63 rd Political Studies Association Annual International Conference The Party's Over? March 2013 City Hall Cardiff

Transcription

1 1 63 rd Political Studies Association Annual International Conference The Party's Over? March 2013 City Hall Cardiff REPRESENTATION AS POLITICAL PROCESS: THE STATE-SOCIETY RELATIONSHIP IN A DEMOCRATIC BRAZIL Debora Rezende de Almeida Professor, Institute of Political Science, University of Brasilia ABSTRACT The incorporation of civil society actors in the process of public policy formation is one of the most significant innovations in the democratization of Brazil. The new institutional architecture has expanded the connections between State and society and inaugurated new forms of political representation in the country, above and beyond electoral dynamics. However, the recognition that civil society actors are, de facto, exercising representation, presents new challenges for democratic theory, some of which will be discussed in this paper. First, we propose to re-examine the concept of political representation, taking as our foundation two main ideas: that representation is a process spread out over time and space, and that it is permanently under construction. Secondly, to understand the democratizing potential of these transformations, the article reinterprets the criteria of democratic legitimacy of representation. It suggests that legitimacy is constructed during representation, not merely obtained through electoral authorization. This does not require us to abandon the tension between representatives and representees (those represented), or the mechanisms of authorization and accountability. It is important to reframe these ideas as a continual process, differentiating between the various examples of representation. Finally, the paper presents an interpretation of civil society in public institutions of co-management in Brazil. We argue that such institutions include a type of authority that grants social actors the power to decide in the name of a broader public. The collective actors who participate in these institutions acquire a positional authority, where they are invested by the State with the legal power to deliberate and decide. This is an authority that is not merely an attribute of their action, but also emanates from the position they occupy in the sphere of the State when participating in a decision-making process. KEYWORDS: representation, democracy, legitimacy, institutions of co-management

2 2 I. INTRODUCTION In the past twenty years, Brazil has been transformed into a democratic country with an impressive variety of participatory practices, which link State and society in a process of public policy formation and redistribution of goods and services (Avritzer, 2009). The most noteworthy experiments have been Participatory Budgeting and policy counsels, whose practices and models of participation have inspired the creation of institutions in different places, and studies of the meaning of contemporary democracy. Until recently, the principal analytical tools, in both the Brazilian and the international literature on the experience of inclusion of civil society, have revolved around participatory or deliberative theory. In this past decade, research has taken an interesting turn toward the representative role of social actors (Vieira and Runciman, 2008; Urbinati and Warren, 2008; Avritzer, 2007; Lüchmann, 2008; Lavalle, et al., 2006; Saward, 2011). Although public voicing of interests is common practice in protest action and expressions of collective action, only recently has political theory begun to call it representation (Lavalle, 2011). Although the theoretical discussion in Brazil does reflect the international discussion, it should be recalled that in practice, civil society representation in the former takes on a different dimension. In Brazil, participation is channeled into the political system, grounded in the incorporation of civil organizations and collective actors into the public policy formation process. These actors are not merely recognized as spokespersons for society; they also have rights and the power to decide (ibid, p. 15). In contrast to international models based on the informal influence of social actors, the institutions that connect the State and society form a part of a process of institutional pluralization of Brazilian democracy. Going beyond the central questions in the recent literature on non-electoral representation, such as what it means to represent and what are the foundations of the legitimacy of representation, the institutionalization that has gotten underway with the redemocratization of Brazil confronts us with several new theoretical challenges. The main aim of this article is to answer these two questions, which refer to the definition of representation and the foundations of its legitimacy, starting from a consideration of the differentiated nature of the representative exercise in contemporary societies. Our investigation is divided into three sections. In the first, we present a redefinition of the concept of representation, based on a critical reading of recent interpretations of political theory. It is our intention to bring to the fore two characteristics of representation that are present in the theoretical writings in which the concept has been re-examined, to wit, the idea that representation is a process spread out in time and space and in permanent construction. Representation involves different actors and arenas, of the State and society, who need to maintain continual interaction with those represented.

