1 - From Procedural to Substantive Accountability in EMU Governance

Summary

This chapter provides the volumes general conceptual framework. It begins by addressing why new approaches to accountability are needed, arguing that accountability literature has reached a stalemate as a result of an impasse between deductive and inductive approaches to accountability in the EU. It then argues that overcoming the stalemate requires developing a generalised framework of what accountability is for, deriving four accountability goods to be used in subsequent chapters. The chapter argues that each of the goods can be delivered in procedural or substantive ways, focusing either on the process by which decisions are made or the substantive worth of decisions themselves. The chapter concludes by discussing the strengths and weaknesses of both varieties of accountability before mapping out how the concepts will be applied across policy fields and institutions in subsequent chapters.

1.1 Introduction: Why We Need a Different Perspective on Accountability in the EMU

The Economic and Monetary Union (EMU) is a major achievement of European integration but also one of its major democratic accountability challenges. No other policy area of the European Union (EU) is as clear about ‘who gets what, when, how’¹ as the EMU. No other policy area has laid bare the necessity of controls – political, legal and administrative – over the exercise of power at the EU level as much as the EMU did during the euro crisis.² And yet the academic and political debate about democratic accountability in this policy field has reached a stalemate. On the one hand, it is widely acknowledged that the EMU suffers from structural flaws determined by a multi-level system that blurs conventional accountability lines between those who hold political authority in a representative democracy (the citizens) and those who make decisions on their behalf (in this case EU institutions).³ With respect to economic policy coordination, it remains difficult to disentangle the individual responsibility of finance ministers acting collectively in the Eurogroup and the Economic and Financial Affairs Council (ECOFIN), although their members are technically accountable to their respective national parliaments and electorates.⁴ The same problem exists when considering decisions made by heads of state and government in the European Council – an institution which has taken a clear leadership role during the euro crisis.⁵ The European Parliament (EP) has virtually no control over the intergovernmental institutions, while its legislative oversight of the European Commission in economic governance remains weak.⁶ In parallel, the Commission saw its powers expanded during the euro crisis by assuming key responsibilities for the coordination of national budgets through the newly introduced European Semester.⁷

On the monetary union side, the European Central Bank (ECB) has always faced criticism from an accountability perspective because its establishment effectively took monetary policy decisions away from Eurozone Member States and entrusted them to a technocratic institution, which, by several accounts, is the most independent central bank in the world.⁸ The expansion of the ECB mandate during the euro crisis has further complicated the situation, adding new accountability deficits with respect to unconventional monetary policies, financial assistance programmes and banking supervision.⁹ Last but not least, the Eurozone Member States established in 2012 an intergovernmental organization – the European Stability Mechanism (ESM) – designed to provide financial assistance to countries experiencing sovereign debt problems. Given its status outside EU Treaties, the ESM is subject only to a limited extent to scrutiny by national parliaments and not at all to scrutiny by the EP.¹⁰ To sum up, the EMU institutional structure illustrates in many ways the ‘impossible accountability thesis’ in the EU, according to which democratic accountability is simply incompatible with the EU’s multi-level governance model.¹¹

On the other hand, scholars have emphasized the need to improve EMU’s democratic accountability since its inception.¹² Demands to make the EMU more accountable follow the standard discourse on accountability in modern governance: at the basic level, accountability requires public officials – whether elected or not – to justify their conduct in front of a higher authority;¹³ at the next level, accountability ensures the possibility to punish those officials found lacking or allow them to make amends for past failures.¹⁴ An accountability relationship thus involves two parties: an account-giver – henceforth ‘the actor’ – who can be a person (member of a legislature, executive, bureaucracy) or a public institution, and an account-holder – henceforth ‘the forum’ – who can also be an individual (a direct superior, a minister, a parliamentarian) or an institution (parliaments, courts, ombudsmen, audit offices).¹⁵ In the policy discourse on accountability, the concept has multiple positive connotations, holding (1) the promise of democracy (by ensuring the answerability and responsiveness of elected officials), (2) the promise of control (through mechanisms designed to oversee executive and administrative action), (3) the promise of justice (through judicial and administrative review of government decisions), and (4) the promise of performance (through target-setting and incentivization of public officials).¹⁶

The problem, however, is that the institutional set-up of the EMU reduces the potential of such promises significantly. For example, in order to address the structural weaknesses of the EP in the EU political system, there is a proposal to institutionalize a subcommittee for Eurozone oversight that would call executive actors to account for their decisions.¹⁷ It is doubtful, however, that such a committee would deliver on the ‘promise of democracy’, given the well-known disconnect between citizens and EP elections – despite the institutional empowerment of the EP in recent years.¹⁸ In a similar vein, since the ECB mandate can only be altered through a cumbersome treaty change, there is pressure for the institution to narrow the mandate on its own – for instance, by excluding itself from financial assistance conditionality, limiting the purchase of government bonds to avoid redistribution or setting more specific objectives to measure the effectiveness of its banking supervision.¹⁹ To put it differently, given the constraints of the ECB legal framework, the ‘promise of control’ present in accountability discourse is largely a voluntary exercise – dependent on the ECB’s ‘willingness for control’ by oversight bodies like the EP and national parliaments, the European Ombudsman, the European Court of Auditors (ECA) or the European Anti-Fraud Office.

