‘Real-existing democracies’ (REDs) seem to be in real trouble.Footnote ¹ Academics and practitioners tend to agree on this and both can produce long lists of ‘morbidity symptoms’ to illustrate it. Most of these lists would include items such as decline in electoral turnout, falling party membership and identification, greater volatility in voter preferences and, hence, electoral outcomes, greater difficulty in obtaining and sustaining majority support for governments, decrease in trust in politicians, parties and political institutions in general, declining centrality of parliament, and increased devolution of authority to administrative bodies.

The one thread that connects all of these symptoms is representation and, even more specifically, representation through political parties competing in elections. Could it be that what are no longer working as they used to and, therefore, generating most of the disaffection among citizens are the partisan channels for aggregating, deliberating and deciding among competing interests and passions? If so, the crisis would not be of democracy itself, but just of a set of institutions that have come to be closely identified with it. And thus the solution is to be found not in getting rid of this type of regime, but in transforming or re-dimensioning the role played in it by political parties and elections.Footnote ²

Government and Opposition has generously put at my disposition a number of recent books that deal in quite different ways with representation. Not surprisingly, most of them presume that political parties are the primary, if not exclusive, channel through which citizens are represented; regular, free and fair elections between these competing parties provide the most important mechanism for holding rulers accountable; and, together, parties and elections determine the legitimacy of the political process as a whole.

The following is not so much an orthodox review of these books, as a selective assessment of what they have to contribute to an understanding of whether or not representation lies at the source of the present crisis in REDs and whether or not political parties and elections are specifically responsible for this alleged crisis in representation. Needless to say, the books reviewed make important contributions to knowledge about many other topics. What I will be looking for will be answers to the following: (1) How does the author (or authors) define the notion of representation in REDs? (2) What institutions are considered to channel and contain the process of political representation and have they remained constant? (3) What are the units (‘constituencies’) that are being represented and at what level of aggregation? (4) What are the characteristics of those persons who claim to be acting as representatives, and how have these changed over time? (5) Finally, how have the citizens who are supposedly being ‘represented’ reacted to any observed changes in the institutions, characteristics or units of those who claim to be doing the representation? Of course, not all of the books being reviewed will provide answers to the questions. None of them pretends overtly to address the meta-issue of whether a crisis of representation in REDs does exist, but all of them provide observations and insights that might help us to discover if this is the case.

Maurizio Cotta and Heinrich Best's Democratic Representation in Europe is by far the most ‘substantial’ book in our collection. It is the product of a major collaborative effort covering the social and political backgrounds of parliamentary deputies in some ten or more countries for almost 150 years. Its database (DATACUBE) is unparalleled in scope and detail and destined to remain an essential source for tracing the historical development of liberal, representative democracy in Europe for many years to come. In an earlier volume the data was analysed country-by-country.Footnote ³ In this one, some 21 authors explore comparatively ‘dimensions of variation’, ‘variations across party families’ and comprehensive topics such as cleavage patterns and elite transformations.

The conception of ‘democratic representation’ is conventional, that is, it focuses exclusively on persons who win competitive elections for positions in parliament. Presumably, the authors would accept the caveat that it would have been apposite to include those persons who lost in those elections and even those who unsuccessfully placed themselves in nomination since the democratic legitimacy of the winners rested on the presence of these other components of ‘free and fair’ competition. In other words, the ‘losers’ and not just the ‘winners’ formed part of the emerging political class and contributed to the eventual success of democracy within the region. Presumably, the enormous data-gathering demands precluded such an effort.

What is deliberately excluded from this vast panorama of emerging and evolving democracy in Europe, however, is representationby persons elected not by political parties to non-legislative positions and, even more, representation by persons selected either by themselves or by specialized publics. This includes virtually all of the institutions of civil society. Maurizio Cotta in his conclusion seems to concede this when he alludes to the ‘almost invisible political dyad’ that has surrounded political parties and representative democracy.Footnote ⁴ If, as some public opinion surveys seem to indicate, the representatives chosen by this ‘selectorate’ are more trusted (and, perhaps, more relied upon) than the politicians chosen by the ‘electorate’, then we may well question whether the ‘deeply innovative’ and ‘increasingly (inclusive)’ historical role of the latter will persist into the future.

