INTRODUCTION: DEMOCRACY, COURTS AND THE DILEMMAS OF REPRESENTATION

The first important work of comparative law and politics to define, analyse and evaluate the global phenomenon of the ‘judicialisation of politics’ was Tate and Vallinder's The Global Expansion of ...

must be made by consensus-an impossible goal. Majority rule, which treats all views neutrally and impartially and is positively responsive to those views most favoured by voters, proves sufficient; offering a public mechanism that gives equal concern and respect to each citizen's interests and views (May 1952). Not only are there grounds, associated with the Condorcet jury theorem (Condorcet 1976), for believing that, when conditions justifying democracy prevail, majority rule is more likely to produce decisions in the public interest than rule by any minority, however well informed, but also democracy offers an alternative to guardianship by an elite with the consequent risks of arbitrary rule (Dahl 1989).
In complex and pluralist societies elite judgements about the public interest-no matter how conscientiously made or expertly informed-will always risk partiality, reflecting a limited knowledge of how citizens will come to think and act in the future. Few if any social and economic issues admit of purely expert solutions. Causal relations are so complex that almost all policies produce unintended consequences, as well as raising normative and evaluative issues for which no consensus is justified. Democracy responds to these epistemological difficulties by operating as an information-gathering process which allows each person affected by a decision an equal influence upon it; constituting a direct or indirect means whereby different views and interests can be related to each other in ways that treat them with equal concern and respect (Christiano 2008).
Ever since the eighteenth century and the American and French Revolutions, it has been customary for democratic decision-making in the West to be mediated by 'representatives', who are somehow or other authorised to act in lieu and on behalf of nominally equal citizens. Representation might seem prima facie to be undemocratic, with the representatives forming an elite who actually rule rather than the people themselves. At best, it substitutes an equal, if imperfect, degree of control over the selection of decision-makers for direct participation in decision-making. Yet, the selection process may indirectly impact on decision-making too. To see why, we need some preliminary conceptual clarifications of what 'representation' and 'representative democracy' mean.
Hannah Pitkin has defined political representation as 'acting in the interest of the represented, in a manner responsive to them' (Pitkin 1967: 209). A representative 'acts for or in the place of the represented' for some practical purpose, being an 'agent . . . related to the represented as "another self", as somehow "standing in his place" in the practical arena' (Diggs 1968: 29 -30). Pitkin's account yields four main ideal types of representation: delegation, in which representatives 'act as' the represented; trusteeship, in which representatives 'act for' the represented; and descriptive and symbolic representation, in which representatives 'stand as' and 'stand for' the represented respectively (Pitkin 1967). A delegate acts 'as' the relevant principal instructs him or her, so that 'the principal acts through the representative . . . in the sense that he shares responsibility for the act of his representative' (Diggs 1968: 32 -3). By contrast, a trustee enjoys much more discretion, being significantly more removed from the principal's will (Pettit 2006). Trusteeship is bestowed by a grantor, who in a democratic context may be either a high official or the public themselves, to act 'for' a third party beneficiary (such as the people) who is not available, best placed or, in some cases, considered fully competent, to act on their own account, and who may have only a limited say in the selection and control of their representative (cf. Griffiths and Wollheim 1960). The normative basis of trusteeship is substantive representation in which an agent serves-but may also define-the interests of the principal. Finally, descriptive and symbolic representatives depict the represented by virtue of either some shared characteristic, such as might emerge from a statistical sample of the relevant population, or a symbolic quality with which those they represent identify.