The mismatch between the structural flaws of EMU accountability and the incremental proposals put forth to reform the system suggests that academic thinking about accountability on the topic is at a stalemate. There is a gap between what is seen as necessary and what is feasible in the EMU governance framework – given the complications of multi-level decision-making within a hybrid institutional constellation of intergovernmental and supranational actors. In this chapter, we identify the cause of the stalemate in the parallel development of deductive and inductive approaches to accountability in the EU. Most deductive approaches typically apply national accountability benchmarks on EMU and subsequently find numerous shortcomings in their institutionalization at the EU level – especially when it comes to the role of parliaments. In contrast, inductive approaches start from the EU’s Treaty framework on EMU and subsequently infer standards of accountable behaviour for different EU institutions. The problem is that the EU cannot meet national benchmarks for accountability, while the principles set in the EU Treaties are too narrow and hence decoupled from generally applicable accountability standards.

To break the stalemate, we propose a new deductive framework for studying accountability more suitable to EMU and the EU setting – which can be applied and drawn upon in subsequent theoretical and empirical chapters within this collection. Drawing on public administration literature, legal scholarship, and liberal and republican thinking in political theory, we develop a normative conceptualization of accountability that seeks to answer a basic question: ‘what is accountability good for?’ Accordingly, we identify four normative ‘goods’ of accountability: openness, non-arbitrariness, effectiveness and publicness. We show that existing mechanisms of accountability can address the normative ‘goods’ in two ways: one centred on the processes through which actors take decisions (procedural accountability) and the other focused on the merit of the decisions themselves (substantive accountability).²⁰ We argue that there are both pay-offs and trade-offs in choosing one alternative over the other. While procedural accountability brings clarity and predictability for the people involved in the process, it tends to detract from the underlying goals of accountability’s four normative goods because it draws public attention away from the policies public officials pursue and their effects to the procedures by which they do so. In contrast, substantive accountability is more complex and costly to achieve but has the merit of maintaining the normative ethos of the concept. After analysing various aspects of EMU accountability, we conclude that procedural accountability dominates: a finding that we encourage our authors to critically explore.

The chapter is structured as follows. We begin by explaining the stalemate of the EMU accountability literature, caught between deductive approaches focused on national benchmarks of democratic accountability and inductive approaches emphasizing contrasting interpretations of different principles set by EU Treaties. We show the need for a deductive, normative perspective that is applicable to the EMU without being specific to it. The second section introduces the four normative goods of accountability and describes the possibilities to enforce them in a procedural or substantive way. The third section applies the new conceptualization to examples across the EMU in order to show how political and legal institutions deliver the normative goods of accountability procedurally and substantively – with an emphasis on the former. The fourth section highlights the limits of procedural accountability, making an argument against its prevalence in the EMU governance structure. We conclude with a call for both our authors and other scholars to apply the distinction between procedural and substantive accountability to different policy fields and institutions within EMU.

1.2 The Stalemate of Accountability Research on EMU: Between Deductive and Inductive Approaches

To make sense of the accountability literature on EMU, we propose a distinction between deductive and inductive approaches to accountability. The basis of the classification is the reasoning behind accountability assessments: how do scholars judge whether an actor has acted accountably in the EMU? Do they start from general definitions and seek to apply them to specific institutions like the ECB? Or alternatively, do they first examine a given institutional setting, for example, the legal framework of the ECB, and then derive general accountability standards applicable thereof? The distinction between deductive and inductive methods is well known in scientific inquiry. From Aristotle to Francis Bacon to William Whewell, philosophers have discussed two directions of the scientific method: the first begins with general and fundamental principles that are then applied to specific cases (deduction), while the second starts with the specific of what is observed and then moves to general and fundamental principles (induction).²¹ Although this chapter focuses on the EMU, the distinction between deductive and inductive approaches applies to accountability research more generally.

In fact, accountability research benefits from the new classification in two ways. First, the distinction between deductive and inductive approaches transcends regular disciplinary boundaries dividing the study of accountability²² – especially visible between political scientists and legal scholars. Browsing through the relevant academic literature, we can identify deductive and inductive studies which examine all the classic institutional mechanisms of accountability, regardless if they are political (elections, parliamentary scrutiny of the executive), legal (judicial review), administrative (investigations by ombudsmen, auditing and anti-fraud offices) or managerial (hierarchy in a bureaucratic organization).²³ Second, the deductive/inductive dichotomy encompasses definitions of accountability that are either normative or descriptive, as distinguished by Mark Bovens.²⁴ From a normative standpoint, what matters are standards for accountable behaviour, which can be either general (deductive) or specific to a situation (inductive). From a descriptive perspective, the interest is in [the appropriate] institutional mechanisms of accountability, which can be borrowed from other contexts in a comparativist effort (deduction) or inferred on an ad hoc basis from the experience of selected actors (induction). Consequently, deductive or inductive studies can have an explicit normative focus on the accountable behaviour of actors or a more analytical focus on the institutional arrangements of accountability.