The editors signal an awareness of ‘the spreading of feelings of dissatisfaction and distrust vis-à-vis the representative institutions and the men and women embodying them’Footnote ⁵ and suggest that this makes a retrospective assessment of their role all the more worthy. But this hint of a contemporary crisis in representation gets overwhelmed by the enormous richness of data and complexity of substance in the rest of the book. To the extent that a subliminal message can be detected, it is a comforting one: partisan and legislative representation has successfully adjusted to changes in the scope of the electorate and the scale of the government. The social background of deputies has changed; the process has been (largely) incremental; and even when disruptions have occurred, the incoming representatives have been rapidly socialized by the incumbent ones. So, even if there are some contentious issues surrounding today's parties and parliaments, the competitive electoral process will resolve them without any need to redesign, much less to transform radically, the institutions involved.

Here and there, however, some less optimistic findings emerge in Cotta's conclusion. The most alarming is the growing evidence of ‘party based professionalism’, the failure to move beyond ‘the confines of an educated middle class’, and the return of ‘public employment as the leading occupation’ of deputies.Footnote ⁶ All of these suggest an expansion of social and cultural distance between the elected and their electors and a concomitant contraction in the authenticity of representation. What if, as Cotta casually remarks, ‘politicians have become a more self-referential group’?Footnote ⁷ Can one still be so confident that the nexus of parties and elections in the future will be as self-correcting as it has in the past?

F. R. Ankersmit's Political Representation offers by far the most intriguing hypothesis about the origins of the present crisis of representation – at least, if one is willing to take a chance on a ‘meta-functionalist’ theory that reaches back over 1,000 years! According to Ankersmit, changes at this level are generated by ‘a political problem whose extreme urgency no one could sensibly doubt and that could not possibly be made to fit the existing political machinery’.Footnote ⁸ This is about as good a generic definition of crisis as one can find (even if the word, ‘crisis’, is not in the book's index). According to him, feudalism arose out of the disruption caused by the barbaric invasions, and absolutist monarchy was the regime that ‘solved’ the schism between Catholics and Protestants. What was the kind of problem that produced representative democracy in continental Europe in the nineteenth century? You guessed it: the permanent fear of civil war due to the polarized ideologies left in the aftermath of the French Revolution and Napoleonic Wars.Footnote ⁹ Only some constitutionalized type of regime that allowed both sides to participate in the political process, thereby limiting persistent domination by one over the other and encouraging both to find compromises in the juste-milieu, was capable of preventing endemic violence. Political parties emerged to express this division between revolutionaries and defenders of the ancien régime and they converted a life-and-death struggle into electoral and parliamentary competition under the auspices of what he calls ‘principled unprincipledness’.Footnote ¹⁰ Subsequently in the same century, when this cleavage waned, the conflict between capital and labour took over and anchored the process of representation firmly along a left–right continuum.

Seen from this ‘meta-functionalist’ perspective, the roots of the present crisis become obvious – at least, hypothetically. The polarizing and threatening nature of class conflict has declined and not been replaced by another one of similar configuration or magnitude. Many citizens have acquired ambiguous class positions that confound their choice of party, weaken their ‘aesthetic’ identification with representatives and diminish their interest in political participation. Deprived of an overriding cleavage, they face a multitude of minor ones – many of which, Ankersmit notes, are the result of the unintended consequences of past public policies rather than being anchored in deeply entrenched social, cultural or ideological antagonisms. Citizens become more divided within themselves than against each other, especially between exploiting short-term advantages and suffering long-term effects. Political parties would seem utterly incapable of processing such cleavages and, therefore, should be doomed to functional extinction. That they have not already done so can only be attributed to their ‘path dependence’ upon past symbolic capital and ‘legal protection’ from competitors. Not surprisingly, Ankersmit regretfully observes that ‘most of the dramatis personae of representative democracy – the political party, the ideological opinion-maker either inside or outside the political party, the party's representatives in the legislative body, the legislative power itself, the executive power – are all on their way out’.Footnote ¹¹