3 3 In the second part, we discuss the foundations of democratic legitimacy. Because of the fact that both Urbinati and Rosanvallon reduce society s political activity to political judgement or indirect influence, the problem of the legitimacy of representation by social actors is not adequately treated. In this paper, the processual interpretation of representation implies a shift away from the idea of legitimacy obtained in the foundational act, i.e. electoral authorization, situating it in the representative process. Thus the legitimacy of representation is based on a twoway relationship, involving both the representative s ability to express different interests and the represented s ability to recognize the legitimacy of such acts. The article discusses the difficulties involved in operationalization of this ideal in experiments with civil society representation, in view of the lack of institutionalized control and the lack of definition of who the constituents are. It proposes a reading of the processes of authorization, accountability and responsiveness that does not involve the electoral dynamics. The third section enters the universe of institutions of co-management and/or social control, which is the term utilized by Lüchmann (2011). The discussion will point out the differences among these institutions in the exercise of representation. Bearing in mind the characteristics of these spaces of power sharing between the State and society in the process of public policy formation, we argue that the first difference between these institutions vis-à-vis informal experiences lies in the authority exercised by civil society. Taking as our starting point Saward s (2011) review, we argue that social actors acquire a degree of in-ness, of positional ( in authority ) authority when invested by the State with the legal power to deliberate and take decisions. In addition to the recognition of that authority by the State, civil society has a kind of authority ( an authority ) which is built on a foundation of practical knowledge or related to experience with the problem or with public policy. The interdependence of these two types of authority (in and an authority) is crucial to the democratization of representation. Two other differences will be highlighted. The first is related to the process of authorization, considered indeterminate and less fluid. The second concerns the relationship between representatives and the represented. Although institutions of co-management do conserve the discursive dimension of control, the relative proximity with the public affected by the policy and with the policy community makes possible more direct accountability relationships. This is in addition to the control that may be exercised by the state actors, because the deliberation is conducted in close contact with them. In this paper, we consider that understanding the specificities of civil society representation is important, in order for us not to fall back on a deterministic assessment that democratic represention by these actors is impossible. Frequently, the absence of a traditional relationship with somebody who gives the orders, authorizing an agent to act in his or her name, has condemned non-electoral experiments with representation to a diagnosis of being nondemocratic or non-representative. Thus empirical analysis of the democratizing potential of institutions of co-management and their effects on political representation depends on a normative understanding of what it is that defines democratic representation.

4 4 II. REFORMULATION OF THE CONCEPT IN CONTEMPORARY DEMOCRATIC THEORY The current widespread dissemination of the concept of representation in political theory has been accompanied by a change in the way the phenomenon is understood. Although widely discussed in the literature, political representation had generally been associated with a discussion on the topic of representative government and the logic of representation in the spheres of the State (Araujo, 2009; Knights, 2009; Lavalle and Araujo, 2008; Vieira and Runciman, 2008). Recent research has gone beyond the study of parties, parliaments and electoral rules to focus on the transformations in representation, from the viewpoint both of the traditional institutions and of the new institutions and practices of representation involving State and society. The appearance and empirical confirmation of non-electoral forms of representation present us with several theoretical challenges. In this section, we will concern ourself with the redefinition of the concept, highlighting the two characteristics considered key to the comprehension of the plurality of representation in contemporary society: the processual and constructive dimensions of representation. The main authors referenced are Nádia Urbinati and Pierre Rosanvallon. The starting point for Nádia Urbinati (2006, 2011) is restriction of representation to questions of authorization and accountability. Inspired by Condorcet, the theoretician of the French Revolution, Urbinati suggests a diversification and extension of the temporalities and modes of expression of political life, in order to reconcile the concept of representation with democracy (Rosanvallon, 2009; Avritzer, 2007). Complex sovereignty is understood to be a process of continual interaction between the people and their representatives. It is expressed by the vote, but also at different times, for example when there is a referendum to revoke a mandate, or a possibility of revising the law. From this viewpoint, sovereignty is an historical construct, and elections are one of the many dimensions of representation and of the relationship between the State and civil society. The people must be continually re-created, and the relationship between the governed and those who govern them is not passive. The manner in which Urbinati suggests rethinking the concept of popular sovereignty, as formulated by Rousseau, has already been examined in various texts (Avritzer, 2007; Araujo, 2009; Loureiro, 2009; Faria, 2010). The important thing in this article is to point out the critique of the notion of sovereignty as alienation of rights and a single act of the will. For Avritzer (2007, p. 452), Urbinati has suplanted Rousseau s view of sovereignty as an act of will, replacing it with the notion of political judgement. In this paper, we argue that it is not a question of mere substitution, but of an interpretation of sovereignty grounded in two components, the will and political judgement. 1 Thus elections are not an act of transfer of sovereignty, but the will 1 Of course this diarchy is clearer in the 2011 article, but already in her 2006 book. Urbatini had stressed that it is necessary to bring the idea of sovereignty into the domain of judgement and opinion; that is to say, these are factors of sovereignty, not their substitutes (ibid, p. 106). It is not clear in her more recent paper how the will regains its

