and foremost when arguments from failure sanction the expansion vis-à-vis parliaments of the powers of actors such as courts or other institutions that do not have the same kind of democratic legitimacy. Structural reform litigation is the most straightforward example for this practice when it sees judges intervening not just in administrative processes, but in the kind of lawmaking that remains the domain of legislatures. But as we have just seen, democracy is also indirectly at stake in other cases, simply because the relevance of democratic authorship recedes in the interest of effective problem-solving.
Third, arguments from failure may run counter to the principle of the separation of powers if institutions exercise powers normally reserved for another branch. This is true whether we are dealing with a system where the separation of powers is applied relatively strictly or more loosely. Arguments from failure challenge the principle of functional integrity, traditionally considered an important aspect of the separation of powers, which in turn constitutes a key principle of constitutionalism and is of relevance at the international level as well.Footnote 7 It entails the idea that each of the three branches of the government exercises a distinct task, sometimes put forward in the strict form of a ‘one branch-one function’ view.Footnote 8 Functional integrity requires mutual respect: As different branches fulfil their tasks, they must not encroach on each other’s spheres and competences.
But why is functional integrity important and what does it mean? Eoin Carolan argues that the distinctness of the three branches reflects the autonomy of the individual citizen as the core value.Footnote 9 Individual autonomy, according to Carolan, is not only protected by individual rights or in individualized procedures. Individuals also have an interest in the collective good. The key yardstick for Carolan is that decision-making procedures are non-arbitrary, which implies that they must be designed to fit their (different) purposes. The different branches realize these purposes (setting out collective policies, local application, ensuring fairness in individual cases) in different ways, which reflect their institutional capacities.Footnote 10 Similar to Carolan, Christoph Möllers seeks to root the separation of powers in autonomy, more precisely in individual autonomy to make our own life choices and collective autonomy to decide together how we want to live in a society and state. Möllers accordingly doesn’t distinguish the different branches in terms of their reasoning, but instead as a matter of institutional set-up. This set-up has to match their function. Hence, legislatures decide collective issues prospectively and in collective procedures. In contrast, ‘[c]ourts adjudicate individualized cases on external initiative, retrospectively’.Footnote 11 Different institutions and different institutional processes thus ensure that both individual and collective autonomy can be realized. This also means, however, that even standard practices such as constitutional review pose a challenge to Möller’s conception, because they allow judges to ‘make decisions of a scope that would typically require democratic legitimacy’.Footnote 12 It is easy to see how arguments from failure that allow institutions to take over the functions of another branch, albeit temporarily, may fall short of these requirements.
Finally, arguments from failure in the context of federal coercion may undermine federalism. How important federalism is depends on context; that is, on the purposes federal arrangements serve in concrete cases. Federalism can be an essential tool to achieve integration within a nation while preserving space for diversity, granting a degree of regional autonomy to those concerned.Footnote 13 The stakes involved are therefore often high – as enduring conflicts in many countries, such as Spain, Canada, Britain and Ireland, Kenya, South Africa, India, the United States and others, demonstrate. Other advantages of federalism may be that it allows for regional experimentation and limits the risks of abuse of political power by dividing it. Federalism, thus, typically creates further veto players,Footnote 14 which can be a good thing, but also problematic when major reforms are necessary. In the more recent literature, federalism has also been advocated by some as enhancing democracy overall by allowing citizens to exercise more power over regional/local matters directly and thus stabilizing democratic regimes.Footnote 15
More could of course be said here with respect to all these principles. But ultimately the details will depend on the concrete case and institutional and legal context. Often, several principles will be involved, and sometimes it will be hard to distinguish between them. Thus, for example, the separation of powers serves the goals of enabling individual as well as collective autonomy, and thus democracy, and is also often considered an important part of the rule of law. It is therefore not surprising that insofar as specific uses of arguments from failure are debated in the existing literature, authors often engage with different principles at the same time.