In the EMU accountability literature, deductive approaches revolve around two general standards of accountability: ensuring democratic control and preventing abuses of power.²⁵ This is visible among authors who underline the accumulation of executive power in economic governance after the euro crisis and the necessity to increase the role of parliaments as a countervailing power.²⁶ While the EP had been marginally involved in economic governance prior to the crisis, national parliaments actually saw their budgetary and fiscal monitoring powers reduced since the institutionalization of the European Semester.²⁷ The importance of parliaments for democratic accountability is grounded in an understanding of the concept as the counterpart to delegation in the ubiquitous principal–agent model. The logic is straightforward: if ‘A is obliged to act in some way on behalf of B’, then ‘B is empowered … to sanction or reward A for her activities or performance in this capacity’.²⁸ Transposed to the EU level, what is required is to (re-)build the democratic accountability chain from voters to elected representatives but especially from elected representatives to executive actors.²⁹ Parliaments are thus crucial in closing the gap between citizens as the ultimate principals of economic decisions and various executive agents such as the European Council, the Council or the Commission. Accordingly, scholars argue that the EMU could improve its democratic accountability credentials by empowering the EP³⁰ and national parliaments³¹ in terms of both decision-making and legislative oversight of executive actors. Despite having different analytical foci, these studies share an implicit normative assumption that parliamentary involvement in economic governance can help Member States – and hence the EU – ‘meet their legitimacy obligations to their own publics’.³²

In monetary affairs, deductive studies focus on ‘preventing the development of concentrations of power’ and ensuring an appropriate system of checks and balances of the ECB.³³ The emphasis here is different because the ECB is a non-majoritarian institution whose need for independence from electoral competition has been one of the cornerstones of the EMU since its creation (Article 130 TFEU). The EP is thus often cited as an ‘accountability forum’ and not the principal of the ECB in monetary policy and banking supervision; consequently, the ECB ‘owes’ the EP transparency and justification of decisions but not obedience or even political responsiveness.³⁴ Conversely, the importance of preventing abuses of power is much stronger in the area of legal accountability, that is, judicial review of ECB decisions by national and EU courts.³⁵ In legal accountability, deductive and inductive approaches are sometimes intertwined, as courts apply the general standard of ‘curtailing the abuse of executive power’ in reference to existing regulations.³⁶ A combination of deductive and inductive approaches can be found for instance in the work of Markakis, who examines the accountability of the ECB in relation to the price stability objective prescribed in Article 127(1) TFEU.³⁷

Conversely, studies that are ‘purely’ inductive use the EMU legal and institutional architecture as the starting point for evaluating accountability. A clear example is offered by case-law analyses of ECB or ESM instruments. In this category, scholars do not apply an overarching accountability definition in order to evaluate judicial decisions; conversely, they selectively employ the Treaty framework in order to identify specific features, for example, the independence of the ECB, which are then connected to different headings and degrees of judicial review, for example, the duty to state reasons as displayed in Gauweiler.³⁸ As a result, there is room for contrasting interpretations of the stringency with which the Court of Justice of the European Union (CJEU) should uphold different Treaty principles. For instance, with respect to the ECB, the introduction of the first unconventional monetary instrument – the Outright Monetary Transactions (OMT, 2012) – divided lawyers on the question of the violation of the Treaty’s ‘no-bailout’ clause³⁹ and the extent to which national courts and the CJEU should intervene to hold the ECB accountable for potentially acting ultra vires.⁴⁰ The ‘no-bailout’ clause had also featured in Pringle⁴¹ in relation to the establishment of the ESM, which similarly divided scholars on whether courts should interpret Treaty articles in light of background teleological expectations about current political circumstances.⁴² In this context, several scholars emphasized the deferential approach of the CJEU towards EMU executive actors, as judicial decisions failed to question if austerity was indeed demanded by ‘the logic of the market’,⁴³ or whether the instruments adopted during the euro crisis were substantively justified by ‘a logic of emergency’.⁴⁴ Regardless of whether studies criticize or endorse court decisions on the EMU, the inductive approach to evaluating accountability is obvious – as the benchmarks for accountable behaviour are derived from specific features of the EU Treaties.

Another variant of inductive studies examines the normative peculiarities of EMU governance. In a recent article further developed in this volume, Steinbach argued that the ‘normative choice for accountability’ in the EMU need not be democratic in the principal–agent sense of empowering political institutions (especially parliaments); conversely, EU economic governance has created an accountability regime of its own, subject to the judgement of the market.⁴⁵ According to Steinbach, the orientation towards the market is institutionalized in the EU Treaties, which highlight the Union’s commitment to create ‘an open market economy with free competition’.⁴⁶ Moreover, the emphasis on the ‘free market’ is seen as a constitutional norm which ‘implies the absence of state intervention in the market-based price determination process’.⁴⁷ Accordingly, states and private actors in the EMU are/should be accountable to markets rather than attempt to create a political accountability regime in a flawed democratic system. In fact, Steinbach sees current criticism of EMU accountability as the result of attempting to substitute economic accountability with political accountability, for example, in financial assistance when creditor states and EU institutions have taken over the position of accountability forum from the market.⁴⁸ The issue is whether accountability to the market is normatively justifiable in a democratic system; after all, the belief that the market will hold actors accountable ‘just the right amount’ is rooted in ordoliberal assumptions of political economy that have been seriously challenged since the crisis.⁴⁹ As a recent study by the ECB acknowledges in relation to fiscal requirements for price stability, a certain paradigm of political economy ‘became constitutional law in Europe before the economics profession could even prove [its] value’.⁵⁰

