Hanna Pitkin's The Concept of Representation is owed a careful re-reading. First published in 1967, a “watershed year” in the United States for studies of political representation, it would go on to shape four decades of empirical research into the practice of American democracy.Footnote 1 That a work of normative political theory captured the attention of empirical political scientists in the mid-1960s, when the behavioral approach to the study of politics was in its heyday, is noteworthy. That this immensely popular text should have been an intricately argued work of critical theory, one that set itself against premises that are foundational to rational choice theory on the one hand and liberal democratic norms on the other, is downright surprising.
The argument of Concept was intellectually radical because it challenged two of the most deeply held convictions of the field regarding the study and practice of democratic representation.Footnote 2 Pitkin rebuts a presumption that is fundamental to both normative and rational-actor analysis of democratic process: that representing can be modeled as a principal-agent relation between legislator and constituency. She counters that political representation is a “public and institutionalized arrangement involving many people and groups, and operating in the complex ways of large-scale social arrangements.”Footnote 3 She also opposes the intuition, definitive for late twentieth-century liberalism, that citizen preferences are and ought to be the “principal force in a representative system.”Footnote 4 Far from merely amplifying or reflecting citizen preferences, Pitkin affirms that political representation is an activity without “guarantee.”Footnote 5
In what, exactly, does that activity consist? While it is the focus of her book, Pitkin does not make this question easy to answer as she shifts between speaking about representing in general terms and political representation specifically without always marking the difference between them. By representing in general, Pitkin means the “making present in some sense of something which is nevertheless not present literally or in fact.”Footnote 6 In political representation, presumably the “something” that must be made present is a constituency, its interests, and its preferences. Indeed, Pitkin does define political representation as “acting in the interest of the represented, in a manner responsive to them.”Footnote 7 Yet, if we take her two radical moves seriously, she cannot mean responsiveness as it is typically pictured: a dyadic relationship driven by citizen preferences. With her emphasis on representing as activity, Pitkin suggests that political representation involves a great deal more creativity than conventional democratic norms allow.
In its most radical features, Pitkin's argument presents what I call the “constituency paradox”: democratic representation must posit as a starting point constituencies and interests that can take shape only by its means. This paradox is radical precisely for casting doubt on “responsiveness” as a normative ideal and operational index of democratic representation. For if constituencies and their interests are endogenous to the representation process, then what does the proponent of democracy gain by urging legislators to respond to them? What does an empirical finding reveal when it tracks policy outcomes and public opinion? Put simply, if constituencies and constituent interest form not prior to, but over the course of, the representative process, then responsiveness to those interests is hardly a reliable indicator that democratic representation is functioning well.
It is odd and remarkable—although few have noticed it—that despite being widely cited, Pitkin's text did not radically reshape the orthodox conception of representation. In 1969, Kenneth Prewitt and Heinz Eulau warned scholars of democratic institutions that “to understand representation [in Pitkin's terms] as a systemic property we would have to re-think many of our conventional assumptions.”Footnote 8 Few heeded this warning. Morris Fiorina noted that it was typical for empirical studies of representation to disclaim responsibility for the “continuing normative and semantic arguments about the meaning of ‘representation’” and then to cite Pitkin's work as a “good survey” of that controversy.Footnote 9 And Malcolm E. Jewell observed in 1983 that, “nearly all [subsequent empirical] research” persisted in defining representation in the very terms that Pitkin called into question, as a property of “the relationship between the individual legislator and the constituency.”Footnote 10 Many distilled Pitkin's central message into a nutshell: “responsiveness is what political representation is all about.”Footnote 11
Her readers had every good reason to do so, given that Pitkin herself used the very term that her own critique should have prompted her to reject. The puzzle of Concept is, as two of her most appreciative readers noted, that “responsiveness” was a surprisingly ordinary place for Pitkin to end up “after considering every conceivable definition” of representation.Footnote 12 Did this term not suggest precisely the “kind of one-to-one, person-to-person relationship” to which Pitkin did not want political representation assimilated?Footnote 13 Responsiveness was a false friend—a term that invited misrecognition by its apparent familiarity. Whereas academic readers took up that invitation by lifting the term out of the context of the argument of the work as a whole, Pitkin herself fosters misreading by failing to provide any “clues” as to how “‘responsiveness’ as a systematic property of the political collectivity … can be measured in ways enabling the scientific observer to conclude that representation has in fact emerged.”Footnote 14 This omission made it much too easy for scholars to read Pitkin's unconventional argument in the terms that were most familiar to them.
