Interview Text Interviews

Dossier / Democracy: Bridging the Representation Gap

Is representative democracy really democratic ?
Interview of Bernard Manin and Nadia Urbinati - New York, April 10, 2007

by Hélène Landemore , 31 March 2008

Hélène Landemore : Bernard Manin and Nadia Urbinati, you both have written books with apparently similar titles, respectively The principles of representative government (1997) and Representative Democracy: Principles and Genealogy (2006). I would like to organize the discussion around the question of whether representative democracy is an contradiction in terms or the true essence of democracy. I will break down the theme into more manageable questions.

Both authors had the opportunity of revising their answers. The latest changes were made after the publication of the French translation.

Hélène Landemore :

Representation need not be democratic, nor is democracy necessarily representative. What brought those two things together historically? And when does the concept of “representative democracy” first appear?

Nadia Urbinati : According to Pierre Rosanvallon, the expression appeared in a letter by Alexander Hamilton to Gouverneur Morris in 1777. It was then used more systematically in the early 1790s, especially by Paine, Condorcet, and Sieyes. In his Bases de l’ordre social, Sieyes made an interesting distinction between two interpretations of representative government, only one of which democratic, although both of them based on elections and thus applicable in large and populous territories. But the former consisted in facilitating “partial meetings in the various localities” whereas the latter only in “nominating deputies for a central assembly.” Hence according to Sieyes, the first could not result in “one general will” because it gave voice to the citizens living in the localities; it was similar to the model proposed by Condorcet. What interests us here is that Sieyes understood very well that there was a distinction between different forms of representative government.

H. L.: Bernard, would you say that the difference is that representative democracy is democratic whereas representative government is aristocratic?

Bernard Manin : No. That is definitely not my claim. Representation does include democratic components, most importantly the opportunity for all citizens to hold representatives to account at the end of the term and to dismiss them if their performance in office is judged unsatisfactory. These democratic elements are real and consequential. The point is that they are not the sole components of representation. Representation is also a government by elites that are not strictly bound to carry out the wishes of their constituents. Thus, representative government combines democratic and non-democratic components. This is why I characterize it as a “mixed” form of government, drawing on the Ancient notion of the mixed constitution that goes back to Aristotle and Polybius. Just characterizing modern representative democracies as systems in which the people are “sovereign”, or “indirectly” govern themselves, obscures the mixed and composite character of such systems. Representative government has never been a simple form of government. Furthermore, over the last decades institutions that were not part of the original arrangement appear to have taken root in a number of representative democracies, such as constitutional courts reviewing legislation and non-elected, “independent” agencies. With the rise of such institutions the mixed character of our democracies has become even more salient.

H. L. : When you say a simple form of government, are you saying that the Ancients did not know any form of representation?

Bernard Manin : Yes, I would say so. I don’t think that the Athenian Council [Boulè] should be viewed as a representative body. While sources identify the Assembly and “the people of Athens”, they do not identify the Boulè and the demos, thus underlining that the Council was not perceived as standing for the people. The Boulè was just a collegial magistracy.

Nadia Urbinati : I agree. Political representation operates in (and should be referred to) the place in which laws are made. In this sense, the 18th century scholars and political leaders recognized that the Moderns had introduced something that the Ancients didn’t know. Perhaps the English constitutional revolution of the 17th century was an important step forward in the construction of representative government. The passage from selection to election, or the method of an open competition for legislative positions, was a crucial turning point in the construction of political representation. Representative government requires to be connected to elections and to pertain to the legislative power. These two elements together bring us to say that representative government is the government of the moderns, not the government of the ancients.

H. L. : When does the concept of representation emerge?

Nadia Urbinati : According to historians of political institutions and ideas the story of representation starts in the Middle Ages, inside of the Church. In that case too the question was to solve the problem of connecting the center and the periphery. The Church sought to represent the community of the entire Christendom and representation was then used as a way of unifying the Christians or connecting the large body of believers. In the Middles Ages, the inscription of the rule of the contract in the public law was advanced. Both religious and secular communities accepted that the decision over the appointment of power was regulated by public law: this appointment entailed that every power of a political kind should bear ‘representation’ of the whole community, as Otto Gierke wrote. Yet Scipione Maffei in a 1736 comparative and historical study on the republican forms of government wrote that the Romans practiced representation in order to give voice to the many nations composing the Empire; he referred to Tacitus who in his Germania described the forms of representation and parliamentary institutions used by the German tribes to voice their claims to the Roman Senate. Representation was there a way of linking the large territory of the Republic by a kind of federative system.