Although analytically distinct, all four forms of representation are related in various ways. Delegation and trusteeship form a continuum, ranging from those who are provided with a strict mandate to act in a certain way in a specific and somewhat limited context-such as proxy voters at a shareholders' meeting who cast their vote as instructed by their principal, to a trustee who has been empowered by others to look after the interests of one incapable of doing so for him or herself-such as the legal guardian of an orphaned newborn. Yet between (and even within) these extremes lie a variety of forms of representation (Pitkin (1967: 121) lists five, and see Bellamy and Castiglione 2011: 122-3) in which trustees are delegated to some degree to act within a given brief, and delegates can exercise their discretion to a certain extent within the constraints of their mandate. For example, a lawyer standardly employs her expertise to act for a client in the manner of a trustee. However, in so doing she must also act under instructions even if these have been granted by a person or body other than her client or merely involves her client's passive assent. Potentially, therefore, her actions may be censured for lack of due care or her advice overruled. Consequently, she also acts in some measure as a delegate of the client either directly or indirectly, via the grantor. Meanwhile, the capacity of agents to act 'as' or 'for' their principal is often enhanced by the degree to which they are viewed as standing 'as' or 'for' them due to certain shared characteristics-be it a shared gender, ethnicity, social background or set of ideological, cultural, religious or other beliefs-or an accepted symbolic attribute, such as monarchs are commonly held to possess.
Within a democracy, how people are represented will change according to context and be shaped by the processes of authorisation and accountability that define the representative's role. These processes delineate the ways and degree to which representative agents act in the interests of and respond to their principals, as well as the extent to which they stand 'as' or 'for' the people by virtue of being themselves in some way or other 'of' them. They also structure how far representatives can be deemed to do so in a democratic manner, by virtue of being authorised by democratic means or being held accountable for their pursuit of democratic ends. Thus, representatives who are subject to strict instructions from the electorate, and subject to recall should they diverge from their mandate, act literally as their voters direct. Here the difference between selecting decision-makers and making decisions is negligible. Similarly, the ancient democratic practice of sortition or a modern version involving sampling supports the selection of representatives who may stand 'for' the people and act 'as' them through being sufficiently similar to them to be representative of their interests and ideas. Here the process of authorisation aims at a democratic outcome even if it does not involve an explicitly democratic input. Yet, neither of these mechanisms is widely used or usable today (Bobbio 1987). Politicians have to take too many decisions, many of which are unforeseeable or involve technical expertise, for them to be subjected to too tight a mandate. Likewise, the people of modern democracies are so diverse along so many dimensions, that sortition offers a haphazard way of selecting representatives who could stand 'for' the people, while sampling begs the question of what features 'of' people should be politically represented.
These problems are in certain respects generic to the very nature of representation and its central paradox: namely, that it involves making present what is not present (Pitkin 1967). Representatives are constituted by a deliberate act of will that stands apart from the representative agent, opening up a 'gap' or source of alienation between them and the represented that can never quite be closed, and which in a certain sense 'sets up' the representative to go adrift (Cheibub and Przeworski 1999;Epstein and Sharyn 1994). At once, the danger arises that the representative's attributes differ from those of the represented enough to INTRODUCTION call his or her representativeness into question. For the range of representatives among whom principals must choose may itself involve biases (Huber and Shipan 2000;Rogowski 1981). For example, in contemporary democracies candidates for public office are drawn nearly always from the upper or upper middle classes (Esaiasson and Holmberg 1996;Norris 1997). Only a minority even of this class can surmount what are often forbidding material barriers to standing for office or attract enough favourable media attention (cf. Manin 1997;Norris and Lovenduski 1995). These circumstances combine to act as a filter, a de facto vetting process (Pitkin 2004). In addition, the voters' discernment may be seriously impaired by intrinsic cognitive weaknesses exploited by the media and other elites (Cox 1997;Hobolt 2009). All of these shortcomings impose severe prior limitations on the voter's act of will, over and above the intentionality gap that emerges between principal and agent (the so-called 'agent drift'). Pitkin even laments that representatives have come to threaten democracy, having 'supplanted [it] instead of serving it'. Representative government has become 'a new form of oligarchy . . .
[o]ur governors have become a self-perpetuating elite that rules -or rather, administerspassive or privatized masses of people', with the result that 'ordinary people [are] excluded from public life' (Pitkin 2004: 335, 339; see also Guarnieri 2013). Others have portrayed modern democracy as 'an elective aristocracy' (Manin 1997: 145) or 'competitive oligarchy' (Manin et al. 1999: 4), since voter choices are confined to an elite class that is already unrepresentative.