2.2 Theoretical Resources
Once we start engaging more deeply with the constitutional theory literature, however, we can see that there is space for defending arguments from failure, at least in some forms and some cases. In what follows, I try to bundle some of the different strands in the literature into what seem to me the most important theoretical approaches to deal with arguments from failure. The first of these is the more recent literature on the separation of powers, both in Europe and in the Global South, where it often appears under the label of transformative constitutionalism; the second is the literature on democratic experimentalism. Both literatures occasionally overlap, but each offers important tools and ideas to address the normative stakes involved with arguments from failure, even though we may be sceptical of some of their claims.
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.
2.2.1.1 Transformative Constitutionalism
Challenges to the traditional model have come in particular from scholars writing on the Global South, who sometimes formulate their arguments in fairly radical terms. A range of scholars have criticized the traditional ‘liberal’ conception of constitutionalism as overly preoccupied with securing individual freedom as opposed to bringing about real transformation.Footnote 19 Pointing to the need for different state institutions to realize constitutional values and emphasizing the shortcomings of existing political processes, they have argued that we need to rethink the meaning of constitutionalism and the separation of powers. In particular, they have contended that a strict model of the separation of powers does not work in many Southern contexts. Instead, they have argued for a degree of flexibility and creativity, including, in some cases, the creation of new ‘fourth branch’ institutions to complement other branches and ensure good governance.Footnote 20 Such approaches are not confined to scholarship either. Thus, the South African Constitutional Court has adopted a flexible approach to the separation of powers and stressed the need for responsiveness between the different branches. Quoting Sandy Liebenberg’s concept of a dialogic relationship of the different branches, the Court has suggested that ‘the limits of each branch’s institutional power will be continually defined and redefined as they respond to the multifarious challenges of South Africa’s evolving constitutional democracy’.Footnote 21
Unsurprisingly, the details of different proposals vary though. Some authors have adopted what David Landau calls a ‘constitutional realization’ model, which seeks to endow courts with a more active role in the context of realizing ‘transformative constitutionalism’ as a countermodel to the traditional liberal understanding.Footnote 22 Though it remains impossible to define clearly what exactly transformative constitutionalism encompassesFootnote 23 – the early definition of Klare described the South African Constitution as ‘social, redistributive, caring, positive, at least partly horizontal, participatory, multicultural and self-conscious about its historical setting and transformative role and mission’Footnote 24 – it seems clear that it requires the state to pursue the transformative vision the constitution lays out.Footnote 25 The constitutional realization model thus seeks to relax traditional understandings of the separation of powers and give courts a broader role in order to realize core constitutional values.Footnote 26
Consider, for example, a 2016 article on the separation of powers by Dennis Davis, a prominent South African judge and academic commentator. Davis suggests that the separation of powers should be ‘developed to accommodate a variety of social, political and economic challenges’, given that ‘the core object of the Constitution which is relevant to this discussion is that a participatory democracy is incompatible with poverty or stark inequality’.Footnote 27 Similarly, another South African commentator, Tim Fish Hodgson, argues that ‘[A]ny interpretation of the separation of powers doctrine within South Africa’s specific constitutional design must be oriented towards the purpose of eliminating inequality and eradicating poverty’.Footnote 28 From that understanding, Hodgson argues, several consequences follow, among them that courts must have a role to play in cases where legislative tools of holding the executive accountable are not employed and that there cannot be a political question doctrine excepting certain kinds of question a priori from judicial review.Footnote 29
These statements are part of broader critiques of constitutionalism’s liberal bent, as well as of transformative constitutionalism itself, which is by now the subject of strong attacks as unable to deliver real change in South Africa.Footnote 30 This connection between the need for change and specific outputs and constitutionalism was spelt out explicitly and early on by Sanele Sibanda, who suggested that:
Put differently, in the South African context the failure to more directly and concretely connect the goals the Constitution’s transformative vision with ideals of delivering truly egalitarianism outcomes and poverty eradication creates room for a charge to be levelled that asserts that as currently conceived constitutionalism is viewed as an end in itself, rather than as a means to an end.Footnote 31
This approach opens the door to arguments from failure quite widely – and in many ways understandably so. In focusing on socio-economic inequality, it is addressing longstanding blindspots of liberal constitutionalism. If our key focus must be on realizing a constitution’s transformative vision in bringing about equality, and existing mechanisms and allocation of competences are failing to achieve this goal, then the overarching importance of realizing that vision suggests a flexible approach to those competences.