5 5 expressed at a determined moment. Authorization is not substitution or renunciation of participation, since the individuals retain their power to form judgements, ideas and opinions (Urbinati, 2006, p. 104). The legitimacy of representative democracy depends both on free regular elections and the activation of permanent communication between civil and political society. In this interpretation, representation has an intrinsically dual nature. On the one hand, it transmits the idea of acting on behalf of another, as an authorized agent. On the other, it leads us to the notion that it is a process of creation of a unity that did not exist previously. According to Urbinati (2011), this duality of representation calls for characteristics which are also of a dual nature, one active and the other passive. To bring this creation to life, representation depends both on the will that authorizes the representative to act in somebody else s name and on politial judgement or monitoring on the part of the represented. 2 Will and judgement, immediate presence (through the vote) and indirect presence (freedom of expression and association) are intertwined in society. As the State needs to be continually re-created and dynamically linked to society in order to pass laws, the dualism between State and society is bridged. For Urbinati, the many forms of information, communication and influence that citizens activate through the media, social movements and political parties are not accessories; rather, they themselves constitute political representation. In this interpretation, popular sovereignty is a source of endogenous tension between the institutionalized power of the State and the extra-institutional power of society. Representative democracy begins with elections but goes beyond them to include both participation and the non-electoral representation exercised by social actors. The latter differs from electoral representation in that it lacks an authoritative power, even though it may have the power to influence the political direction of a country (Urbinati, 2011). Urbinati s contribution lies in presenting a theory of democratic representation that is not confined to representative government or its traditional institutions. Thus, mainly in her 2011 article, the author incorporates civil society s non-electoral representation as an expression of political judgement. However, civil society does retain the negative power to influence, investigate and censure. This power is negative for two reasons: it has the function of stopping, status in representative politics and is reconciled with judgement. She seems to indicate that sovereign will is expressed in intermittent electoral moments (the sovereign will is tied to the act of authorization, and to direct presence, with the authority to take decisions); and political judgement, which also constitutes a factor in sovereignty, is expressed in citizens continual influence and power vis-à-vis the representative institutions (their indirect, informal and non-authoritative presence) (2011, p ). 2 Here Urbinati is clearly contesting the hobbesian thesis, according to which representation, as an act of creation of a unity that did not exist prior to the pact, has no obligation to the will of the individuals. For Hobbes, it is the unity of the representative, not of the represented, that makes a person one. In view of the fact that representation is not an act of portraying a previously unified body, it can be performed by an individual body of a representative nature. Contrary to the parliamentarian authors who argued that representation must be in the image and likeness of the population, Hobbes argues for an absolute power, in which the representative acts on his own behalf (Skinner, 2005, p. 173).

6 6 controlling or changing the course of actions of the elected representatives; and it can be expressed through direct channels of participation such as the referendum and recall, as well as by indirect, informal participation, through social forums, movements, civil associations or street demonstrations. One of Urbinati s contributions lies in her renewed interest in representation as key to democratic politics. In contrast to the utopian vision of a society that is participatory at every moment and in every affair that touches the individual, politics needs to be seen, through the lense of representation, as an indirect activity. Likewise, her comprehension of the negative nature of sovereignty is relevant, in that society does not govern itself, nor does it lose control of representive politics. Nevertheless, there is one problem that needs to be resolved, especially in her 2006 paper, concerning how the new forms of representation of civil society are integrated. Although at the end of the book she does propose a circularity between public judgement and elected representatives, she winds up assigning critical weight to the political parties in the unification of the interests and ideas of society (Loureiro, 2009). Furthermore, society only responds to the actions of institutionalized representation, and takes a passive, or non-active, attitude in decision making. Despite these limitations, her contribution to the discussion of sovereignty and representation is worthy of note. For Araujo (2009, p. 55), Urbinati does not establish a fixed separation between the two poles, State and society, into which, according to the traditional formula of liberalism, representation must fit. It is representation that splits the social space asunder, that produces and reproduces this division between State and society. This dynamic notion of representation as an activity that mediates between the diverse partialities of social life and politics is key to a comprehension of representation as an act of manifestation of sovereignty. As expressed in the title of this paper, Urbinati suggests thinking of representation as a political process, in contrast to the static and privatist model of a contract that confined it to a relationship between elected representative and represented (Urbinati, 2006, italics in the original). Rosanvallon (2008) likewise perceives the act of voting as part of the positive power, and the acts of society, or counter-powers (in the sense of keeping vigil, sanctioning or impeding and judging) as constituting the negative power of the people. The actions of society produce a set of effects, and not necessarily a formal authority or explicit decisions that might be called political. Despite the similarity with Urbinati 3 with regard to the perception of society s negative power, there is not, in Rosanvallon s thinking, a search for a model of democratic representation in which elections play the role of unifier, howbeit temporarily. The French theoretician s processual view of representation is marked by a contradictory coexistence, where society s mistrust plays a democratic role of resistance to the constituted powers. Practices considered apolitical, and the people s mistrust, can help generate decisions that are more in harmony with 3 While Nadia Urbinati uses the register of political theory, seeking a normative model for thinking about representative government, Rosanvallon is closer to social theory, where representation only makes sense if thought of over time and with continual reconfiguration.