As a result, it is often thought that at best representatives can act as trustees. This position is identified above all with Edmund Burke, and given its influence his critique of delegation and defence of trusteeship merits some discussion. Burke claimed that representatives ought not to be chosen because their views coincided with those of their voters but as individuals who could be trusted to exercise their independent judgement in an informed, conscientious and impartial manner. As he informed his prospective constituents in Bristol, though representatives ought to pay attention to the opinions of their electors, government and legislation were matters of judgement and reason and they elected representatives to employ these intellectual and moral powers on their behalf.
Certainly, gentlemen, it ought to be the happiness and glory of a representative to live in the strictest union . . . and the most unreserved communication with his constituents. [. . .] But his unbiased opinion, his mature judgement, his enlightened conscience, he ought not to sacrifice to you, to any man, or to any set of men living. These he does not derive from your pleasure; no, nor from the law and the constitution. They are a trust from Providence, for the abuse of which he is deeply answerable. Your representative owes you, not his industry only, but his judgement; and he betrays, instead of serving you, if he sacrifices it to your opinion. (Burke 1774) On the one hand, representatives ought to possess sufficient experience and expertise for their reasoning to be superior to that of most of their electors As he put it when, true to his word, he championed free trade with Ireland in the teeth of his constituents' decided preference for protectionism (Ayling 1988): Faithful watchmen we ought to be over the rights and privileges of the people. But . . . I should be ashamed to show my face before them, if I changed my ground as they cried up or cried down men or things or opinions,-if I wavered and shifted about with every change . . . as to the detail of particular measures, or to any general schemes of policy, they have neither enough of speculation in the closet nor of experience in business to decide upon it. They can well see whether we are tools of a court or their honest servants . . . but of the particular merits of a measure I have other standards. (Burke 1780) On the other hand, and in many respects more importantly, he contended deliberation in parliament would hardly be meaningful unless representatives could reason independently. If representatives were delegates of particular interests then they would be unable to deliberate with other representatives on the general interest of the nation. A representative had the duty of being a Member of Parliament rather than a member for a given constituency: Parliament is not a congress of ambassadors from different and hostile interests; which interests each must maintain, as an agent and advocate, against other agents and advocates; but parliament is a deliberative assembly of one nation, with one interest, that of the whole; where, not local purposes, not local prejudices, ought to guide, but the general good, resulting from the general reason of the whole. (Burke 1774) The democratic credentials of the Burkean representative are weak but not non-existent. Representatives are authorised by the electorate, and Burke credited them with the good sense to be able to choose wise and informed decision-makers, even if they were not up to making wise and informed decisions for themselves. Trustees are also accountable to those who have chosen to trust them, who may throw them out if they feel that trust has been abused or their representatives simply turn out to be less wise and informed than they had thought. However, these democratic aspects of his account have usually been thought insufficient to remove its essentially elitist caste.
The accuracy of Burke's argument as an account of actual existing representative democracies also proves questionable. In certain respects, and despite the problems noted above, contemporary politicians are far closer to delegates than Burkean trustees, and can avoid some of the problems Burke associated with delegation. Unlike Burke, today's politicians campaign not so much as individual reasoners as members of parties and on the basis of their party's manifesto (Klingermann et al. 1994). These manifestos are drafted with an eye to winning electoral support and so attempt to accord with the opinions-the express preferences-of sufficient voters to gain election. Moreover, within the legislature politicians tend to be whipped to adopt the party line, and remain remarkably faithful to their electoral commitments. Nevertheless, that need not mean that representatives are delegates of particular interests in a narrow sense. Because they are for the most part elected in national elections that in parliamentary systems lead to the formation of governments, incentives exist at the electoral phase to produce programmes of government capable of winning the support of at least a majority of public opinion-a requirement that the need in many systems for governments to build and sustain coalitions continues into the legislative phase. Meanwhile, parties have ideological commitments that constrain the arguments and policies they will choose in order to appeal to the electorate and the alliances they will make.