Bernard Manin : The origins of representation must undoubtedly be found in the Middle-Ages, in the context of the Church and in the context of cities in their relation to the king or the emperor. The idea was to send out delegates having power to bind those who sent them. There lies the origin of representation. A given community delegated some members with powers to bind those who appointed them. That’s the kernel of the notion of representation. Then the technique got transferred to other contexts and used for other purposes.

Nadia Urbinati : There were also private practices and institutions, like the advocates or the lawyers.

H. L. : What is the role of Hobbes in this story?

Nadia Urbinati : Hobbes used the strategy of representation in an importantly new way, that is to say in order to create the sovereign state. Representation was in his system a way for giving legitimacy to the absolute sovereign while disempowering the people, who were only subjects. It was an interesting way of giving legitimacy by taking away power from the people; representation as a fiction to create the absolute sovereign.

Bernard Manin : Hobbes may articulate with particular cogency the idea of a sovereign authority that acts in the place, and on behalf of the subjects. However, the fact that Hobbes’s theory is particularly striking to us is not evidence that it had a major impact on actual historical developments. As we just noted, representative institutions and techniques vastly antedated Hobbes. Note also that Hobbes does mention at all elections as the method for appointing the sovereign authority. In the case of representation, to be sure, Sieyes read and employed Hobbes to articulate and justify some of his views about government. But I don’t think that any such recourse to Hobbes can be found amongst the American Founding Fathers. Tracing Hobbesian ideas amongst American revolutionists and constitution makers sounds complicated, at the very least.

Nadia Urbinati : Quentin Skinner recognizes rightly the role of Hobbes in creating the representative system in an anti-republican function. But one may even say that Hobbes did not use representation as a political institution and, in this sense, didn’t construct representative government (although employed representation within the context of the state). Indeed he used representation not in order to create a government that was representative of people’s opinions or accountable to them. Perhaps we should disassociate representation from that tradition, which was a way of getting the sovereign state an absolute power, not creating a government that enjoys people’s consent and is authorized by the electors. Electoral representation is a break with absolutism while it is certainly an open door to the democratic transformation of the government. The 18th century is thus more interesting if we want to see the different avenues that the idea of representative government took. I think the American case is very interesting, also because the founders organized representation in the making of their republic rather than in theorizing.

H. L. : You both describe a different set of principles for representative government and representative democracy. What are they and why do you differ?

Bernard Manin : My own focus is on concrete institutional arrangements. I call them principles because they have been stable over time. But by principles I do not mean abstract propositions, much less ideals and values. My approach is positive and analytical. Such a perspective, I grant, entails some limitations. I adopted it for the sake of manageability. I identify four such institutional arrangements that have remained unchanged since the establishment of representative systems. 1/ Those who govern are appointed by election at regular intervals. It is not just the fact public officials are selected by election that characterizes representative government, but the fact that such elections are recurring. In his famous definition of democracy Schumpeter fails to mention the recurring character of electoral contests. Repeated elections entail critical implications, however. In their actions while in office those who govern have an incentive to anticipate the retrospective judgment of voters at the end of the term. Thus, elections do not only select leaders, they also affect the actions and policies of those in power. At the end of their term public officials are held to account. In representative government elites govern, but at the same time these elites are accountable to ordinary citizens. It is worth noting that in his definition of democracy Schumpeter makes no mention whatsoever of political accountability. We see here with particular clarity the combination of democratic and non-democratic components. 2/ Those in power enjoy some measure of independence in the policy decisions that they make while in office. They are not strictly bound by the wishes of their constituents and by the platforms presented to voters. Note that this arrangement leaves room for some influence of voter wishes over the actions of elected officials. It only provides that exact congruence between the two is not mandatory. 3/ The third principle is what I term “freedom of public opinion”. While representatives have a certain measure of discretion in their action, the people or any segment of the populace retain for their part the right to voice their opinions and grievances and to press their claims upon those in office at any time. Even Burke, one of the most fervent opponents of binding instructions to representatives, insisted, in his Third Letter on a Regicide Peace [1796-1797], that the people retain the right to manifest their views and wishes at any time “without absolute authority, yet with weight”. A similar idea can be found in the last clause of the First Amendment to the American constitution. This clause consecrates “the right of the people peaceably to assemble, and to petition the government a redress of grievances.” Representative government has never been a system in which the people elect their representatives at regular intervals, while remaining quiet in the interim. This is yet another point missed by Schumpeter and his followers [1]. 4/ The last feature is that public decisions are subject to trial by discussion. Saying that public decisions are “subject to trial by discussion” does not amount, I would insist, to characterizing representative government as government by discussion. Discussion is not a procedure for making decision. It is a method for trying, scrutinizing, and testing public decisions. Those are the four principles.