2.2.1.2 Recent European Scholarship
Interestingly, European scholarship has moved in a similar direction in recent years, albeit in a less radical key. In particular, Aileen Kavanagh has recently set out a more collaborative understanding of constitutionalism, requiring institutions to work together in a way that ensures ‘1) mutual responsiveness; 2) mutual respect and support; and 3) a commitment to the joint enterprise as a “shared cooperative activity”’.Footnote 32 Ideally, Kavanagh argues, each institution will perform the tasks it can do best and support other institutions in their respective roles.Footnote 33 Thus, while each branch plays a ‘distinct role’ in modern constitutional states, functions may be shared among them. The question is how far this reaches and what it requires in terms of the distinctiveness of the different roles of different branches. Kavanagh argues that even where institutions exercise functions shared with other branches, they will nevertheless perform their roles in different ways – and this ensures distinctiveness. For example, she readily accepts that courts may make law, too, but points out that they do so ‘in different ways’ to the other branches, more ‘piecemeal and incremental’.Footnote 34
Even earlier, Nick Barber put forward a similarly flexible conception of the separation of powers rooted in an idea of the state that is not merely a threat to individual liberty but provides a foundation for efficient government.Footnote 35 Though Barber acknowledges that the division of labour brought about by the establishment of different branches may be important, he argues against treating this division too rigidly. In particular, Barber suggests that institutional roles should be understood as interdependent, arguing explicitly that ‘where one institution is failing, the other needs to accommodate this’.Footnote 36 Barber thus views the separation of powers not as an absolute principle but would allow for limitations in what are essentially cases of dysfunction or failure. Even more explicitly, he argues that:
An extreme activist court may emerge to compensate for an exceptionally supine legislature; a competent executive may dodge enforcing the rulings of a corrupt judiciary. If the compensating institution is reformed, and brought back in line with the requirements of the separation of powers, the problem within the constitution may be exacerbated. Whilst in these systems the separation of powers should be worked towards, it should be applied carefully: the failing institution may need to be reformed first.Footnote 37
It is easy to see that from this perspective arguments from failure can be legitimate tools as part of a broader, more flexible approach to institutional competences. The advantage of this kind of more flexible approach put forward by Barber and Kavanagh is not only that its understanding of the separation of powers arguably better captures constitutional reality, but also that it is normatively attractive. This is true in particular if we focus on the need to realize certain constitutional goals or values in political systems that function less than well. It also speaks to the kind of argument for judicial intervention put forward first by David LandauFootnote 38 and most comprehensively in terms of responsiveness by Rosalind Dixon,Footnote 39 which I discuss in more detail in Chapter 5.
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 Discussion
These different strands in the literature have in common that they provide some room to argue for expanding existing powers of state institutions in cases of failure, at least in certain circumstances. The more recent European and South African literatures on the separation of powers and constitutionalism stress the need for legal flexibility. They emphasize the importance of enabling different institutions to collaborate to achieve key constitutional goals – in a South African context in particular the eradication of poverty and the achievement of a more equal society. The democratic experimentalism literature adds a concern with procedures, suggesting not just the ends to which institutions exercise their powers matter, but also the ways in which they do so, stressing the importance of inclusivity and participation. Both strands of literature, however, also raise important questions that we need to address.
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
urgency of resolving the particular dispute before them, the severity of the alleged interference with rights, and concerns about the appropriate division of labour in the constitutional scheme. Timing is also important. If the legislature is the best-qualified and most legitimate branch of government to tackle a particular issue, then it should be given time and opportunity to do so.Footnote 62
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 Working Out a Framework
In light of the above, the best response to the risks and challenges that come with arguments from failure is to take them seriously as a legal argument and be as clear as possible as to what they are and what they are not and under what conditions they might be legal.