Pitkin's thought takes an important turn with her second major text, on Wittgenstein. In that work, she proposes a conceptual shift from responsiveness to responsibility, and coins a term that I believe captures the radical implications of her earlier work. The core of Concept, rendered (retrospectively) in the language of the Wittgenstein study, is that acts of representation are “quasi-performative.”Footnote 15 If representing is a “making present,” then the emphasis is on the making.Footnote 16 In politics, acts of representation do not simply reflect constituencies and their interests but help to bring them into being.Footnote 17
This notion of “quasi-performativity,” which I explicate here, makes an important normative advance. It redirects attention from the typical focus on responsiveness—which tracks congruence between public policy and changes in public opinion and constituent preference—to a system-wide measure that I will term representation's “constituent effects.” As such, it bears on the contemporary debate regarding democratic “oligarchy” in the US, and on what Jacob S. Hacker and Paul Pierson have called “winner-take-all politics.”Footnote 18 Through the lens of Pitkin's work, this current work on the political effects of economic inequality is important not for calling attention to the “differential responsiveness” of the US political system, but for making its constituent effects a focal point of political analysis.Footnote 19 That is, it frames a crisis that has more to do with structural power—the “bias” of the contemporary democratic system—than with measurable, quantifiable differences in influence of its competing interest groups.
My argument proceeds over five sections. It begins by situating Pitkin's work in the context of the tradition of engaged political theory at Berkeley, and then in relation to empirical studies of representative democracy. Next, I offer a close reading of Concept, emphasizing its most radical strands of argument, and linking those radical strands to the concept “quasi-performativity” from Pitkin's Wittgenstein book. This concept, I argue, reorients the analysis of democratic representation from responsiveness to responsibility. In conclusion, I use this new vocabulary to characterize the contribution of recent empirical scholarship on democratic inequality to normative analysis of democratic representation. Here I suggest that normative analysis of democratic representation needs to focus on assessing the organizational dynamics that favor the upward transfer of wealth, while thwarting the mobilization of groups that would oppose that transfer.
Pitkin and the Berkeley School
Pitkin composed her text during the heyday of direct action, when the sit-ins, street protests, and occupations by civil rights activists, Students for a Democratic Society, and Berkeley's own Free Speech Movement threatened to make “representative democracy” an oxymoron.Footnote 20 The US political system was in the throes of transformation, as the Supreme Court's landmark decision in Baker v. Carr prompted waves of redistricting challenges in the states. For all that, there is little evidence of political turmoil in Pitkin's text. She makes no mention of student activism at Berkeley. Although she notes Baker v. Carr, the questions Pitkin poses have nothing to do with the issues that followed from that case, such as whether the right to vote implies a right to be represented, and whether representation, in turn, has to be both fair and equal—guaranteeing not just that votes count equally but that they are equally decisive.Footnote 21 Instead, Pitkin frames her text as an inquiry into the “question of what ‘representation’ means.”Footnote 22
Pitkin's preoccupation with representation during this period set her apart from her “Berkeley School” colleagues, who made the events on Sproul Plaza a centerpiece for their political theorizing.Footnote 23 In a series of essays originally written for the New York Review, Sheldon Wolin and John Schaar proposed to view Berkeley with its vast bureaucracy and ties to the military and to corporate agriculture as the “technological society” in microcosm.Footnote 24 They read the student “rebellion” as an expression of the powerlessness experienced by the middle class more generally in a US that was growing “ever less effectual in coping with the human environment and ever more destructive in dealing with the natural environment.”Footnote 25 By contrast, Pitkin's focus is academic. She speaks to empirical scholars who were expressing frustration at the failure of normative theorists to provide guidance for political science research. As Heinz Eulau, John C. Walke, William Buchanan, and Leroy C. Ferguson put it, the “history of political thought is studded with definitions of representation, usually embedded in ideological assumptions and postulates which cannot serve the uses of empirical research without conceptual clarification.”Footnote 26 In Concept, Pitkin sets out to use the “tools and techniques” of linguistic analysis to clarify the concept of representation by “formulating a one-sentence definition of [its] basic meaning,” and by providing an “overview of its use.”Footnote 27