Nadia Urbinati : To the four principles described by Manin, I would add something else. I think that democracy (or better saying the democratic transformation of representative institutions through universal consent by voting) introduces something interesting. By democracy I mean here the universal franchise, including adult men and women, and also the specialization and pluralization of civil society (what we call today democratic society). Democracy in this broad sense introduces two crucial elements that mark political representation: one is that of advocacy (it has to do with the third of Bernard’s four points in some sense), the other is that of representativity. In other words, representation needs to have a correlation with civil society through forms of political associations, that is to say aggregative forms that are able to express, control, claim, survey and set a current of relationship between the inside and the outside of states institutions. Of course this current consists in an informal politics, one made of influence and public judgment more than authoritative will; but it is very important in order to capture the peculiar character of political representation in democratic society. Representation is not just having the people vote for individual candidates. It’s much more interesting giving them a voice in the intermediary time between elections. The parties and associations make possible the performance of this function. This is what advocacy does.

The other element is the representativity of representation. Representation is not a substitution for, but a way to identify with. When I go to vote, I really do two things: I select somebody to seat in the assembly but I also want somebody who is close to my ideas or represents them as much as possible. I don’t choose a competent bureaucrat or an expert, because the job of the lawmaker is not like that of a bureaucrat or a magistrate (who are not supposed to express their ideas when do their job or even act according to their own ideas). I choose someone close to my own views because I have some views on how to make a better law or change a law, and want my views to be heard. The job of a representative is exquisitely political. Representativity is a sort of vicinity in ideas and ideology between the candidates (and then elected) and the electors. It is also important for what it does inside of the assembly, where lawmakers have to operate as members of the deliberative setting while being in contact with the extra-parliament. Without the connection between the elected and the citizens, pluralism inside of the assembly would simply reflect the personal view of the lawmakers with no correlation to civil society: the representatives would represent themselves. An assembly made in such a way would resemble that of a direct democracy (in which each citizen represents only herself and her own ideas when voting on a given issue). But representation is not direct democracy. The construction of parties and associations is important, I would say essential to representative government,. Indeed a representative Assembly is not a gathering of individual delegates, but a collective body of representatives, that is to say individuals caught in ideological separations/alliances. In this sense, political representation is a radical violation of private or juridical forms of representation. The representative is not just there for himself or for me as a private person, but for me as an equal part of the demos or citizens-elector. Political representation is actually a violation of private or juridical representation also because it excludes imperative mandate: I can’t remove the representative as I wish even when she says or does things I personally disapprove. Hence thanks to political parties, the single representatives are related both to the citizens and to the general interest. Parties are not like factions, to use an expression of Machiavelli; they are a way to connect the particular and the general interest, whereas factions are about taking over the general interest and putting the private interest in its place.

H. L. : Would you then subscribe to the idea that representation is not the second best of direct democracy?

Bernard Manin : Exactly. On this point, Nadia and I are in complete agreement. Representative democracy is not the second best of direct democracy. It is a different system. In my view, democracy, whether direct or indirect, is a simple form of government, while representative democracy is a mixed form involving a variety of elements.

H. L. : Bernard Manin shows in his book a process of democratization of representative government, in the move from parliamentary democracy to party democracy to today’s audience democracy. But in the end representative government, even democratized, is still a partly elitist regime. It is a mixed regime. For you Nadia, the representative model of democracy does not involve an elitist element. In that sense it can be opposed to the electoral model of democracy, which presents this elitist dimension. Is that correct?

Nadia Urbinati (laughing) : Bernard is more of an elitist than I am.

Bernard Manin : On my account elites play indeed a critical role in representative government. This is so because elections necessarily select people possessing uncommon characteristics that are positively valued by voters. A candidate that did not stand out by some favorably judged trait would not win in an electoral contest. That said, the elective method does not determine the particular content of the distinctive and positively valued characteristics that get candidates elected. Such characteristics are determined by the preferences of voters, that is, of ordinary citizens. Voters choose the distinctive qualities that they want to see in their representatives. These qualities may consist in a variety of things, including uncommon ability to articulate and promote a given political opinion. Even in that case we are still dealing with elites in that people who are exceptionally able to defend an opinion possess a talent that most people sharing that opinion do not possess. This is what I mean by elites.