Overall, the point of this review is to show that there are problems with both deductive and inductive approaches to EMU accountability. First, deductive studies drawing on principal–agent theory are stuck in a vision of accountability designed for the nation-state that is simply unattainable in the specific institutional setting of EMU governance. National parliaments and the EP are important accountability forums, but they cannot be expected to deliver in the same way as legislatures within national democratic systems of government. In the EMU, the delegation chain from voters to elected representatives to executive actors is either short-circuited (in the case of the EP) or too tortuous to function properly (in the case of national parliaments). In contrast, inductive studies display different problems, namely the replacement of general accountability standards with situation-specific benchmarks assessing the performance of an actor in a given setting. In EMU, the EU Treaties constitutionalize certain principles (like the ‘no-bailout clause’ or the ‘free-market orientation’) that remain open to political contestation and ordinary decision-making domestically.⁵¹ One of the difficulties of this constitutionalization is that EU institutions can be held accountable for the extent to which their decisions comply with Treaty principles but not for the principles themselves. Political accountability, however, may concern the overall principles governing EMU (a level of contestation that inductive studies – as they are oriented by these principles themselves – cannot capture). Normative standards of accountability should be broader than policy-specific benchmarks for holding actors accountable in a particular context. We introduce such an approach in the next section in relation to the EU setting.

1.3 The Four Normative Goods of Accountability in Modern Governance

The objective to conceptualize accountability beyond the nation-state is usually associated with global politics.⁵² In this context, Michael Goodhart has observed that all approaches to global accountability share a similar understanding of the term, namely the ‘question of making those who wield power answerable to the appropriate people’.⁵³ In his view, this standard definition is based on a Westphalian notion of the state that is unworkable in world politics. In the EU, the absence of a European demos and the resilience of national demoi⁵⁴ implies that accountability cannot be organized around the ‘appropriate’ forum because this is simply not feasible in a large-scale political unit where citizens have few opportunities to influence governing decisions.⁵⁵ In this context, Goodhart proposes to shift our thinking about accountability ‘from who is entitled to hold power to account to the reasons why accountability is justified in democratic theory’.⁵⁶ His interest is in the nexus between democracy and human rights, linking accountability to emancipatory human rights norms that ‘constrain the exercise of power and enable meaningful political agency’ in the global arena.⁵⁷ We agree with Goodhart that the conceptualization of accountability beyond the nation-state must include normative standards for holding power to account. However, we believe that focusing on human rights reduces accountability to the role of an instrument necessary to achieve other democratic objectives rather than giving it credit as a democratic goal in itself. Accordingly, we hold that any meaningful conception of accountability must begin with an understanding of the normative goods to which accountability is aimed. Drawing on liberal and republican thinking from political theory and the broader public administration literature, we distinguish between four such goods.⁵⁸

The first good is openness. Liberal thinkers from Bentham onwards have long argued that public confidence in official action is likely to be increased where public policy is conducted under the public gaze (what he termed ‘publicity’).⁵⁹ The openness of public policy has thus been linked to a number of public goods, such as the avoidance of corruption,⁶⁰ the improvement of public knowledge and the republican demand that free citizens should enjoy ‘non-domination’ through the ability to question and contest official action.⁶¹ We might therefore want accountability because we see it as a device to ensure that public action is open, transparent and contestable.

The second such good is non-arbitrariness. There is a deep tradition in accountability research of tying accountability to notions of principal–agent theory in which accountability is a device for (political) principles to control (administrative) agents to whom they have delegated powers.⁶² This is a narrower instance of a broader accountability good, namely that those who wield public power should do so in a limited manner and that they should exercise coercion only to the degree necessary to achieve their goals.⁶³ Non-arbitrariness is also therefore linked to more general limits on public action such as human rights or due process guarantees that seek to regulate the relationship between the individual and the state.⁶⁴ Accountability – by making officials answer for conduct – provides a means by which arbitrary distinctions or applications of power can be identified and later remedied.

The third good which accountability seeks to render concerns effectiveness. While openness and non-arbitrariness seem highly normative values, accountability may be sought for more utilitarian reasons, namely that accountable officials are more likely to deliver high-quality services. From this perspective, accountability holds the promise of performance.⁶⁵ By making an official answer for their conduct, and by offering the possibility to correct potential errors, accountability is a mechanism to improve the efficacy and responsiveness of public policy.⁶⁶ Here, the premise is that the need to justify and even correct conduct will likely improve, and encourage reflection upon, the design of policy-making or implementation.

The final such good is one of publicness or that official action should be oriented towards the common good (and therefore justified by public or universal reasons).⁶⁷ This involves demonstrating both that officials were not personally enriched and that their decisions are fairly balanced, taking into account different societal interests and perspectives. Once again, accountability is a key device for ensuring the publicness of official action in this sense – when parliamentarians scrutinize government agencies, or courts conduct judicial review, a key demand is that actors show how their activities forwarded the national or collective interest (with different accountability forums likely to disagree on what a fair balancing of societal interests would entail).⁶⁸ Accountability is thus a device to advance the normative good of public policy grounded in the public interest.

Having established the normative goods of accountability, the question is how they can be delivered in practice. Our proposal is to distinguish between procedural and substantive ways of providing the four normative goods of accountability. To put it simply, actors are procedurally accountable if they can demonstrate that the processes or steps they followed in performing their tasks were open, limited (non-arbitrary), effective and/or public. By contrast, actors are substantively accountable if they can demonstrate that the decisions themselves or the outcomes to which they led were open, limited, effective and public. We further explain the distinction below.