This means, first and foremost, that we need to distinguish arguments from failure from other, more standard arguments about the need for purposive interpretation and the effective realization of rights and other constitutional values or goals, as I have argued in chapter 1. It is therefore crucial to remember that arguments from failure – unlike such broader arguments about updating constitutional norms and effective realization – are geared towards justifying a temporary exception from existing competences as they are commonly understood, rather than providing arguments for updating certain competences and for bringing about permanent changes. However, in practice, we often combine and blur them with other arguments into a broader functionalist approach to legal interpretation. This is precisely what we need to avoid. Such an approach feeds into an effectiveness paradigm of public law where each institution is accorded the power to do whatever it can to maximize output. Once this happens, we are not far from Somek’s dystopian depiction of ‘governance by situations’Footnote 73 where legality in its traditional sense has ultimately disappeared.
We also need to say more about its prerequisites: Under what conditions might one institution step in for another? What institutions are we talking about? What constitutes failure, and how can we distinguish it from political disagreement? And how can we make sure that allowing for such arguments does not pave the way for a new paradigm of public law, rooted in effectiveness and further removed from any democratic legitimation?
In what follows, I endeavour to answer these questions. Doing so means engaging with the central normative stakes and the literatures discussed above, which come with certain implications for the legality and legitimacy of arguments from failure. Together, these implications add up to a framework, albeit one that consists of a set of questions to ask whenever we encounter such arguments, rather than a set of ready-made answers. I will put some more flesh on this skeletal framework in the following chapters, where I address a range of typical settings and situations for arguments from failure.
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.
2.4.2.1 A Violation of Law
First, our failure should normally be anchored in some breach of law. In other words: While the failure or dysfunction itself may involve issues that are hard to capture in legal terms, such as the capture of an institution by external interests or problems of institutional design, we should understand in legal terms either the inaction or the problem that needs resolving. We should understand them, in other words, as a violation of law. While this is usually a relatively straightforward issue, it can be tricky to determine in individual cases if legal rules have been violated. That said, this kind of question is of course the bread and butter of courts.
2.4.2.2 Inability and Resistance
Thinking about the problem that needs fixing in terms of a violation of law makes clear that the failure in question must be understood in terms of the inability and/or unwillingness of the responsible institution to remedy that violation. The distinction between problems of capacity/inability and unwillingness/resistance has been emphasized in the literature on structural reform litigation, with some authors calling for more robust intervention in the latter case.Footnote 76 However, we should not exaggerate that difference. In reality, if we encounter an argument from failure, there is always at least a degree of unwillingness/resistance. This is because complete governmental inability would imply that no intervention can fix the problem, including interventions by other institutions that try to step in. What appear as cases of inability are therefore usually cases that will require multi-institutional coordination or oversight and/or involve higher costs. Those needs are not easy to meet. They therefore typically involve intervention at higher levels of government that has not already been forthcoming, because of negligence and/or resistance.
2.4.2.3 No Comprehensive Framework
The more there is already a set of checks and mechanisms of control provided for within existing institutional arrangements, the less room there is to invoke failure as a basis for acting outside or in addition to those arrangements. In other words, there exists typically a gap in the law when it comes to remedying failure, that is existing rules do not provide means to address the situation. We can borrow here from earlier writings on state necessity prerogatives, which similarly presuppose that there are gaps in the law in the sense that there are no explicit legal rules to address cases of emergency.Footnote 77 Thus, for example, if state practices lead to a massive and enduring violation of human rights/constitutional rights, we must ask if and to what degree there already exist mechanisms and institutions to tackle the problem.
If we seek to determine whether the existing constitutional rules are comprehensive, it also matters if a constitution is old and hard to change, as is the case for the US Constitution and many international treaties, or if it is more recent and perhaps easier to change. Where formal constitutional change is in practice almost impossible, arguments for a purposive interpretation as well as arguments from failure are more likely to be convincing.