I don’t think, however, that this argument amounts to taking an elitist position. Elitism as a normative principle holds that it is desirable that people who are objectively superior to others should occupy superior positions. My account of representation does not imply any such view. First, I am not arguing that elections select people that are objectively superior to their constituents. The claim is only that elections select candidates endowed with characteristics that are subjectively valued by their constituents, –rightly or wrongly. Second, I am not offering an argument as to whether it is desirable that positions of power should go to people possessing distinctive and favorably viewed traits. I mainly establish that such an outcome is a necessary feature of representative systems. I do argue, it is true, that these systems are consistent with the normative principle that political power should come from the free consent of those over whom it is exercised. That is the case as long as voters have the effective possibility of choosing the distinctive traits of their representatives. But I don’t go further than this circumscribed claim. A more ambitious normative perspective would require a lengthy and intricate argument given the tight mixture of egalitarian and non-egalitarian dimensions. I decided that such an argument was beyond the scope of my project and abilities. Thus, in essence my argument about elites is positive, not normative. One may recognize the actual importance of elites while not endorsing elitism as a value.

H. L. : But then Nadia, assuming that Bernard is right that representative government is partly elitist today, and assuming with you that normatively speaking it shouldn’t be, have we ever experienced what true representative democracy is?

Nadia Urbinati : Not exactly. When you read Bernard’s book’s last chapter, you are told that we cannot talk of a crisis of representation because representation was since its inception instituted in order to contain rather than to implement democracy. How could we demand our governments to act in a way (democratic) they were not designed for? In this sense it is futile to speak of a “crisis of representation.” Yet there are times in which we do feel that there is a disconnection between us and our representatives –although this is simply an opinion and destined to remain pretty much so. Is this tension part of what representative government means? It is a fact that there are moments in our society in which we think or write that there is a disconnection. Why is this so? Even if it can’t be measured or quantified, there is a sense of disconnect or lack of representativity. What interests me is the democraticness of representation. If it is true that representative democracy has to do with the judgment of the people rather than the will of the people, in order for a government to be fully democratic, we need to have more than simply regular electoral systems or party systems. We need for sure to take care of the information system (because judgment is a crucial component of the kind of presence that the people exercise in an indirect or representative government) and of the very problematic presence of private money in electoral campaign. Information is very important in a system in which indirectness and mediation are crucial, where we receive all inputs in the form of elaborated or digested information and nothing is first hand or face to face. We don’t have any instrument to make a competent judgment independently of the media. So it is true that the question of private money in the campaigns, media independence and pluralism, are genuine problems because they can jeopardize the possibility of equal participation in the system. Hence there are reforms to be done on these issues if we are to be more in tune with a representative system.

Representative democracy is certainly no less democratic than direct democracy. Paine was right when he said that representative democracy surpasses direct democracy. In direct democracy every citizen is here for himself or herself and there is no need to create a connection between citizens and the institutions. But in a representative government the Parliament and the institutions are always connected in a mediated way to the people outside. The second thing that representation does is that of giving democracy’s stability. Paine used to say that representative democracy is superior in that respect too. In direct democracy assemblies are the place of a direct confrontation of the individual citizens and they can easily give rise to harsh and radical conflicts or to situations in which only the majority rules with no constraint or one in which factions may easy rule. Mediation is a good remedy. But it requires to be regulated.

H. L. : Bernard, Nadia says that you deny that there is a crisis of representation. On the other hand, you talk about the different crises of representation as stages in the metamorphoses of representation. So is the current “crisis of representation” just an illusion of perspective? Just one more metamorphosis?

Bernard Manin : For the notion of crisis to be analytically helpful, we should use it only under specific conditions. It must be the case in particular that well established changes and trends seem for some reason inconsistent with the constitutive characters of the object under consideration, potentially threatening its survival. If changes do not reach that level, diagnoses of “crisis” are little more than weakly informative clichés. Any change in a given domain, and in particular any evolution that is still little studied and poorly understood, then becomes the occasion to declare a crisis. Two factors, in addition, favor the proliferation of diagnoses of crisis: on the one hand, the propensity to idealize the past, and on the other hand, the fact that the announcement of a crisis in any kind of activity is more likely to attract the attention of publishers and readers, even academic ones, than analyses of this activity in ordinary times.

I don’t think that the conditions required to justify the diagnostic of a crisis of the representative system are met. Here, the two most serious indicators of a possible crisis are first the decline in the rate of electoral participation and second the relative disrepute affecting political officials. This is not the place to enter a technical discussion of these indicators. One can observe, however, that their analysis is still an object of debate amongst specialists. More importantly, it is unclear why the developments expressed by these indicators should be viewed as inconsistent with the principles of representative government.