7 7 democratic principles. 4 On this point, the author argues that the crisis of representation is not a recent characteristic, but rather an attribute that has accompanied the democracies throughout time. The difference nowadays is said to be in the way society expresses its discontent, and its criticism of the representative process. Above and beyond society s mistrustful actions, Rosanvallon (2009) presents three forms of legitimacy in ascension in the contemporary world. For each type of legitimacy, tied to the ideas of impartiality, reflexiveness and proximity, Rosanvallon (2009) lists a set of institutions that are said to have the role of expressing them. 5 For the purposes of this paper, it is important to stress his view of the forms of proximity, which are directly related to representative government and the actions of society. The legitimacy of proximity is not tied to any special type of institution, but to a set of social expectations concerning the behavior of the governors, to bring them closer to the citizens (ibid, p. 35). The different manifestations and institutions of civil society are perceived as part of what is called the democracy of interaction. There are two political functions of the democracy of interaction: a labor of justification (a confrontation of the explanations of power with the interventions of society), and an exchange of information between power and society (an instrument of government for one and a form of recognition for the other (ibid, p ). Although it is not possible to describe in detail the author s argumentation concerning the role of the different forms of legitimacy, it may be pointed out that civil society also plays an indirect role in its formulation. That is to say, society either exerts pressure on and surveillance of the acts of government, or classifies the actions of the latter through providing information. At any rate, the important thing is his expanded conception of political as something that surpasses the dimensions of institutionalized politics. Bearing in mind the limitation of the electoral moment, in which the people, which exists only in its instaneous and evanescent form, is given a voice, Rosanvallon advocates a permanent representation. It is characterized by a constant clash between the manifestations of society (of mistrust and dissatisfaction with the policies of the representative government) and institutionalized politics. The extended temporality of representation in Rosanvallon (2009) and the recognition of the permanent tension among the different forms of legitimacy imply, as a consequence, the criticism of modern separation between the institutional arrangement and society. Returning to an interpretation of representation as circulation (Urbinati), or as a permanent process (Rosanvallon), can help us to perceive that the foundations of legitimacy of representation reside 4 For a discussion of the problems of mistrust in democracy, see the critique by Urbinati (2010), Unpolitical democracy. Political Theory, n. 38, v. 1, p The appeal to impartiality is present in the regulatory agencies, technical counsels and commissions, because they are considered to be more objective and suited to the search for a unanimity attained in an impartial manner. As for reflexiveness, its main example lies in the Constitutional Courts, which have the role of multiplying the partial and plural approaches to the common weal. The reflexive role of democracy, instead of the distancing present in the ideal of impartiality, is stressed.

8 8 in the constant capacity to mediate between the two poles, State and society, going beyond the moments of supposed electoral unification. This is the intention of several contemporary authors. Even though it is not possible to go into the various proposals being put forth, one may state that what there is in common among these projects is an attempt to disentangle the concept of representation from a static view in which it is seen as a fact or property of a determined collective body or structure. The idea that representation is a political process, an open and infinite project that shifts around in time and space, is what unites the various authors (Urbinati, 2006; Rosanvallon, 2009; Dryzek, 2000; Saward, 2010; Bohman, 2012). The constructive dimension is the complementary part of this understanding of representation as a political process. This viewpoint implies avoidance of unidirectional analysis, which focuses solely on how the representative is formed. It is necessary to observe the different parts and moments of the representative process, emphasizing the active role of the represented and the disputes about what it is that is represented. The constructive dimension of representation calls for a vision of representation constituted in the course of time, in which the constituency is fluid and dynamic. Ankersmit (2002), with his esthetic understanding of the concept, defends the interpretative dimension of representation. For that author, there is neither a determined interest regarding what is to be represented nor a unified view of the represented, because these are also created through representation. The author criticizes the restricted and naïve conception in terms of a principal-agent relationship, because the interests of the electorate are not cohesive, nor are they expressed or given at the electoral moment. Borrowing from the artistic world the idea that a form of representation is never a simple copy of a pre-existant external reality, 6 but rather always a creation of something new, Ankersmit calls our attention to the permanent state of tension between the representatives and those they represent. Whereas the literature on representation has not devoted much space to a discussion of the represented, Saward (2010) argues that representation must be thought of as an activity. Bearing in mind that it is impossible to break down the distance between these two poles, representative and represented, the concept of representation will benefit from a more dynamic and plural notion of politics, in which the questions, who represents an what is being represented are frequently under construction. Thus it is not merely the definition of who the people are that is in dispute, but also who is the agent of representation. It should be recalled that an interpretation of representation as a dynamic phenomenon was already present in the classic book by Hanna Pitkin (1967), in which she defines it as a creative activity, a substantive acting for others. As a process of action, representation necessarily involves an overlapping of activities, a plurality of arenas and publics. 6 The example generally given by the author refers to a painting of a landscape which the painter is unable to replicate; rather, he creates a representation of it on the canvas.