And in consideration of alternatives, the legitimacy and capacity of the institutions involved also matter. Some institutions may be better suited to deal with the problem, and some may enjoy a greater degree of legitimacy. For example, when courts prescribe detailed policies from the bench, akin to a legislator, this is likely to be more problematic than when they interfere with the administration. Democracy will be at stake in the former case, but much less so in the latter. This in turn has implications for assessing which institutions may put forward arguments from failure more broadly.
2.4.2.4 Judicial Review
Given the associated risks of abuse, any exercise of power on the basis of arguments from failure should moreover be considered justiciable. As with many legal issues, analyzing arguments from failure will involve a more stringent judicial control in some respects – for example when it comes to identifying a violation of rules giving rise to the central problem – and a more restrained approach to other aspects. But this poses no uniquely difficult challenge. Many of the considerations relevant to failure involve fairly standard legal questions that judges can realistically be expected to grapple with, some vagueness notwithstanding. Of course, where courts themselves intervene, this can be difficult. But even here, unless the highest courts are themselves involved, appeal procedures provide some measure of control.
2.4.2.5 Further Considerations
It bears emphasizing that not all problems disappear in those cases where there is a legal basis for institutions to step in when others are failing. The standard example for such explicit failure provisions is clauses allowing for federal coercion, as briefly discussed in Chapter 1. Though this book is primarily concerned with the expansion of powers on the basis of failure without a specific basis in law, the frequently vague framing of such provisions leaves lawyers to fill in the details and in doing so to work to minimize the risks of abuses. In engaging with such written powers, the first task of any lawyer will of course be to grapple with the legal text itself. However, the considerations discussed in this book should be considered relevant in such cases, too. While there can of course be no question with regard to the existence of a gap in the law, the requirement of some legal anchor in the form of a violation of law together with an assessment of functionality, necessity and costs, as discussed in Chapter 2, should be understood to play a role here too.
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.
2.4.3.1 Military and Presidential Rule
To begin with, some institutions in constitutional democracies should never be allowed to step in, even where others are failing, because their intervention would be at odds with core principles of constitutional democracy. The most important example of this is the military. Military takeovers or coups often involve claims of taking over responsibility for the public welfare, sometimes in response to real problems as in Gabon,Footnote 78 NigerFootnote 79 or, indeed, Thailand.Footnote 80 This is because military coups do not simply stretch existing competences, but rather replace democracy with an authoritarian system wholesale, albeit for a temporary period (or so their defenders typically claim). This is distinct even from very activist courts seeking to address specific problems in light of persistent political failures to do so.
A more difficult question is whether executive interventions on the basis of failure should similarly be considered incompatible with constitutionalism. Consider intervention by presidents through emergency decrees or other executive decrees as an example. Thus, in the US arguments about congressional dysfunction and gridlock have long played an important role in justifying an expansive use of presidential powers including issuing executive orders, though arguably we are dealing with an expansive use of existing powers here rather than an assumption of new ones, and thus are outside the scope of arguments from failure.Footnote 81 Generally, the answer will have to be more nuanced here. In some systems, such as Germany, where presidents only hold a ceremonial role – in response to the Weimar practice of governing through emergency decrees – we should similarly reject intervention in such cases. The Basic Law was explicitly designed not to repeat the lessons of Weimar, and there is accordingly no gap in the system to be filled. In presidential systems without a comparable history, however, the situation is less clear. Here, too, though, the risks for abuse are considerable. It matters therefore if we are dealing with one specific targeted intervention in an individual case or rather a shift to a broader, albeit temporary, presidential rule by decree. The latter case involves another wholesale change of constitutional regime and thus falls outside of the scope of failure arguments – and this takes me to the next point.