In his thorough study of established democracies, Mark Franklin shows, for example, that electoral turn out has declined since 1945 [2]. He stresses, however, that this decline is limited and that one could as well wonder why participation has changed so little. His own theory attempts to answer the two questions. But Franklin mostly shows that this decline constitutes only an average over elections held in each of the countries. The central phenomenon is that participation fluctuates according to the particular character of each election. The number of voters rises, sometimes massively, when elections are perceived as important, involving high stakes, and close-fought. The recent French elections spectacularly illustrate this fact, but the phenomenon of fluctuation can be observed in all democracies. There is no reason why the fact that people vote mostly when the stakes of an election seem to be high and when the outcome is expected to be tight should held as inconsistent with the functioning of representative government.

A number of studies also show that citizens who declare a low level of trust in political officials do not withdraw in political apathy and disinterest. They are on the contrary more likely than average to engage in diverse forms of political participation, non electoral as well as electoral [3].

The viability of representative government would be threatened if citizens systematically stopped taking an interest in it and stopped taking part in the different forms of political action that it offers them. This does not seem to be the case. The picture that emerges is more that of a change in the rhythm and the terms of political involvement. Nothing indicates that the system is not able to adapt to such changes. Representative institutions have already demonstrated their adaptability. That 18th century arrangements survived the social dislocation caused by the industrial revolution, even serving the pacification of class conflict and the political integration of the working class offers the most impressive evidence of such adaptability. This ability to adapt is not just an empirical fact. One can identify its causes. On the one hand, as already said, this system is mixed, constituted of several elements between which relationships are not rigorously determined. Thus, the system grants the authority to decide to elected officials only. But it also guarantees the freedom to express opinions and grievances at any time. The weight that elected officials must give to these expressions is not rigidly fixed. The fact the relationship between citizen wishes and decisions by representatives is not fully specified makes adjustments possible. On the other hand, the system makes visible the dissatisfactions that it engenders. Electoral competition gives incentives to remedy these dissatisfactions. Thus, freedom of information and expression of opinions makes us aware of the relative disrepute in which politicians may fall. Those politicians are aware of it too. The prospect of facing new competitors that may not suffer from the same disrepute leads to the search for antidotes. Representative government thus contains self-regulative and even self-transformative mechanisms. Given the capacity for transformation of representative government, we should be quite demanding in terms of the criteria that allow us to speak of its “crisis.”

H. L. : How about riots?

Bernard Manin : Riots in the projects around Paris and large cities are of course the sign of a failure. But why consider this failure as the sign of a crisis of representation? First, these riots are specific to France and do not happen in every representative democracy. Second, one must distinguish between a political system and the specific policies that it produces in this or that domain. France has certainly failed, so far, to integrate the inhabitants of « banlieues » into the common social, economic, and political life of the nation. It does not follow that the French system of government is flawed. Rather, the conclusion is that the policies followed in that domain were not the right ones. More generally, however, the eruption of public disorder is not necessarily the worst of evils. Much depends on what kind of answer is given to such disorders, and also on the severity of the events. Within certain limits, public disorders can also represent a powerful incentive to address recalcitrant problems or injustices. Machiavelli argued that dissensions between the Plebs and the Patricians, and the troubles that ensued, did not cause the ruin of Rome, but contributed in fact to the longevity and the balance of the Republic by constraining the Patricians to alleviate the Plebs’ grievances. His observation is still relevant.

Nadia Urbinati : We can talk about a short circuit of the system. The distance between the “legal” country and the “real” country – the sign of what we perceive as a lack of a sympathetic adhesion between representative institutions and citizens, as Pierre Rosanvallon wrote in La contre-démocratie—may be a way for the system to self-adjust, a kind of self-medication. In Machiavelli’s opinion, riots or, more pertinently popular upheavals, were taken as an impulse or a stimulus to political change. In today’s mature representative democracies, the lack of reform imagination can be more of a problem. What according to me is more of a problem is audience democracy. Bernard is right descriptively to diagnose a move from party democracy to an audience one. But should we be only descriptive when we diagnose this phenomenon? Audience democracy is not a violation of representative government? In my view this new form of Caesarism or populism is a violation of representative democracy. An acritical identification of the masses with a leader elected out of a campaign he manipulated through a media system is still a violation of the principles of representative democracy. The rise of audience democracy is truly problematic in my view, even more than the riots.

H. L.: Bernard, in answering Nadia’s question, can you perhaps touch on the figure of Ségolène Royal and her alleged populism (cf. her proposal creating citizens’ juries and her general defense of more participatory forms of democracy )?