9 9 Political representation is primarily a public, institutionalized arrangement involving many people and groups, and operating in the complex ways of large-scale social arrangements. What makes it representation is not any single action by any one participant, but the over-all structure and functioning of the system, the patterns emerging from the multiple activities of many people (Ibid., p , author s italics). Further on, the author points out other modes of representation, the descriptive and the symbolic, and different agents who do the representing, such as State officials, interest groups, lobbyists, and people who act before government agencies. According to Wampler (2012, p. 673), Pitkin shows us a way to theorize about where and when citizens can act as participants, representatives and represented. Despite this advance, the last chapter of the book winds up paying attention to one specific type of action: speaking on behalf of others in an established institutional context. In this conclusion, the author seems to dillute the plurality established at the start of her book, when she points out the inescapable paradox inherent to representation: making what is absent present. The contribution of contemporary democratic theory leads us to think about other modes of action, in which the représentant deserves to be called representative; thus, it recovers the continual and plural aspects of the concept, in which it is difficult to establish a unity. Starting from these references, I propose, in this paper, to accentuate the processual and constructive character of representation, in order to interpret the potentialities and limitations of civil society representation. Seen in this way, representation is an activity that continually shifts in time and space, and that involves different actors and arenas, from the State and in society. In this process, there is not always a prior definition of what is represented and who represents. As will be discussed in greater detail in the next section, democratic legitimacy depends on ongoing interaction between representative and represented. It should be recalled that the concept of representation as relationship is already present in the work of Young (2000). For her, representation is a differentated relationship among actors engaged in a process spread out over time and space. It is a process that depends on mediation and connection between the parties (Ibid., p. 127). To be sure, the very fluidity of the concept gives rise to a number of questions in relation to its democratic legitimacy, which I shall leave for the next section. Furthermore, one runs the risk of associating just any public expression with representation, dilluting the concept or the differences among types of representation. For example, Saward(2010), in defending a shift from the idea of representation as a fact or fixed quality of the State to the view that it is an event, proposes examining the representative claims. The problem is that the author winds up incorporating in just any form of citizens initiative whatsoever an expressive form of representation, or what he calls self representation. It is not clear what the difference is between participation and self representation. The notion considered in other articles, that a person is at the same time participant, representative and represented, seems more appropriate in dealing with plurality and overlapping functions (Ibid, 2006, 2008). In institutions of co-management in

10 10 Brazil, this dimension seems very clear. A member of a policy counsel, for example is a participant in a civil society organization, a participant on the counsel and a representative of the organization in the counsel and, at the same time, is represented by other people and speeches given within that venue. Furthermore, for political scientists concerned with eliminating the ambiguous character of representation, its fluidity and indetermination may transform the concept into something empty and nebulous. However, it is important to remember Pitkin s warning that to represent involves a paradox. The constructive sense of representation shows that the unity that may be obtained through democratic representation will always be subject to disputes within the terms proposed for this system. Besides, recognizing that representation is a process in construction, involving various actors and mobilizing different conceptions of authority, is not the same as saying that it is always democratic. III. RETHINKING THE FOUNDATIONS OF DEMOCRATIC LEGITIMACY The view of representation as permanent and processual demands as a complement a discussion of democratic legitimacy. In shifting the evaluation of legitimacy from its exclusive focus on the foundational act, i.e. authorization, and situating it in the process, we need to ask what basis we have for evaluating civil society representation. Thus we need to bear in mind the dual nature of representation, involving a relationship between the two parties (representative and represented), as well as the creation of a unity that did not exist before, as propounded by Urbinati. For the theories of representation in the spheres of the State, the legitimacy of the representatives, as creators of unity, is given in the moment they are constituted. Electoral authorization permits the free, egalitarian expression of the will of the individuals at the foundation of power. To the degree that the vote expresses both agreement with the method by which the representative is selected and with the political results themselves, authorization also generates a sense of political obligation (Manin, 1997, p. 85). With regard to the relationship between the paties, the individuals are in a position to evaluate, in institutionalized moments of control, the degree to which a government s decisions approximate their own interests: prospective and retrospective evaluation based on the ballot. The limits placed upon vertical control and democratic expectations in elections are not negligible. According to Przeworski, Stokes and Manin (1999), the vote can be seen more as a promoter of electoral democracy than as true vertical control of the representative. Furthermore, democracies that obey the rules of democratic procedure may not produce substantive results such as economic equality and accountable governance (Fox, 2000). This forces us to recognize that the marriage between processes and results is contingent, and that accountability is always an imperfect process, dependant on the interaction between vertical and horizontal controls and the