2.4.3.2 Democratic Failure
Can arguments from failure justify interventions in response to democratic failure as such, rather than some more narrow instance of failure? This question partly goes back to the democratic experimentalism literature. If politics in a particular country are highly elitist, if the parliament is de facto unable to reach decisions on most issues because of a high degree of political fragmentation and/or polarization, if the electoral process is skewed by money and/or tailored to preserve the power of those inside the system, we may ask if and to what degree democratic procedures and institutions still deserve respect. Thus, David Landau has argued that where the democratic system in place is deeply dysfunctional, courts may step in to address important social questions.Footnote 82 Similarly, Roberto Gargarella has emphasized the dysfunctionalities plaguing a range of Latin American democracies and called not just for reforms in the ‘engine room’, that is in the political organization, but on that basis also defended judicial intervention, particularly where it has come in participatory forms.Footnote 83
I broadly agree with these approaches. However, with regard to arguments from failure, I would insist on proceeding on a case-by-case basis. In other words, democratic failure as such should not be taken as a justification for institutional expansion and should only be considered relevant insofar as it affects specific problems – for example because it means that certain violations of law will likely persist. The reasons for my more cautious approach lie once again in the risks that come with such general assessments of democracy.Footnote 84 What constitutes an improvement of one meaning of democracy may well constitute an infringement according to others.Footnote 85 Should, for example, the fact that certain minorities or indeed women are not statistically adequately represented in parliaments and face structural hurdles in the political arena lead us to consider a democracy insufficiently open and thus ultimately insufficiently democratic? Should we therefore adopt laws compelling parties to draft electoral lists that are equally representative, or are such rules in turn incompatible with democracy and the freedom of voters and political parties? Or consider the influence of money in politics. Do such influences at some point turn the system into a plutocracy no longer deserving of democratic respect, and if so, what point is that? And so on.
What is ‘democratic’ and what isn’t is thus not an easy question to answer, even within a specific domestic constitutional system. And the risks of allowing arguments from democratic failure are enormous.Footnote 86 This is not to say that there is no way to identify cases of democratic malfunction; indeed, sometimes it may be obvious that the democratic system is not working well. But in many others, it will not be.Footnote 87 Permitting this kind of general assessment invites not just over-generalizations and clichés, but it also opens the door to populists seeking to portray the political system as hijacked by elites disconnected from the real people. Thus, in Germany today, far-right supporters sometimes argue that there is no real democracy in the country, comparing the contemporary German governments to the former GDR.Footnote 88
We might ask if those risks can be contained if we were to restrict this kind of argument to institutions outside of the fray of party politics, such as courts. But courts and judges are hardly apolitical. Judges, too, may engage in judicial populism, portraying themselves as the ‘true friends’ of the people in contrast to corrupt political elites.Footnote 89 Generalized distrust is strongly linked to personal ideological preferences, and this makes it a suspicious ground for judges to expand their power. That said, there are some situations where dysfunctionality is not a matter of contest but rather common knowledge, and international courts or organizations may have assessed the situation in similar terms. It is largely this kind of situation that both Landau and Gargarella have in mind and where accordingly we can see the importance of such a broader assessment.
Nevertheless, arguments from failure should still focus on more specific cases of inaction or enduring violations of law by state institutions rather than be deployed in response to democratic failure as such, albeit the existence of the latter may have consequences for the former in concrete cases. This follows not least from their purpose: to allow only for temporary deviations from standard rules.
2.4.3.3 Disagreements, Legislative Failure and the No-More-Democratic-Means-Test
In many cases, different institutions and actors will disagree – both about what the law is and about what should be done to comply with it. And this is likely to give rise to accusations of failure. Where some will see a fraudulent election, others will suspect a sore loser. Where some see a violation of social rights, others will point out that governmental capacity and funds are finite. Can we meaningfully distinguish such disagreements from failure? The correct answer to these questions will depend on the case and institutions involved.
If we take representative democracy seriously when it comes to arguments from failure, we will be particularly worried about accusations of legislative failure offered to justify intervention by other, less democratic institutions, such as courts or governmental or administrative officials. It is easy to see how accusations of failure in this context may simply be opposition rhetoric, deployed by actors seeking to expand their own powers. This risk will be particularly high in presidential systems when the government is divided into different institutions that may be held by opposing political parties as is often the case in the United States. But disagreements about what constitutes adequate performance may arise in other contexts as well. Though this was not the actual legal justification offered, when the UK government under Johnson prorogued parliament, some commentators defended this as a necessary response to parliament’s refusal to fall in line with the government’s Brexit proposals and thus in terms of a kind of parliamentary failure.Footnote 90 As it was, the UK Supreme Court did not accept the government’s approach – and for good reasons.Footnote 91 Accusations of failure may play a role in other contexts as well. There are plenty of instances, for example, where governments have attempted to put pressure on courts in legally doubtful ways, often in response to previous jurisprudence, which governments consider to be judicial overreach.