1.4 Providing Accountability Goods: Procedural versus Substantive Ways

How can accountability be procedural, and how can it be substantive? The simplest way of understanding the distinction is through the categories of public law.⁶⁹ In this context, judges often distinguish between reviewing parliamentary acts on procedural or on substantive grounds.⁷⁰ When conducting a procedural review, a judge will enquire into the robustness of the process through which a parliamentary act was adopted.⁷¹ When conducting a substantive review, what is important is not the process of adopting an act but its substantive provisions per se and their likely impact. To give an example, if a Court is enquiring whether a statute setting out minimum requirements for religious schools infringes the right to freedom of religion, it might assess the infringement either procedurally (did Parliament consider the impact of the bill on freedom of religion, or incorporate the views of religious minorities, when adopting it?) or substantively (is the statute likely to infringe religious freedom, or is it in fact neutral vis-à-vis different systems of belief?).

If we transport this distinction to the world of accountability, procedural accountability suggests an accountability relation oriented around the process by which a particular decision was rendered. If we are holding an actor to account procedurally, we are calling them to account for, and justify, the procedural steps they undertook in forming or executing a policy decision. If we are holding an actor to account substantively, by contrast, we are calling them to account for and justify the substantive worth of the policy decision itself. Thus, a parliamentary committee examining the implementation of the bill above might also seek to hold a school inspectorate either procedurally or substantively accountable. Procedurally, they might ask how often religious schools had been inspected or how parents of religious minorities had been consulted in drawing up guidelines for inspection. By asking such procedural questions, the committee is not calling into question the substantive worth of the inspectorate’s decisions but confining itself to examining the steps the inspectorate took to fulfil its mandate.

Parliamentarians might also seek, however, to hold the inspectorate substantively accountable – did the implementation of the statute achieve the goals (e.g. of improving school standards) it originally sought, or why did the inspectorate choose to prioritize the inspection of one set of schools or one aspect of the school curriculum over another? In the latter case, what is at issue is not the form of decision-making but its substance, that is, did the actor being held accountable make substantively worthwhile, just or efficient decisions? The official is thus being held accountable against a substantive rather than procedural benchmark: they are being asked to explain and justify the worth of their action.

From a conceptual perspective, process is either unimportant or instrumental in this case – the inspectorate could be judged by parliamentarians to have implemented the statute justly or effectively even in circumstances where the procedure by which they had done so was inadequate just as the inspectorate could demonstrate a robust, transparent and inclusive procedure yet still be seen by parliamentarians as substantively failing to adequately explain the correctness or efficacy of their decisions. In simple terms, the two ‘forms’ of accountability carry different lenses through which to understand whether accountability has been adequately rendered.

The notions of procedural and substantive accountability overlap to a certain extent with the distinction between process and outcome accountability found in social psychology.⁷² The interest there is in the micro-behaviour of individuals and how they respond to different types of evaluation standards set by accountability forums, which can focus on processes or outcomes. In contrast, we take a macro-level approach and discuss the abstract forms (procedural or substantive) through which the normative goods of accountability can be delivered in practice. Our concept of accountability is therefore not relative to a specific accountability forum, but it is centred on general democratic ‘goods’ considered inherent in the term. For the purposes of illustration, these are applied to the EMU context in the next section. The purpose of the edited collection is to expand and apply these goods in a manner more comprehensive than this schematic overview can provide.

1.5 Procedural and Substantive Accountability in the EMU

The four normative goods of accountability can be identified across the EMU governance architecture. The examples below focus on diverse cases that are ‘representative in the minimal sense of representing the full variation of the population’.⁷³ The purpose is to show the predominance of procedural ways of providing the normative goods of accountability in the EMU.

Let us start with the first good – openness. Transparency has been for decades a key concern of accountability research.⁷⁴ Most importantly, however, transparency seems a good that can be fulfilled without a demand for substantive justification. An official can therefore satisfy the demand for openness procedurally by providing the public with information and documents on a regular basis. In this scenario, the task of accountability forums, such as parliaments, auditors or courts, is to enquire into the procedures by which citizens can access official information and to demand reform if these procedures are found wanting. Many accountability requests in the field of EU economic governance are of this nature. To take a specific example from banking supervision, the largest number of questions asked by Members of the European Parliament (MEPs) to the ECB are requests for information, seeking to address information asymmetries between the two institutions.⁷⁵ Keeping in mind that ‘transparency is a necessary but insufficient condition for accountability’,⁷⁶ it would make sense for MEPs to first ask for information before acting on it substantively. But when political accountability does not move beyond transparency requests, the value of openness remains procedural.

The value of openness can also, however, be met substantively. The test of substantive openness is not the de jure but the de facto openness of official action. To be substantively open, an official must not simply provide information, or demonstrate transparent procedures, but provide information in a sufficiently relevant and timely way that it is likely to be used by accountability forums, such as parliaments, courts or citizens.⁷⁷ The test of substantive accountability is thus one of whether official action is in fact regularly probed and contested. This establishes an obligation on accountability forums too, namely that they utilize their information rights to understand policy decisions and to make clear to the public the substantive choices, including the achievements and errors, of public actors. To return to the example above, the key difficulty in the parliamentary accountability of the ECB is not only the availability of information but its volume and complexity.⁷⁸ MEPs simply do not have the expertise to identify the most relevant or salient questions that would allow them to substantively challenge ECB decisions.⁷⁹ Substantive openness often requires additional resources, raising difficult questions about who should bear its costs.