11 11 different modes of representation. Despite these limitations, one must acknowledge that in the last analysis, the vote makes it possible for everyone to have an influence within the representative activity. It is when one transfers this notion of legitimacy to non-electoral experiments that different problems arise. How may we regard the recognition of authority (the right to speak for or in the name of others) and of representative activity (the degree to which the representation approximates people s interests and preferences) in a model without institutionalized control and a clear definition of who the constituents are? This is a problem that Urbinati and Rosanvallon do not confront, for the political judgement exercised by social actors plays a role of approximation or sympathy, or even contestation, between their ideas and the decisions of the State. In addition, one must not forget that Urbinati, for example, is talking about an open society, free and pluralistic, in which political judgement may be exercised by different actors, in different instances. Political equality as a condition of legitimacy, a guide to judgement, and a promise to be fulfilled is the reference point in relation to which the political process of democratic representation defines its goals and language, and projects and submits itself to the critical judgement of the citizens (2006, p. 6). To be sure, in the face of the material, cultural and political inequality that impedes the autonomous manifestations of individuals, we need to consider that access to civil society s mechanisms of influence and citizens critical judgement are also unequal. Miguel (2011) is one of the writers who have denounced these limitations. For that author, representative activity is a form of exercise of power marked by conflict and defense of interests, and may not, therefore, dispense with the notion of autonomy. The challenge of democratic representation is to permit that individuals and groups capable of autonomously defining their own interests find themselves represented by agents with whom they have an identity and a capacity to communicate. This depends in turn on social conditions more conducive to political participation and individual autonomy. The author acknowledges that there is no easy way out of this dilemma, in view of the structural inequalities of capitalist society and the absence of universalization of the material informational and cognitive resources that permit its development. For Miguel, these problems cannot be resolved through replacing the autonomous participation of individuals with a representation of social actors, who would supposedly speak in the name of the excluded. Civil society representation, about which different authors have been theorizing, removes the institutional instruments of authorization and accountability. Instead of focusing on the refinement of democracy via the plurality of representatives of civil society who present themselves as spokespersons for the needs and welfare of others, he proposes expanding participation in the public debate and the dimension of production of interests. For that to occur, the social processes that coerce the formation of interests must be taken into account. Despite the importance of denouncing the inequalities that permeate the processes of State-society relations, two problems with this analysis should be pointed out. First, with regard to the view that non-democratic attitudes are also present within civil society. This is a viewpoint

12 12 that has been spreading among students of participation, who have long since abandoned the idyllic vision of society as a pole of virtues (Dagnino, 2002). I argue in this paper that the acknowledgement of inequalities must not be allowed to put the interpretation of societal representation into a straight jacket, considering it to be necessarily nondemocratic, as Miguel (2011) suggests. Secondly, Miguel s (2011, p. 59) verdict that civil society ativity is a false solution to the problem of exclusion and autonomy, because of a lack of concern with the mechanisms of authorization and accountability, is not fair to most of the authors who have been rethinking representation in recent years. The notion that every form of representation claiming to be democratic must pass the test of public control is supported in the lion s share of the literature in which the concept has been re-examined. Despite the absence of a procedure of the eletoral type, one may suggest less formal and more permanent processes of authorization, and maintenance the dialogue between the representatives and those they represent. In this section, we will discuss the foundations of authorization, accountability and responsiveness in experiments with civil society representation, whether self-authorized (Urbinati and Warren, 2008) or instituted through random methods (Brown, 2006), which emphasize influence rather than decision making power. III.1. Non-electoral authorization Although in informal experiments in civil society it may not be possible for individuals to directly authorize or de-authorize the representative, they can express some kind of choice. This is the defense put forward by Saward (2011), for whom what distinguishes democratic from nondemocratic claims is the degree to which the former demonstrates a sufficient degree of chosenness, within the context of a reasonably open society. In other words, there is reasonable evidence that the constituents of a given representative demand accept this claimant as speaking and acting on their behalf, with regard to a certain set of proposals (ibid, p. 90). In those spaces and among those players that are self-authorized (Urbinati e Warren, 2008), there is no formal transfer of power such as occurs with an election or the authorization received by counsel members, associations and unions. Nor does the authorization express the manifestation of a preference that can be aggregated. The authorization or choice must be sought in the public recognition of those who claim to speak for individuals and groups. For example, recognition of social actors as spokespersons on topics and policies must be sought in the public sphere of discourse, based on the public justification of these representatives (Maia, 2012). In that sense, the authorization is intrinsically connected with the mechanisms of accountability. As Castiglione and Warren (2006) show, where there is no authorization, accountability can provide it retrospectively and reflexively. The authors point out that the authorization can be given through the ability to attract members, convergence of purposes, descriptive characteristics or the