The first and most obvious point to make in response is to recall that arguments from failure presuppose a violation of law. Constitutional norms sometimes stipulate legislative or parliamentary obligations, as do other statutes or indeed internal parliamentary rules. In other words, by not acting or acting inappropriately, parliament can sometimes be understood to violate its legal obligations. However, if parliamentary action or omission is considered to violate not constitutional but statutory law, this should generally be considered insufficient as a basis for intervention, given that parliament is violating rules authored by itself, which it may change. Thus, the mere refusal to deal with an urgent political problem – for example Brexit – without a constitutional basis for compelling such action should not be considered sufficient to warrant intervention by others unless such an omission violates norms of a higher status.
Insofar as there is a serious disagreement, it will thus have to be a disagreement either about the meaning of constitutional norms or about what to do with respect to a violation of constitutional norms, that is about how to remedy it. In this situation, a lot depends on which institution is seeking to step in. For example, when a court steps in to deal with rights violations according to its view, but not according to the government’s view, scope for governmental disagreement is usually limited, as I argue in Chapter 5. Courts are meant to have the last word on how to interpret legal rules, at least in most cases in most systems.Footnote 92 By contrast, if the executive or a president seeks to intervene on the basis of her understanding of rights, which is different from parliament’s, and/or we are dealing with a disagreement about how to respond, the situation is more difficult. Here, too, the principle of democracy suggests that, generally speaking, parliaments as representative institutions are better suited than the executive branch to develop and act on their own understanding of the constitution and to decide what needs to be done to respect, protect and fulfil them. In most countries, therefore, interventions by executive officers seeking to take over parliamentary tasks should as a default be considered illegal, with only some narrowly tailored exceptions to that rule, for example where the political party system and accordingly, the parliament have been clearly dysfunctional for a long time, as in Landau’s Colombian example.Footnote 93
And this leads us to a broader point I want to make: namely the need for a no-more-democratic-means-test for arguments from failure. In other words, institutions may only expand their powers to deal with a problem where there are no other more democratic means that would address the problem with comparable efficiency. Of course, such an analysis may be difficult to conduct, and in many cases, the fact that massive rights violations have persisted over a long time will have to suffice as an indication that no other more democratic means are likely available. In other cases, however, there may be significant political mobilization around the issue in question; there may be an ongoing parliamentary debate, and so on. If so, we should leave room for disagreement to unfold within the ordinary democratic channels. In such cases, stepping in will typically not be the appropriate response. In contrast, where parliaments step in to deal with issues traditionally reserved for other branches, we should be more open towards such interventions.
2.5 Conclusion
This chapter has shown how arguments from failure play a role in a range of existing debates on constitutional theory. In particular, recent scholarship on the separation of powers with its emphasis on effectiveness and the need for collaboration and the literature on democratic experimentalism open up space for such arguments. However, such approaches often come with high risks of opening the doors too widely to a general logic of efficiency, with potentially disastrous consequences for democracy and the rule of law.
The first and most important conclusion to draw from this, then, is that arguments from failure need to be as clearly delineated as possible if they are to be considered legal. It also means that arguments from failure presuppose a certain legal context where expectations of institutional collaboration are appropriate. Moreover, intervention can only ever be legitimate if the existing legal framework cannot be understood as comprehensive. Failure should be understood to stem from an inability or unwillingness to remedy rights violations and accordingly be considered justiciable. In addition, arguments from failure can only justify interventions in concrete cases or instances, rather than justifying a wholesale shift of power within the political system at large. And finally, the democratic stakes matter, and any intervention requires that there are no other more democratic ways to end the violation of law: It must pass the no-more-democratic-means-test.
Open access