The second good – non-arbitrariness – also carries procedural and substantive elements. Procedurally, public actors are commonly bound by statutes or other rules which specify their substantive mission. When adopting legislation, or making specific decisions, officials are commonly under a duty to explain why particular decisions are necessary to fulfil their mandate (which may be subject to judicial review).⁸⁰ Similarly, public institutions may adopt procedures to ‘mainstream’ rights-based limitations into their policy-making, that is, to conduct impact assessments or other exercises by which officials may demonstrate that human rights have been taken into account.⁸¹ Here, arbitrariness is limited procedurally in the sense that public actors bind themselves via process-based limits on their action; subsequently, accountability forums such as courts and parliaments are able to verify these limits; for example, was a rights-based impact assessment conducted? To provide an example from the EMU, the Commission committed, as part of the Juncker Commission’s promise to improve the EMU’s social dimension, to produce social impact assessments to assess and minimize detrimental social rights implications of future EU financial assistance. Such an assessment was conducted in relation to the third Greek bailout.⁸²

Substantively, however, non-arbitrariness carries a higher bar. As with openness, the important element is whether public action was de facto limited and non-arbitrary. This would require not only that actors are bound by limits but that they demonstrate how limits constrained their activities. Non-arbitrariness also concerns whether a given policy arbitrarily discriminates against a given group in society or infringes an individual’s rights. From the perspective of an accountability forum, the key question would be whether a particular group in society or (for a judge) a core autonomy right is disadvantaged by virtue of how a policy has been designed. To return to the definition of the procedural/substantive distinction of the previous section, the existence of a procedure to mainstream human rights considerations within an institution would not fulfil this requirement if the outcome of such mainstreaming violated a core right. To continue the example of financial assistance, while the Commission indeed conducted a social impact assessment for the third Greek bailout, academic commentary on this assessment has been highly critical.⁸³ These criticisms range from the inadequacy of the assessment to the question of how it actually fed into policy-making (with no indication that it led to any meaningful changes in how the stability programme for Greece was designed or implemented). The key test for substantive accountability as non-arbitrariness is thus whether policy choices in EMU plausibly aimed for and achieved non-arbitrary results.

The third accountability good, effectiveness, would seem to be inherently substantive in nature – as it concerns whether planned policies resulted in particular outcomes. Nevertheless, the ‘explosion’ of auditing in the 1990s as a means to control the government suggests that effectiveness can also be implemented in a limited procedural way.⁸⁴ In 2016, the ECA evaluated the operational performance of the Single Supervisory Mechanism (SSM) and the role of the ECB thereof.⁸⁵ The evaluation report included a section on the ‘difficulty in obtaining audit evidence’, as the ECA officially complained that the ECB provided the auditing team ‘very little’ of the information required.⁸⁶ The ECB was given the chance to respond to the report, emphasizing its different understanding of ‘sufficient information’ for auditing purposes. Specifically, the ECB claimed that ‘all audit evidence covered by the Court’s mandate to audit the “operational efficiency of the management of the ECB” had been provided’ and that any exception concerned documents that were not related to the SSM’s operational efficiency.⁸⁷

The problem here concerns the legal limits imposed on auditing the ECB’s substantive effectiveness by the ECA. According to EU primary law, the ECA has ‘full power to examine all books and accounts of the ECB’ but only with respect to examining the ‘operational efficiency of the management of the ECB’ (Article 27, Protocol no. 4 TFEU). The conflict between the two institutions regarding the SSM report reflects the different interpretations of ‘operational efficiency’ and the substantive need for information to make such an assessment. In the absence of said information, the ECA’s report was reduced to evaluating procedural aspects such as staffing, for example, whether on-site inspections of banks should be run by ECB staff as opposed to representatives of national supervisors.⁸⁸ Despite these legal limitations, the ECA clearly understood the value of substantive effectiveness, as the report recommended that the ECB develops a public and formal performance framework to assess the effectiveness of its supervisory activities.⁸⁹

Lastly, the final accountability good of publicness can also be rendered procedurally or substantively. In both cases, publicness requires that public officials demonstrate the orientation of their conduct towards the common good. In the EU context, this may take on a specific meaning, namely the duty of EU actors to demonstrate that their policies (such as in EMU, country-specific recommendations) take the interests of the EU as a whole into account (and not just of selected industries or states). A common mechanism to achieve accountability as publicness concerns establishing procedures for public participation in government action from notice and comment procedures to more intensive forms of citizen participation.⁹⁰ An accountability forum may therefore assess whether a robust process of consultation existed when adopting public policies. This can also include proportionality review in the sense that such review commonly requires, in order for limitations on rights to be justified, that officials demonstrate that their policies restricted rights in pursuit of a ‘legitimate aim’.