Admittedly, both the electoral and the legislative process prove less high-minded than Burke had aspired for them to be, with bargaining often more important than arguing and reasoning. Given the diversity and complexity of modern societies, noted above, it can be difficult for parties to accommodate all views and interests sufficiently adequately to give them an equal voice. Though pluralist theorists of democracy have noted that in modern societies the frequent need to build majorities on the basis of shifting coalitions of minorities makes the 'tyranny of the majority' less likely than liberal theorists imagined, it remains a logical and occasional real possibility that groups with less bargaining power than others will be INTRODUCTION unfairly marginalised in the democratic process. Meanwhile, and in some respects more significantly, the need for coalition making can alienate voters if they feel it entails representatives making compromises that sacrifice principles to power-a feeling likely to be exacerbated the less voters feel they can identify with politicians (Bellamy 2012). As Norberto Bobbio noted many years ago, the practices of any workable representative democracy seem doomed to offer citizens a series of 'broken promises' by failing to live up to the largely unachievable standards set by ideal participatory theories of democracy (Bobbio 1987: 26 -7).
Some commentators have seen the courts as a potential remedy for these democratic shortcomings: [P]residentialism and parliamentarism are not the only forms of democratic system on offer, as the political scientists customarily claim. It is logical to think that the third branch -courts -could also be dominant in some sort of democratic governments as presidents and parliaments can be. What seems to get in the way of this idea, however, is the persistent view that courts are undemocratic, therefore unsuitable for being the lead institution in a democratic polity. (Scheppele 2005: 44) Nevertheless, they claim that the flaws of representation and representative democracy, especially that 'what voters voted for [i]s not what they got', are such that it may be argued that courts can be 'more popularly responsive' and 'democratically thoughtful' bodies than parliaments (Scheppele 2005: 52).
Courts can be representative in some ways. A common law system especially, which is built on judicial precedent instead of a legislated code, empowers courts to carry on, independently of legislatures, functions that may merit being called representative. An enormous number of varied interests come before the courts over time, each advocated by its own attorney. Of course, a similar range comes before political representatives as well. However, we saw how commitments to a broader constituency or their party and sheer time constraints may mean that they fail to represent them adequately. By contrast, attorneys are obligated to act on behalf of their clients alone. As we noted, attorneys act as both delegates, following their clients' instructions, and as trustees, who have the expertise to steer the case through the legal jungle (cf. Griffiths and Wollheim 1960). As such, they do not presume to know their clients' interests better than they do, merely how best to pursue them most effectively in court. In these respects, attorneys may be more democratically representative than elected politicians.
Courts that involve a common law trial jury can also offer a form of representation in the sense of sortition or sampling. However, to be fully effective, jurors must be aware of their power of nullification (cf. Spooner 1852). A striking aspect of contemporary judicial practice in the United States, where the judicialisation of politics began (Dezalay and Garth 2002;Tate and Vallinder 1995), has been the marginalisation of the jury (DeWolfe Howe 1939; Ostrowski 2001)-to the point where many commentators complain it has become little more than a rubberstamp whose independent judgment is routinely suppressed by the 'instructions' of judges who may not be impartial (Federal Judicial Center 1996: 100;Sand and Rakoff 1993). 1 When functioning properly, though, the jury may be said to be reasonably representative of the community's conscience over a large number of cases.