13 13 public visibility that is promoted. The important thing about this view is that it accentuates the dialogue that is to be established with the representatives, so that they may reappropriate the representative demands or choose their representatives. Allied to this discursive dimension there exists, in the spaces that are formed (whether by the initiative of society or the State), such as deliberative polls and citizen juries, a concern with the method of selection of the representatives. In these cases, methods that are alternative to elections, such as the draw or random sample selection, are once again taken up, with the aim of guaranteeing the representativity of the representatives (Fishkin, 2009; Brown, 2006; Parkinson, 2006). This is an attempt to have the representatives mirror the characteristics of the population, or be selected in proportion to the size of the group and the intensity of those interests in policy results. In view of the difficulty in operationalizing more comprehensive and inclusive methods like the lot, those experiments have, in practice, adopted sample by quotas. To that end, certain cleavages and/or socioeconomic, gender or ethnic characteristics of the groups that are to participate are defined. The difficulties involved in determining the types of cleavages, and the natural overlapping among them, are just some of the limitations of this type of authorization. Just as in self-instituted spaces, the legitimacy of representative activity must not be divorced from public judgement. III. Non-electoral accountability With regard to control and public judgement of the representatives of civil society, Lavalle and Isunza (2011) argue that in contrast to electoral accountability, in which the act of consent encompasses both the moment of definition of the representative and the mechanisms of institutionalization of control, civil society accountability depends on a temporally shifting process. This leads us to a diachronic control relationship, in which representation must no longer be measured by the number of citizens who are able to participate in these forums, but rather by the nature and quality of the democratic representation obtained in the participatory process (Warren, 2006, p. 4). Thus, we might also think of accountability as based on a constant movement of information, argumentation and sanctioning between representative and represented, to use the three dimensions proposed by Schedler (1999). 7 Despite Lavalle and Isunza`s relational interpretation of accountability, we have, in this paper, suggested that the two directions of an analytical enterprise, vertical and horizontal, 8 be 7 Although civil society does not suffer the direct sanction of the vote, one might think of the symbolic sanction (such as a loss of credibility with a determined public) to which it is subject when it acts in a manner not responsible to the interests of those it claims to represent. 8 The literature on accountability has made great advances along these dimensions in recent years; among these, the crosscutting nature of these processes and the diversity of the actors involved (state, collective societal and

14 14 maintained, as in O`Donnell`s classical formulation. This orientation may help us to construct methodological strategies to evaluate the relationship between representatives and the represented, and of the former with other agencies and actors of social control. On the vertical dimension, the representative must develop ways to connect with those represented, who will be the target of the demands and of the discourses pronounced by him or her. As for representatives who are self-authorized or instituted by random methods, there are many fragilities in societal control. Nevertheless, it must not be forgotten that often the maintenance or survival of these actors depends on how well they are able to answer the questions they themselves put forward.therefore, the representatives of society need to first foment public discussion, in order to be able to bring to the attention of public opinion the positions they defend and represent. A good example of how to counterbalance the opinion that emerges in a deliberative process with the population s direct recommendation was utilized in the Citizens Assembly of British Columbia, responsible for discussing the electoral system in the province (Brown, 2006). In that experiment, the orientations that emerged from the deliberative forum were submitted to a subsequent referendum, in which the population had the right to express their approval or disapproval of the proposals. Secondly, one must not forget the control exercised by the elected representatives themselves. Without communication with the formal venues of decision making, social actors will be incapable of influencing the public agenda. In this relationship, society contributes along both the informative and the argumentative dimensions of the institutionalized decision making process, as Rosanvallon pointed out. Besides being subjected to the symbolic sanction of the state or societal actors, who do not acknowledge them as legitimate interlocutors. Finally, other forms of control, less institutionalized and formal, may be cited; the media are an interesting example. They can play a key role in publicizing the acts of the civil society representatives, generating information and compelling argumentation (Lavalle, Houtzager, Castello, 2006b). As for the horizontal dimension, the dialogue of the representatives with their organizations and other institutions and spaces is the key point. In cases in which the representative belongs to some organization or other body, it can serve as a locus of accountability. Despite the critique of the tendency of leaders to acquire autonomy vis-à-vis the people they are supposed to be representing, as already pointed out by Robert Michels (1982), Fox (2000, p. 11) argues that this is not inevitable. The existence of internal mechanisms of horizontal accountability in civil organizations, as intermediary instances of participation, is a crucial complement to the control of the leaders, who do not need to deal directly with atomized members or individuals. These groups or instances of coordination help to share information and individual) stand out. In this paper, the vertical and horizontal dimensions are taken up once more, solely for the purpose of general illustration of the relations between representative and the represented, and of the levels of control.