Bernard Manin : Let me clarify something first. I am not arguing that political parties are obsolete or getting obsolete. My initial formulation was perhaps lacking in clarity. In any case, that was not what I had in mind. What I meant was that at the stage of “audience democracy” –which I thought of calling “democracy of the publics”– stable partisan allegiances and loyalties are declining. They have not disappeared, certainly, but they are being eroded. A growing number of voters no longer vote for political parties based on stable loyalties, determined by preexisting class and religious cleavages, regardless of context and circumstances. These less loyal voters rarely switch between opposing parties, sometimes they switch between members of a coalition, and most often oscillate between abstention and voting. This is not to say that parties have lost their importance. Parties are still essential in two domains. First, votes in the Parliament are still determined by partisan alignments; voting party discipline is still the rule in the parliamentary arena. Second, parties are still dominant in the electoral arena. Parties have adapted, no doubt, to the personalization of electoral choice (another trait of “audience democracy”). But this adaptation allowed them to retain a central role. It is parties that finance, prepare, and organize electoral campaigns. You were mentioning Ségolène Royal. She precisely illustrates the preeminent role of parties in the electoral competition as well as their adaptation to the personalization of power. To be sure, her personality was central to the electoral campaign. But she did not rise to the candidacy because she had independently reached celebrity status in the show business, the arts or the world of sports. She became a candidate because she was chosen by the socialist party as the result of a procedure chosen by the leadership. Of course, Ségolène Royal was not one of the main leaders of the party, but for over twenty years she made her entire career in and by the party. There are strong reasons to think that by choosing a woman who was not part of the circle of established leaders, the Socialists made a deliberate choice meant to express a desire for change. Political parties are by no means about to disappear. In all democracies, they have adapted both to the erosion of stable partisan loyalties and the personalization of power.

Nadia Urbinati : This is an important addition and emendation you make Bernard. Because in your book, you presented the party as one possible and transitory moment, a way of being of representative democracy in a specific historical time, the time of mass-parties; moreover, you presented audience democracy as the future of representative government, and in any case as a form of representative government. Actually, you even wrote that once “independent of individual partisan leanings” citizens would be more autonomous in their judgment because whatever their partisan opinions, they would “receive the same information on a given subject as everybody else.” On your benign evaluation of audience democracy I strongly disagree because I think this is a new form of populism that while does not make citizens more independent in their judgment it is less open to their check and their participation. Segolene Royal uses populist means along with other means, such as the party system. That is fine, also because she did not own or directly orchestrated the media. I agree with your diagnosis of a crisis of strong ideological partisanships. In any case issues that are translatable into values, ideas, are always there. But the distribution of party lines has changed, not however the ideological character of alliances. Without parties or part-like kind of aggregations, however, I can’t imagine any form of representative democracy. As I said, it would be a form of Caesarism or populism; but don’t call it representative democracy. The case of Italy is important in that regard. Mussolini was a patent violation of representative democracy; but Berlusconi too is a violation (although only the former caused an exit of the constitutional system altogether).

Bernard Manin : I entirely agree with Nadia to say that there is a violation of representation when a leader aims to embody by himself alone the entire community beyond its many divides, rather than a particular view of the community and the common good. Cesarism most certainly transgresses a fundamental norm of representative government, especially if the candidate to the transcending position actually succeeds in disqualifying his potential opponents and in preventing them from even entering the competition. But this is not what we observe in established democracies. The personalization of electoral campaigns does not mean the abolition of differences between existing options. Each proposed direction temporarily takes a personalized face, each party identifies with their leader of the moment, but the result is that several leaders are, in fact, confronting one another in the competition.

Nadia Urbinati : A liberal conservative theorist like Guizot had some good points when he argued that representative government designates a society in which the pluralistic interests of civil society should not directly enter politics but should operate in a mediated way. Italy is an eloquent example of a country in which this mediation is thinner and thinner and on some occasions almost absent. The chief of a private and powerful corporation (fully engaged in the media business) enters the political arena and “does” everything by himself. That is a violation of representation. Representation is a system of indirectness and mediation. Audience democracy can become a regime in which a corporate interest creates its own party, its own media to promote its own interests. The violation of the representative principle is when civil society enters the political sphere unmediated. It becomes factious. The direct transfer of the social (be it economic or religious or cultural) to the politics is a patent violation of representation.

H. L. : Some people advocate new ways of ensuring a better representativity of democratic assemblies, for example through random sampling and lotteries rather than elections. What do you make of these suggestions?

Nadia Urbinati : I’m somehow critical of these ideas, and particularly the idea of deliberative forums amongst randomly sampled citizens (Fishkin). Randomly selected bodies might represent certainly a novel contribution in so far as they might function as supplement to existing forms of representation. Yet should these forms grow, they will bring new challenges because any randomly selected deliberative body will inevitably generate opinions that differ from public opinion. These deliberative bodies might become tools that elites use to legitimate policies while bypassing electoral accountability, or they might substitute for broader citizen judgment and participation [4].