Arguably, though, courts are limited as representative fora because the least representative actors in any legal system are the judges. With the exception of some American states, judges are not elected and cannot be voted out of office. This precludes counting them as 'delegates' of the people, even if the elected representatives who appoint them respond to direct or indirect political pressures when doing so. At this point, some commentators, invoking Burke, try to assimilate judges to trustees (Sajó 2012;Stone Sweet 2003). The analogy, however, is tenuous, as the Burkean trustee is not only authorised directly by his constituents, but is directly accountable for breaches of his fiduciary duties to them-be they the result of corruption, bad luck or incompetence (Pettit 2006). As Burke remarked, notwithstanding the independence of the representative's policy judgement, his constituents 'can well see whether we are tools of a court or their honest servants', and can vote out those who prove to be the former or whom merely displease them (as they famously did to Burke). Yet, at least in Europe, the trend has been toward removing the judiciary from any democratic authorisation and accountability. Instead, judges are increasingly authorised by and accountable to no one but themselves (Guarnieri 2013;Parau 2012Parau , 2013Piana 2010). If this trend becomes consolidated, the supposed 'beneficiaries' of their 'trusteeship' will find themselves in no position to ensure judges perform their duties.
In so far as representation by attorneys and juries is mediated by judges, the claim that courts can act as democratic representatives proves decidedly limited. Courts may constitute an alternative venue for 'representation' (cf. Saward 2010), but only 'by default' as it were, in that contemporary polities offer a limited repertory of institutions to which one can turn to further one's claims. More importantly, courts may work for some people and types of claim, but, like parliaments, may not work for others. Every court case features a loser as well as a winner, and those with deeper pockets and better organisational resources tend to fare better than the less well-off or harder to organise.
Judges undoubtedly resemble political representatives in having to arbitrate between conflicting interests, but this shared feature highlights the judicial role of politicians rather than the representative character of judges. A higher law like the King's writ or a statute of parliament or a written constitution may be 'committed' to judges for interpretation, but the irreversibility of the commitment and the wide scope for judicial discretion in legal interpretation erodes the comparison with even the weakest forms of democratic trusteeship. Judges may act as trustees of a decision (such as a statute, constitution, or international convention) that was made or legitimised democratically. However, trustees can betray their trust. Indeed, the common law of trust is concerned to a large extent with the remedies that are available in those cases where the trust is abused by the trustee (Pettit 2006). Two key questions arise as a consequence: what happens when judges go beyond their allotted task of upholding democratically endorsed decisions? And how can one tell when they have done so? Judges might be thought to have betrayed the trust placed in them when their decisions reflect their own preferences rather than those of the elected representatives who crafted the laws, constitutions or conventions in question. Such documents are notoriously vague and incomplete, and in any case must be deployed to meet individual cases and new circumstances that their democratic framers could never have anticipated. Hence, judges will always need to exercise some discretion. The difficulty lies in providing criteria and a process for deciding when discretion has ceased to be democratically legitimate, and a remedy for those situations where this is believed to be the case.
The process of judicialisation has raised these issues from the beginning and continues to do so across different jurisdictions and levels of governance, be they national or supranational courts. From classic US Supreme Court cases such as Dred Scott v. Sandford, 60 U.S. 393 (1857), which infamously held that African Americans, whether slave or free, fell outside the protection of the Constitution and so could not be US citizens, and denied the federal government had the right to regulate slavery in the territories; to some of the latest decisions of INTRODUCTION contemporary courts, such as that of the Chamber of the Second Section of the European Court of Human Rights (ECtHR) in Lautsi v. Italy banning crucifixes from Italian classrooms, or the cases of private property restitution in Romania after 1989 that resulted in elderly tenants who had been born on the property and lived there all their lives being evicted onto the street (Parau 2012)-courts have made decisions that are questionable in democratic terms, issuing neither from a democratic process or serving democratic ends. Of course, they have also made decisions that may be unpopular with sections of the electorate and their representatives yet are arguably supportive of democratic rights, such as the ECtHR's decision in Hirst v. UK (2), which questioned the imposition of a blanket ban on voting by all prisoners serving a custodial sentence, whatever the severity of their crime or length of their sentence. Courts may sometimes serve democracy better than parliaments. Yet, as the popular reaction to the Hirst decision illustrates, when they can be said to have done so is rarely uncontroversial, and the ability of parliaments to override their decisions on such occasions has become increasingly difficult as the power of judges has grown. This special issue does not propose to solve this conundrum but merely to explore its different aspects from a variety of political and socio-legal perspectives. This goal is achieved by examining the dynamic between courts and representative democracy at various stages of the judicialisation process, from prospective institutional design, through concrete actuation, to retrospective evaluation.