At this point, what I anticipated was for the author to turn to (and to extol) the shift to ‘civil society’ where citizens can presumably find a wide diversity of interest associations and social movements catering to represent their fragmented identities, exploit them to set policy agendas and forge shifting alliances to control the subsequent behaviour of their leaders – in other words, a shift towards some type of ‘associative’ democracy. Instead, he accords no explicit importance to civil society and foresees that a new form of ‘plebiscitary’ democracy will emerge, in which the role of citizens will be confined to passively approving or rejecting the performance of state bureaucracies and experts in elections every four to five years. The only hope, he argues, lies in reviving political parties – something that is ‘eminently practicable, but even an absolute must for the survival of (representative) democracy’.Footnote ¹² Unfortunately for the reader, his only suggestion about how to do this is an aphoristic reference to ‘a reideologization of party programs’ without any practical advice about how to make these units of representation ‘take up a clear and widely recognized position in the dilemmas with which the citizen feels confronted’.Footnote ¹³

Joan DeBardeleben and Achim Hurrelmann have edited a collection of essays on Democratic Dilemmas of Multilevel Governance that directs attention to a quite different dimension of representation (and that sandwiches it between ‘legitimacy’ and ‘accountability’ in the subtitle). What happens to it when the European Union comes along and superimposes an additional layer of bargaining and decision-making upon the usual national ones? There are almost always several levels of government within national REDs, but these are normally embedded within an over-arching and pre-established framework of central authority – even in the most federal of systems. The EU has been constituted by a series of treaties among member states that leaves ambiguous (or co-determined) the location of sovereignty in an expanding number of policy arenas. If it were just another international organization (as some have argued), the issue of representation would be irrelevant since it should suffice for citizens to focus their collective efforts upon influencing national governments that are supposed to be able to veto or to defect rather than accept any new obligation that is not acceptable to its citizens. All of the authors in this volume seem to agree that the EU has succeeded in at least sharing sovereignty with its member states in many key arenas and, hence, deserves to be a distinctive target for directing representation – not to mention, for calculating legitimacy and demanding accountability.

The editors identify three dilemmas intrinsic to multilevel government (MLG). All of them have major implications for the practice of representation. The first involves the lack of congruence between the demand for authoritative political action and the sociocultural identities of those affected. In other words, Europeans want the EU to act on a variety of issues that can no longer be handled effectively by national states, but they continue to identify, communicate and sympathize predominantly with their co-nationals. The previous books presumed a degree of congruence between culture, economy and the demos that had historically resolved this dilemma. The EU has yet to accomplish this (and may never do so), which makes it much more difficult – both normatively and empirically – for representatives to connect these domains in a way that can generate legitimacy and accountability. The second dilemma refers precisely to this last property. The EU in order to solve common problems effectively has been endowed by its member states with a decision-making apparatus of formidable complexity, but the actors involved do not form a supranational government that can be directed or sanctioned by European citizens. They can vote in parliamentary elections at that level, but the deputies they elect have only very limited ‘co-decision’ powers over specific decisions and no capacity to sanction the decision makers with dismissal. This system of MLG has lots of checks and balances built into it, but none of them is unambiguously responsive to voter preferences – either in the formation of the executive or its subsequent performance. Instead, the EU has attracted (and sometimes subsidized) an extensive system of access for a self-selected group of non-elected representatives and this leads the editors to their third dilemma. As impressive as the EU has been in incorporating these European associations and movements into its processes of negotiation, deliberation and decision, this has introduced a major new source of distortion since the capacity for collective action at this level is much less equally distributed than in the national one. ‘Privileged groups’ of ‘stakeholders’ that are small in number, concentrated in space, and better endowed with material resources and expertise are much more likely to be well represented than those with numerous, dispersed and less endowed ones. The Commission of the EU has made efforts to compensate for this, but this has done little to diminish the perception by citizens of unequal access to a very remote, complex and obscure decision-making process.