H. L. : What if the goal were simply to redress some objective imbalance, like the lack of women in the assembly? The idea is that random sampling would yield an assembly resembling more the real composition of the population than the current system.

Nadia Urbinati : Yes, but with Fishkin’s deliberative forum the issue seems to be not that of making representative democracy more representative, but of bypassing representation. They divide the population in young and old, rich and poor, workers, or any other category. In this way they give the impression of representing society like in a mirror. They presume also that population is subdivided into kinds or corporate bodies, and that the individuals of each group reason in the same way or have the same views. What is the goal? Is it to give more legitimacy to the decisions of the law-makers? Or to create an alternative?

As for the quotas, in theory, they are a violation of the most basic principle of democratic citizenship (equal liberty). But sometimes it might be necessary to violate principles to make right something that is deeply wrong (this was the case of affirmative action, for instance). A Parliament that is only a male Parliament is, seen from outside, so blatantly unrepresentative! Supose you have a Parliament that is 90% female : don’t you think that men would feel not truly represented ? It is true that male representatives may be representative of my ideas also; I can identify with them but still, since the citizenry is made up of both men and women... But quotas are instruments for protecting a minority (for instance national or linguistic) that would otherwise be absorbed within the majority. Yet women are not a minority to be protected. Actually instead of using quotas, I think it would be more appropriate to adopt a 50% politics: each list of candidates should be composed of an equal number of male and female candidates. This would allow for a true freedom of choice.

Bernard Manin : I am not hostile to institutional experiments. The constitutive flexibility of representative government allows for the introduction of complementary systems. As to the imbalance in the composition of representative assemblies, it is of course undesirable. However, it is not undesirable because each category of the population can only be well represented by members of this category. This would be the principle of mirror-representation, against which representative government was explicitly established, especially in the United-States. It is not for the framers of the constitution to determine by whom women or the descendants of immigrants should be represented. This choice belongs to these citizens themselves.

The problem, rather, comes from the fact that an important, systematic and enduring gap between the composition of the assembly and the composition of the population is a sign that some categories of population are very likely the victims of persistent handicaps in running for office and getting elected. How else could one explain the constantly minuscule proportion of women in the assemblies, if not by the existence of material and cultural obstacles encountered by women who desire to become candidates? There is here, as Nadia rightly says, a violation of fundamental political equality. The solution does not consist in establishing permanent quotas, but first in identifying the causes of these recalcitrant handicaps, and second in remedying them through appropriate social policies (such as making the pursuit of a political career easier for women desiring to have children). Finally, if these handicaps are found to stem from cultural prejudices, one can resort to incentives, or even to some form of legal obligation. But it is preferable that such policies remain temporary (until the prejudices disappear) because they are contrary, after all, to the principle of equal freedom in choosing one’s representatives.

A structural handicap that deserves a particular attention concerns not social-demographic categories, but those that are called the “challengers” in an election. In several democracies, the incumbents benefit from diverse advantages (in terms of resources particularly) when they run for another election. The incumbents’ advantages are of course the challengers’ handicaps. Such a configuration is doubly flawed. First, it goes against the renewal of the political personnel, feeding the image of a political class that is closed and self-reproduces come hell or high water.

Nadia Urbinati : In Italy too. They stay in power for twenty years and even more. The other thing is to avoid the accumulation of offices or functions. In Italy you have a person playing on different tables and institutions. This is a terrible problem of morality as well. On many occasions, those people are judges in their own causes. The accumulation of careers or tenures (no matter whether electives) is a problem. Concerning the issue of which system, the Parliamentary or the Presidential, is more in tune with representative democracy: I don’t know if you’d agree or not, Bernard, but it seems that the Parliamentary system is more flexible than the Presidential system, at least because of the possibility of anticipated elections, which is sometimes good.

H. L. : Taking Nadia’s normative concept of representative democracy as the standard, on a continuum from the least representative to the more representative democracy, where would the USA, France and Italy fall?

Nadia Urbinati : I can’t answer that. It’s difficult. But in Italy, we produced the best laws in the 1970s: the laws regulating divorce and abortion; the charters guaranteeing unionization and the possibility of implementing forms of industrial democracy; the new universal system of health care; the new family law stating relationships on an equal ground. In terms of civil and social legislation, the proportional system was able to produce better laws, more general and less partisan laws.

H. L. : It is interesting that you judge of this in terms of results rather than in terms of the representativeness of the procedures.