Cristina Parau explores judicial empowerment in terms of institutional designs that are prior to any actual policy-making by judges, showing how certain templates of judicial governance of transnational origin, which fundamentally alter power relations between judges and elected representatives, have become the norm in Eastern Europe amongst the new EU members and the latest candidate for EU accession. The disempowerment of elected representatives is a major consequence of this trend. It has been driven by transnational networks of legal professionals and by the 'dormancy' of parliaments. It is hard to avoid the conclusion that the transnational legal professional community has simply assumed a trusteeship role. Eric Ip addresses the institutional origins of the UK Supreme Court, arguing they lie in 'rational' interactions between political actors. He analyses the power dynamic involving the Supreme Court and parliament, finding that as the courts continue to be deferential to parliament so parliament is induced to reciprocate, complying with Supreme Court rulings even when they impinge on its supremacy. In their respective ways, then, Ip and Parau both discuss the role of parliaments in the judicialisation of politics.
By contrast, Daniel Kelemen argues that the Court of Justice of the European Union (CJEU) and the ECtHR may be understood as consistently democracy-reinforcing. Both can lay claim to democratic authorisation via the appointment process, and both serve democratic ends. Lisa Vanhala and Rachel Cichowski explore in their turn how far resort to the courts may be regarded as itself a democratic process. They examine the important role of civil society activists in initiating the process of judicialisation by bringing policy issues before domestic and supranational courts which formerly would have been decided by democratically elected representatives. They then evaluate the implications for representative democracy at the domestic level, concluding that judicialisation is a double-edged sword for representation.
Richard Bellamy and Carlo Guarnieri likewise explore the consequences for representative democracy of the judicialisation of politics but come to more negative conclusions. Bellamy considers the claim made by many legal and political theorists that courts have in many respects superior democratic qualities to the electoral process. He critically reviews three key arguments to this effect: the first maintains that courts provide an exemplary form of deliberative democracy based on public reason; the second that courts allow for citizen participation, and the third that courts provide a mechanism for representation. He contends that though courts do uphold and even presuppose certain democratic norms, such as treating all as equal before the law, they possess these features and exercise them in a democratic manner not through being independent of or above electoral democracy, but because courts depend upon and are a part of the broader democratic political system. It is the extent to which that remains the case that is the focus of Guarnieri's contribution. He reviews the trends evident in the latest reforms of judicial institutions and their impact on the role of judges in Europe. He notes how the costs and benefits of the move towards ever greater judicial independence and its implications for representative democracy have yet to be adequately assessed.
Finally, Ran Hirschl provides a commentary on the special issue as a whole. He situates the various contributions in the context of an overview and assessment of the global movement towards the judicialisation of politics. His Epilogue draws on the empirical research investigating the correlations between judicialisation, on the one side, and democratisation and human development, on the other. He concludes that there is little reason to believe that judicialisation advances either of them. The link between democracy and development proves far stronger, while democracy upholds the rule of law rather than vice versa. In sum, courts may supplement democratic politics in many important ways, not least by offering new avenues for representing the interests of citizens. However, their capacity to do so depends on their being embedded within an effective democratic social and political system. The dilemma prompted by the judicialisation of politics is that if taken too far it risks undermining the very democratic structures that render it both legitimate and effective.

1.
US v. Thomas, 116 F.3d 606, 614 (2nd Cir. 1997): 'Nullification is, by definition, a violation of a juror's oath to apply the law as instructed by the court -in the words of the standard oath administered to jurors in the federal courts, to render a true verdict according to the law and the evidence'. This suggests that the modern practice is repressive of jury nullification. On the other hand, the precedents of the common law are clear. For example, in Bushel's Case