The 11 chapters bring together a wide range of topics, including two dealing with the reaction of new member states from Central Eastern Europe to their participation in MLG. For the student of representation, the most pertinent will be the ones in Part III that focus more specifically on the third dilemma: Lawrence LeDuc on the problematic impact of holding direct elections for the European Parliament; Tapio Raunio on the difficulties facing national parliaments in learning how to play the politics of MLG; and Justin Greenwood on the efforts of the Commission to reach out beyond privileged business and professional groups to the more encompassing units of ‘organized civil society’.

Terry MacDonald, in Global Stakeholder Democracy, directs the student of representation to a yet higher level of aggregation: ‘the global’, and introduces a concept that has already become pervasive in political jargon: ‘the stakeholder’. The dilemmas identified by DeBardeleben and Hurrelmann were formidable enough at the EU level; they become literally unsurpassable at the global level. There are simply too many persons potentially eligible to participate in the choosing of representatives and, worse yet, these potential citizens are grouped into political units of very different size and capability. Unless…the theorist simply abandons the founding principle of democracy, namely, citizen equality with regard to making binding decisions, and replaces it with the new principle of stakeholder liability or as MacDonald puts it: ‘individual(s)…should be identified as the legitimate subjects of democratic control in accordance with the impact of this public power upon their lives’.Footnote ¹⁴ Forget about the notion that all adults have an equal right and inherent capacity to participate in politics – even if they usually have to delegate this to someone else – and restrict access only to those who are ‘constrained in their autonomy’ by the impact of whatever is being decided.

Conveniently enough, once you have performed this radical surgery on democratic theory, you will find a large number of self-certified groups who claim to be doing exactly this at the global level – non-governmental organizations (NGOs). With no plausible claim to be elected as representatives by citizens, they are presumably ‘selected’ by those most directly affected and concerned, usually through the mechanism of voluntary financial contributions by individuals.Footnote ¹⁵ In extreme instances, when the presumed beneficiaries are incapable of communicating (they may not even be human beings!) or too impotent to provide resources, the NGO may be uniquely capable of representing their interests and protecting them from suffering the adverse impact of public authority.

Needless to say, the plausibility of the entire argument rests on a convincing mechanism for authorizing and eventually holding bona fide stakeholders accountable. Anyone can make such a claim to representation (vide Bono and the protection of African interests, or Greenpeace and the survival of whales), but what makes such claims legitimate and, hence, likely to engender voluntary compliance with the decisions they produce? MacDonald's treatment of this is painful: well intentioned but manifestly insufficient to the task. He begins by assigning the concept of ‘communities’ to different stakeholder groups – an obviously distorted assumption. His stated intent is somehow to narrow the notion of stakeholders to only those that are ‘democratic’, but offers no operational criteria for doing so. After some exploration of examples – e.g. is the stakeholder community for an intervention by a given (and presumed) humanitarian NGO composed only of the population it has chosen to help or should that community not also include other populations that were deprived of such benevolent assistance? – he absconds and admits that ‘the stakeholder criterion will need to be interpreted [by whom?] and operationalized [by the same whom?] within specific political contexts…it is impossible to say something more about how this contextualism would play out…the delineation of stakeholder communities [is] a far from straightforward process.’Footnote ¹⁶

If the discerning reader has already learned that identifying whether a crisis of legitimate representation exists now at the national level is difficult and may be even more problematic at the EU level, just try to imagine how impossible it is going to be in the future to make reasonable normative judgements when one has shifted from citizens to stakeholders and from established polities to the emerging ‘democratic global order’ that MacDonald is so enthusiastically advocating.