Nadia Urbinati : That’s because these results corresponded to a Parliament that reflected the quest of society for important changes. There was a connection between the Parliament and society. It was as if the Parliament was able to understand this. Now it is no longer the case.

Bernard Manin : Procedures are no doubt important. But in present day France, it is more important to address substantive problems, such as integration of the descendants of immigrants or labor market regulation for example, than to think about procedures for addressing such problems. Focusing on procedures and on democracy in general may unduly relegate to the background the discussion of more pressing and concrete questions.

Nadia Urbinati : I agree with that. If these issues are not dealt with it is perhaps because the social issues are peripheral in the agenda of the representatives. They are not enough pressing issue.

H. L. : How do you conceptualize deliberation in relation to parties?

Nadia Urbinati : According to me, deliberation is a dialectical exchange of ideas and opinions not however related to a cognitive conception of politics, as in the tradition of Rousseau and, under certain condition, of Habermas. To be consistent with this view of deliberation it would actually imply to erase representation or better saying question an assembly in which representatives construct alliances/oppositions, thus share some similar ideas with people outside or endorse a value-politics. Yet these partisan distinctions are needed because even if representatives are of course free to change their mind during their parliamentary tenure, it is highly desirable that they do not change it completely so as not to be recognized anymore by their electors. A cognitive conception of political deliberation can hardly allow for such commitments.

Bernard Manin : I agree. I was talking about eroding partisan loyalties, not about lesser attachment to values.

Nadia Urbinati : If you are open to a complete change, you dilute all forms of commitment.

Bernard Manin : Not necessarily. We may remain attached to values or principles, and yet reasonably change our opinion about the party, policies, or persons that, in one context or another, seem to us closest to our principles.

Nadia Urbinati : With this specification I agree.

H. L. : How do you connect representation and participation?

Nadia Urbinati : Participation, even in the extreme and exceptional form of riots, may sometime be a way to flag opinions otherwise unheard. Participation is not an alternative to representation. It is a way to check that representation is working.

Bernard Manin : Obviously, burning cars is not a good thing, but it is not civil war either. It would also be very undesirable if the circumstances associated with such behavior remained in closed enclaves, far from the public eye. In any case, democracy is not consensus.

Nadia Urbinati : The question is whether these riots create a form of closure, as a ‘law and order’ issue, or whether they stimulate a political debate about how to solve the problems they signal.

H. L. : One last question on the European Union. There is there a talk of a democratic deficit. When the public is so large, how visible can the European public be and isn’t it legitimate here to speak of a crisis of representation?

Nadia Urbinati : Europe, like America, is a very large geo-political space, although its political system enjoys less democratic representativity than the American one. The question is: How can we design institutions that are representative? There is room for institutional innovation. In Europe it’s now a middle way. It seems that if we want to have a European Union, we need new institutions: this means a Federation, restructuring the electoral system and creating European parties. Let’s take the American example.

Bernard Manin : As much as I am convinced that the reform of French institutions is not a priority, I am also convinced that the European Union is the object on which institutional ingenuity should be deployed. I am in perfect agreement with Nadia on this point.

Nadia Urbinati : So far Europe seems to be closer to a confederation. As a federation, it would be a new state. Until we have a federation or a new state, we can’t even have a Parliament endowed with a sovereign power. The true European question is thus one of constitution, a foundational issue rather than one of ordinary or representative politics.

by Hélène Landemore, 31 March 2008

Further reading

 Nadia Urbinati’s web page

 Bernard Manin’s web page

To quote this article :

Hélène Landemore, « Is representative democracy really democratic ? . Interview of Bernard Manin and Nadia Urbinati - New York, April 10, 2007 », Books and Ideas , 31 March 2008. ISSN : 2105-3030. URL :

Nota Bene:

If you want to discuss this essay further, you can send a proposal to the editorial team (redaction at We will get back to you as soon as possible.


[1I elaborate at greater length on this point in the post-script to the German translation of my book. See “Publikumsdemokratie revisited. Nachwort zur deutschen Ausgabe”, Kritik der repräsentativen Demokratie, Matthes & Seitz, Berlin, 2007

[2Voir Mark N. Franklin, Voter Turnout and the Dynamics of Electoral Competition in Established Democracies since 1945, Cambridge University Press, 2004.

[3Voir, entre autres, Pippa Norris, Democratic Phoenix: Reinventing Political Activism, Cambridge, Cambridge University Press, 2002

[4Bruce Ackerman, We the People: Foundations. Cambridge, UK: Cambridge Univ. Press, 1991, p. 181.

Our partners

© - Any replication forbidden without the explicit consent of the editors. - Mentions légales - webdesign : Abel Poucet