In the context of EMU, examples of both can be found. In the 2015 ‘Five Presidents Report’, the leaders of the EU’s main institutions committed to greater involvement of the social partners in fiscal coordination processes such as the European Semester, with some commentators arguing that this has led to a gradual ‘socialization’ or re-balancing of the Semester’s policy priorities.⁹¹ Elsewhere, several ECB programmes – most notably the OMT, the Asset Purchase Programme (APP) and the Public Sector Purchase Programme (PSPP) – have been subject to judicial challenge, with the CJEU asked to rule on whether their effects on national economic competencies were adequately grounded in the ECB’s mandate of producing stable prices for the Eurozone as a whole.⁹²

Substantively, publicness concerns not just the existence but the effects of participation and reason-giving. With respect to participatory governance, many studies of the phenomenon in different national and transnational settings complain of its elitist, or perfunctory character, questioning whether participation actually leads to policy change.⁹³ Substantive accountability regarding publicness would therefore concern the question of how participation affects outcomes, or how the knowledge garnered via participation was utilized in the policy process. In respect to proportionality review, substantive accountability would not merely require officials to posit the aim to which their policies were directed but would allow scrutiny of the suitability and necessity of those policies, given those negatively affected by them (such as those whose fundamental rights were infringed).

Once again, demands for substantive accountability as publicness can also be found in the EMU context. To return to the examples above, while the participation of civil society actors in the European Semester may be an end in itself, the tying of this participation to a debate about the Semester’s policy priorities illustrates the weakness of accountability in this policy context. The real question is whether the involvement of civil society in the Semester matters or is simply another abstract commitment boldly stated in policy documents only to be safely disregarded when substantive decisions about EU fiscal policy are made. Similarly, those analysing proportionality review of the ECB have repeatedly questioned whether such review in the area of monetary policy is meaningful or whether the standard of review provides the ECB with such a margin of discretion as to make judicial review practically meaningless (or ‘incomprehensible’ as provocatively put by the German Constitutional Court).⁹⁴ In this sense, what matters for accountability as publicness is not the mere provision of reasons for action relating to the common good but whether these reasons meaningfully orient the conduct of economic policy-makers.

With all four categories, the guiding distinction between procedural and substantive accountability is between process and merit. For the former, the connecting point between actor and forum is the steps taken to make public action accountable; for the latter, the basis for interaction is the merit of official action vis-à-vis alternatives. This leaves a crucial question: from an institutional design perspective, why would anyone choose an accountability regime focused on one form of accountability, rather than the other? This is the subject of the next section.

1.6 Procedural Accountability – and Its Limits

There are good reasons why institutions may favour procedural accountability. The main reason concerns the clarity and predictability of the standards used to orient accountability. Under the substantive reading, a potentially broad set of standards are at play, with the actor under a heavy and potentially limitless justificatory burden. To take one of the four categories discussed above, requirements of publicness in public policy are complex and may be subject to significant disagreement.⁹⁵ This diversity in interpretation applies to both actors and accountability forums. Assuming that officials may be accountable – in a complex and ‘networked’ modern polity – to multiple accountability forums, representing an array of interests, finding an appropriate balance able to satisfy these interests can be an overwhelming task.⁹⁶ In the EMU context, the ECB’s bond-buying programme was of such volume and complexity to have potentially limitless consequences on a wide variety of societal interests; under the circumstances, how could the ECB adequately demonstrate that this policy was non-arbitrary, effective and oriented towards the common good? Substantive accountability widens the set of standards orienting the conduct of the actor, potentially confusing both public officials themselves and the accountability forums that must scrutinize them.

In this regard, procedural accountability seems to be much more straightforward. Here, the primary duty of the actor is to follow an established process (under the assumption that, if followed, this process will lead to open, non-arbitrary and effective outcomes oriented towards the common good). The job of the forum is then to verify that the correct process has been implemented. As a result, a significant burden – of calculating and adjusting conduct according to its concrete effects – is lifted and externalized. This may be particularly important in a judicial context, where judges may lack both the knowledge and legitimacy to interfere in complex economic debates. In social psychology, experimental research has demonstrated that procedural (or process) accountability is linked to situations when actors lack knowledge about the specific outcomes they are expected to achieve; consequently, they shift their focus to the quality of the decision-making process.⁹⁷ This suggests that procedural accountability can easily become decoupled from substantive outcomes.

The potential ‘replacement effect’ of procedures for substance therefore represents the first major limitation of procedural accountability. Indeed, there are important implications to ‘lifting the burden of substantive justification’ for actors. Jane Mansbridge has famously discussed the interplay of material and moral incentives in the process of delegation.⁹⁸ As she argues, the introduction of material incentives may destroy moral incentives by making what was once a duty (to act in the right way) a material question (is this action in my interest?). A similar risk applies in relation to procedural accountability. As discussed above, procedural accountability is based on the assumption that, if actors follow the correct procedures, they will then orient their activities towards correct substantive outcomes. The risk is that the opposite applies – that if actors begin to care only about procedures at the expense of the substantive goods, those procedures are designed to secure (and without reflecting on their adequacy). According to Roy Heidelberg, this logic reflects a technical conception of accountability that ‘allows for an actor to dismiss criticism of a policy or action by appealing to an obedience to procedural rules or, at worst, to justify doing the wrong thing in the right way’.⁹⁹

In the EMU, the substitution effect of procedural for substantive accountability is significant. As we have argued elsewhere, ECB accountability in particular is grounded in procedural devices such as transparency, which are seen as better suited to the institution’s operational independence.¹⁰⁰ Nevertheless, transparency obligations tend to carry the functional mission of the ECB to the outer limit, as accountability debates become hijacked by discussions over the secrecy regime of the ECB.¹⁰¹ As a result, the focus on transparency limits the substantive contestation of the actual merit and distributive implications of ECB decisions. While accountability forums like the EP are thus permitted and even encouraged to ask questions of ECB officials, their ability to contest and seek to influence the direction of monetary and supervisory policy is minimal.¹⁰² Accountability in substance seems needed but is also excluded a priori by the EMU institutional structure.