15 15 generate alternative options and counterproposals, and to form potential leaders. Understanding the complexity of these interactions is essencial if we wish to avoid a deterministic affirmation that democratic representation by social actors is impossible. Control may also occur among institutions and spaces of representation. From this perspective, the recent literature on deliberative theory, focusing on the idea of systems, has proposed to de-territorialize democratic representation, so that it may be perceived as dependant on the different networking that occurs in the public and political spheres (Parkinson, 2003, 2006; Hendriks, 2006). So it is necessary to go beyond the analysis of individual institutions and processes, in order to examine the interaction of the system as a whole (Mansbridge, et al., 2012). In view of the understanding of politics as a process in which it is hard to balance citizens different interests and viewpoints, this literature establishes a division of deliberative labor, in which it is necessary to combine different modes of representation and participation. The solution is to assess the degree to which deliberations that occur in one locale complement or promote negative shifts in others. This implies that the societal representation exercised on a specific venue, such as a policy counsel, may be subject to public control by other counsels or institutions. Or when an organization for the defense of women s rights is subject to the public control exercised by a broader network of organizations working in the same area. That same literature makes a fruitful contribution to evaluation of the vertical connections between representative and represented. The theory of systemic deliberation permits the analyst to perceive the degree to which civil society representatives are subject to control by or in interaction with other elected actors or society.the question to be investigated is the following: when are the discourses that are pronounced able, or not able, to go from one sphere to another; and what are the variables that explain this complementarity or shift? III.1. Non-electoral responsiveness In relation to responsiveness, the principal problem is the reduction in popular sovereignty due to the impossibility of direct influence on the mechanisms of societal representation. Thus the legitimacy of speech or action on behalf of the represented, in a manner that is sensitive to his or her demands, concerns not the direct consent of the individuals, but the way this creation takes place. One of the important proposals along these lines has been presented by feminist theory, which argues for the presence of certain groups to defend their interests. These traditionally marginalized groups could help give public visibility to the demands of their fellows and raise the awareness of elected representatives, to the degree that they share their ways of seeing the world with their represented (social perspectives, for Young, 2000), as linked to certain socially structured patterns of life experiences (Miguel, 2005).

16 16 According to Young (2000), the notion of social perspectives shifts responsiveness to the field of sharing certain characteristics, and to similarities in conditions of fellowship. Perspectives are important starting points in the construction of interests, and do not, in and of themselves, determine the contents of the deliberation (Young, 2000). As a proposal operating within electoral politics, the notion of perspectives has been adopted, for example, in policies of quotas for women and other excluded groups, and in the adoption of random mechanisms for filling public positions (Miguel, 2005). With regard to civil society representation, perspectives offer a way out for those interested in expanding the plurality of voices present in the spheres of decision making. Along these lines, random mechanisms are used to form deliberative groups taken from population samples, which study political problems and contribute to better informed decisionmaking, including, as they do, groups that had previously been absent from political discussions. Representation by perspectives is complemented by other analyses that stress the role of thematic affinity (Avritzer, 2007). Representativeness can be generated through the defense of themes and policies with which the participants have widespread experience, acquired in the course of their history of participaton and involvement with politics. What civil society organizations, social movements and individuals are representing in these cases is a discourse on rights or themes, not a specific set of people. Along similar lines, Dryzek (2001) emphasizes discursive representation, in which the role of the actors is to give voice to discourses that are consistent with the public sphere and not expressed in elections. That author goes so far as to propose constituting chambers of discourse, to bring together representatives of different discourses in a deliberative mini-public (Dryzek and Niemeyer, 2008). It should be pointed out that these are elements that may activate responsiveness or help stimulate discussions that may foment the formation of the interests of those said to be represented. Therefore, it is not a question of defending eloquent speakers who represent without the consent of the represented. Public control is key to democratic representation, but the way it is operationalized varies from one experiment to another, as we have seen above. Despite these differences in the operationalization of accountability and responsiveness, what is common in contemporary democratic theory is its avoidance of a notion of previously given and subsequently aggregated preferences. Representation takes on a sociological dimension of construction of preferences, and even of the represented himself or herself, as the ideas of Saward (2010) seem to clearly show. From the author s perspective, it is necessary to examine the representative claims, an approach that stresses the active side of representation and makes room for a creative normative endeavor in the radicalization of our ideas about who and what can be genuinely politically representative. In the words of Abers and Keck (2008b), civil society representatives are activists promoting causes, much more than delegates of constituents with pre-defined preferences.