Robert E. Goodin's Innovating Democracy is less useful for the purposes of this article. He has nothing to say about any imagined crisis of representation in ‘real-existing’ democracies. The term ‘crisis’ does not even figure in this book's index either. The reason for this can be deduced from the subtitle of his book: Democratic Theory and Practice after the Deliberative Turn. Goodin is more interested in exploring (and extolling) the benefits of deliberation than in resolving any crisis in the practice of REDs. He focuses on the intrinsically democratic benefits of ‘making use of mini-publics’ and consciously avoids any reference to potential innovations in institutions that have what he calls ‘constitutional power or statutory authority to determine outcomes’.Footnote ¹⁷ Subsequently, he expands his purview to include mechanisms of accountability wielded by discursive deliberation on a larger scale prior to decision-making and sequenced intervention on a more specialized scale by complex networks of civil society organizations. Throughout, he seems to assume that introducing such innovations will supplement rather than displace or transform the more classic institutions of representative democracy.

The interesting exception comes when Goodin takes the highly innovative step of imagining what might happen to REDs if they became ‘non-party democracies’. Unfortunately, he does not specify how this might come about, since the process of abolition or demise would surely have some impact upon the subsequent outcome. Nevertheless, the thought experiment is a bold one. Just suppose each candidate nominated him or herself, campaigned alone and was elected to parliament with no prior commitment to coordinate his or her behaviour with any other deputy. According to Goodin, all of the effects would be negative: each representative would promote a ‘personality cult’; each would surround him or herself with a distinctive clientele; each would engage only in constituency service by providing patronage; each would focus on administration rather than politics or the public interest; each would appeal exclusively to identity groups rather than substantive interest groups; and, finally, ‘what [would be] lacking is the “politics of ideas”, practiced in any systematic way.’Footnote ¹⁸

Ironically, this list of negative features is almost an exact replica of the critiques that are made of existing political parties! Far from being what Goodin calls ‘ideational facts’ embodying a principled conception of the interest of the system as a whole and, therefore, capable of providing a convincing ratio for the measures they take, contemporary parties form a crucial part of a ‘self-referential’ political class that colludes to protect its privileged interests, while opportunistically benefiting a surrounding set of favoured patrons and clients. What is surprising is that Goodin, a self-professed proponent of deliberative politics, denies the possibility that non-party elected representatives would make use of such devices within the legislature in order to form governments, control executive autonomy, discuss proposals of law, vote in shifting majorities and presumably be rewarded with re-election by their constituents for having acted in such an independent and responsible fashion. Is there some good reason to expect that candidates elected without the benefit of party labels or programmes will be much less able to deliberate subsequently than their partisan precursors?

Andrew Rehfeld's The Concept of Constituency makes no explicit contribution to our knowledge of whether there exists a crisis of representation and, if so, what is causing it. Again, there is no item for crisis in his index, although the author does refer frequently to the issue of democratic legitimacy and its dependence on the legitimacy of political representation. He neither argues nor presents any data demonstrating that the legitimacy of the latter as currently practised is in decline and, therefore, in urgent need of reform.

Rehfeld's implicit contribution, however, could be the most important of all of the books being reviewed. He raises in a systematic, critical and provocative way a dimension of representation that all of the others have ignored or, better, taken for granted – namely, the subunit of the polity within which representation takes place.Footnote ¹⁹ It is not the individual who is being re-presented; it is that individual (or family, firm, social group, etc.) within a pre-established constituency. Historically, this first involved professional categories (guilds) or status groups (Stände) in autonomous medieval cities and only later did territory become the defining criterion. The territorial base of representation has become so habitual that it is almost neverquestioned, despite the fact that the ‘communities’ supposedly involved have changed radically in their stability, size and composition.