Second, procedural accountability is limiting from the perspective of accountability forums. As indicated by the classic accountability literature, one challenge faced by accountability forums is information asymmetry, or the difficulty of holding accountable actors with greater knowledge of a specific topic. In principal–agent theory, agents are expected to ‘shirk’ their obligations before principals either by hiding information before they are appointed (adverse selection) or by hiding their behaviour while on the job (moral hazard).¹⁰³ Procedural accountability extends these problems, as explained by Heidelberg:

Because the accountor is exposed to the rules and subject to them, it is not unusual for the accountor to have a better understanding of how the rule system works than the accountee, in which case, the rule system is as much an instrument for the accountor as the accountee.¹⁰⁴

The discussion about the exceptions to the application of the Stability and Growth Pact (SGP) by the Commission is a clear example of an actor exploiting the rule-based system introduced by the excessive deficit and macroeconomic imbalance procedures. In fact, the Commission’s past record of sanctioning some Member States but not others for budgetary or macroeconomic transgressions has raised important accountability questions of arbitrariness and equal treatment.¹⁰⁵ MEPs have regularly accused the Commission of unfair treatment of a few Member States, for example, for the failure to take action against France’s excessive deficit in 2014 or Germany’s macroeconomic imbalances in 2015.¹⁰⁶ The problem is that the relevant legislation – the Two-Pack and the Six-Pack – gives the Commission ample discretion to calculate budget deficits by deciding which expenditures, fluctuations and one-off investments are taken into account and which are excluded.¹⁰⁷ This has two important implications. On the one hand, the Commission benefits from information asymmetries in fiscal governance, as it is the only institution with the [expert] knowledge to navigate the ‘maze of alternative or even-conflicting rules, part legislative, part non-legislative, which few understand’.¹⁰⁸ On the other hand, the Commission can make strategic use of these rules in order to explain away decisions not to sanction some Member States, as shown by the justification given for the French and German decisions to MEPs during hearings in 2014 and 2015.¹⁰⁹ Overall, procedural accountability seems to favour the actor at the forum’s expense.

To sum up, of the two forms of accountability, which is more suitable for the EMU? The argument of this section was not that procedural accountability is of limited value universally. In fact, there are advantages to procedural accountability, namely its clarity and predictability, which make it an important avenue for providing the normative goods of accountability in the EMU. This may be particularly important in an EMU constitutional settlement that gives particular institutions – namely the ECB – high operational independence. But without a substantive component, the normative goods of accountability remain focused on processes of decision-making, which are insufficient for evaluating decisions in a policy field that decides ‘who gets what, when, how’¹¹⁰ as much as EMU.

Substantive accountability can be more complex and costly to achieve; for instance, in order to balance information asymmetries between forums and actors, the former would need to acquire expertise in many fields, which requires additional resources. In some cases, this might be infeasible – no matter how desirable or demanded substantive accountability becomes. Under the circumstances, political architects of EMU accountability will have to decide if the costs of substantive accountability are worth the payoffs or, alternatively, if they are willing to accept the trade-offs between substantive and procedural accountability which the EMU’s institutional set-up currently entails.

1.7 Conclusion

In this chapter, we introduced a new normative framework for analysing accountability in the EMU. The framework has been determined deductively by surveying the relevant literature in political theory, law and public administration in order to identify four goods that accountability is supposed to ensure: openness, non-arbitrariness, effectiveness and publicness. All of these can be achieved in a procedural or substantive way, but the latter imposes higher standards of accountable behaviour for actors – as they have to demonstrate the merit of their decisions rather than defend the process through which those decisions have been reached. Different types of accountability forums can examine an actor’s conduct in line with the four goods: parliaments through legislative oversight, courts through judicial review, ombudsmen and court of auditors through administrative review. When transposed to the EMU, there seems to be a predominance of procedural accountability that remains limited because of the ‘replacement effect’ of procedures for substance and the weak position of accountability forums vis-à-vis actors with expert knowledge of the policy area. Our argument is that there is a clear need for more substantive accountability in the EMU across different mechanisms of accountability – political, legal and administrative: a claim that we invite our authors to question, probe and elaborate in the chapters that follow.

We have sought through the chapter to add to the academic literature on EMU accountability by constructing an approach that is deductive without being rooted in a nation-state view of the concept and which identifies normative standards of accountability that are not inductively derived from the EU Treaties alone. The benefit of the approach is that accountability is evaluated on the basis of norms, rather than owed to a specific principal – part of a would-be democratic chain of delegation. Acknowledging that the term is often used interchangeably with ‘answerability’ or ‘responsiveness’,¹¹¹ we argue in favour of a change of analytical optics: instead of obsessing about the appropriate forum to whom actors should answer or respond, scholars should focus on the question ‘what should actors be accountable for?’ This is particularly applicable at the EU level, where the distance between the ultimate democratic principal (the citizens) and their supposed agent (EU institutions) remains great. We thus call on other authors to break the stalemate of EMU accountability research by researching the extent to which national and EU institutions provide the four normative goods of accountability, and subsequently, by showing how account-giving by EU institutions can be made more substantive in the future. We hope to have provided useful conceptual tools for the more empirical papers that follow this chapter.