What if the ‘real’ problem underlying political representation today has little or nothing to do with parties and elections, but lies in the type of constituency in which this process is embedded? Unless citizens are choosing within collective units that are meaningful to them, why should the winning representatives be regarded as legitimate – no matter how honestly the votes are counted, how fair the contest or how significant the alternatives? Territory may have seemed the ‘natural’ and logistically efficient solution in the past, but why continue to rely so exclusively upon it in the present?

Rehfeld proposes an original alternative. He would have citizens assigned randomly to 435 non-spatial single-member constituencies for elections to the House of Representatives upon their registering to vote and then competition would take place between nominees of political parties and selection would be by simple plurality.Footnote ²⁰ In concluding, he engages in a thought experiment about what might happen in the (highly unlikely) event that his proposal was accepted. Needless to say, he discovers many virtues – not the least of which would be a revival of political parties.Footnote ²¹ I personally find virtually all of his imaginative inferences implausible (and very Americo-centric), but I applaud his effort at provoking a discussion concerning this obscure dimension of representation in REDs.

I have been unfair to the authors of these works and I apologize to them for this. Not one of them set out to write a book about ‘the crisis of representation’– not even about ‘the crisis of political parties’. If you wish to explore these themes, I suggest reading one or all of the following: Jean-François Thuot, La fin de la representation et les formes contemporaines de la démocratie, Montreal, Éditions Nota Bene, 1998; Pippa Norris (ed.), Critical Citizens: Global Support for Democratic Governance, Oxford, Oxford University Press, 1999; Susan Pharr and Robert D. Putnam (eds), Disaffected Democracies: What's Troubling the Trilateral Countries?, Princeton, Princeton University Press, 2000; Pippa Norris, Democratic Phoenix: Reinventing Political Activism, New York, Cambridge University Press, 2002; Robert Putnam (ed.), Democracies in Flux, Oxford, Oxford University Press, 2002; Russell Dalton, Democratic Challenges, Democratic Choices: The Decline in Political Support in Advanced Industrial Democracies, Oxford, Oxford University Press, 2003; Archon Fung, Designing Democracy, Princeton, Princeton University Press, 2003).

I have, however, learned some things from reading and reviewing the six books above:

1. 1. Our authors (with the important exception of Ankersmit) do not bother to waste much effort in defining representation, relying implicitly on a ‘commonsensical’ version of what it seems to mean in contemporary REDs. This makes it all the more difficult for them to identify changes in its practice and shifts in its focus.

2. 2. Political parties are still considered to be the most salient and significant institutions of political representation, despite evidence of growing popular distrust of them and increased reliance on other channels of political expression, e.g. interest associations and social movements.

3. 3. Competitive elections are still considered the most salient and significant means for ‘certifying’ the legitimacy of representatives and the claims they make on behalf of the represented, despite evidence of the increased role played by self-appointed and/or organizationally selected ‘spokespersons’.

4. 4. Territorial constituencies are still considered the most appropriate and reliable political units within which interests and passions should be aggregated, despite evidence that these units have changed considerably due to greater mobility and that citizens identify strongly with functional or ideational constituencies.

5. 5. The ‘deliberative turn’ in normative democratic theory has failed to come to terms with the likelihood that political deliberation in REDs is much more likely to involve communication between formal organizations rather than between individual citizens and that these two are not equivalent.

6. 6. Scholars working on representation typically choose to focus exclusively on one level of political aggregation – usually the national state – and are, hence, insensitive to the fact that contemporary REDs (especially in Europe) are embedded in multiple levels of decision-making with overlapping, cross-cutting and sometimes cumulative claims by representatives.

7. 7. The authors of these volumes do not seem to be concerned about an alleged ‘crisis of representation’, either because their data or speculations do not support such a strong and alarming inference or because, even if they are aware of some negative trends in contemporary REDs, they seem to be confident that normal and well-entrenched processes of adversarial social relations, partisan competition, regular and honest elections and anticipated reactions by rational politicians will automatically correct for such momentary disruptions and serve to bring the expectations of representatives and represented back into line with each other.