As I noted, Pitkin defined representation as the “making present in some sense of something which is nevertheless not present literally or in fact.”Footnote 28 Judged by the task she set herself, this definition is a failure. Its qualifications and hesitations make it singularly ill suited to empirical testing. If representing means that something is to be made present but need not be present “literally or fully in fact,” then how are we to know when the represented is not present? As Adam Przeworski, Susan Stokes, and Bernard Manin have observed, “a government may be representative even if it is not responsive to the expressed wishes of the people, and conversely, it can be unrepresentative even if it follows the signaled preferences.”Footnote 29
Pitkin herself acknowledges that “to say that something is simultaneously both present and not present is to utter a paradox.”Footnote 30 Yet this is precisely the insight she hopes to convey to empirical colleagues caught up in the “mandate-independence” controversy, the question whether a democratic representative should be more like a delegate or more like a trustee. If the represented is both present and not present in representation, it turns out that following a mandate and being independent are “both involved in the concept of representation.”Footnote 31 It makes no sense to pit these against each other as alternative conceptions of political representation because “there is no rational basis for choosing between them tout court.”Footnote 32 Pitkin wants her colleagues to see that they are in the grips of a “conceptual problem” that no amount of empirical or historical evidence could settle because the ambiguity about the activity of representing derived from the “paradoxical requirement imposed by the meaning of the concept”: it requires the represented to be “made present” but not to be present “literally or fully in fact.”Footnote 33 The represented, in short, is “simultaneously both present and not present” in the act of representing.Footnote 34 Only when this “paradoxical requirement” is satisfied do “we begin to speak of representation.”Footnote 35
It is tempting to read this as David Runciman has done, as a “paradox of presence” that requires the represented to be “present and absent at the same time.”Footnote 36 It is worth asking, however, whether being “present and absent at the same time” is as paradoxical as Runciman assumes it to be. Certainly, if presence and absence are presumed to be opposites, then it is impossible to be both at once. But does this presumption hold? Jacques Derrida has argued that presence and absence are not opposites but complements, terms that—by our habit of taking them to be opposed—work to sustain the idea of “present being” or “simple presence,” the belief that “reality” exists prior to and as a condition of the possibility of its being represented.Footnote 37 That something needs to exist before it is represented may seem obvious, but it is precisely this common sense—which he terms the “metaphysics of presence”—that Derrida aims to call into question.Footnote 38 He argues that it deprives acts of representation of their agency by relegating them to a secondary role as “derivative” of an original which is temporally prior to and more authentic than its representation.Footnote 39 Whereas this view makes representing imitative and passive—the mere “reproduction of a presence”—Derrida counters that representing is active: it is “repetition” that lends to objects and to reality more generally both unity and “permanence.”Footnote 40
As Iris Marion Young has observed, it is a hallmark of conventional understandings of democracy to insist that the people be taken literally, “like at a town meeting” where “citizens meet in one place and make their decisions on one occasion.”Footnote 41 Young regards this insistence that “authentic democracy” must be face-to-face as a perfect illustration of Derrida's “metaphysics of presence,” because it relegates political representation to a secondary role and implies that, however necessary it may be, it is “impossible” to reconcile with democracy.Footnote 42 Young's argument is important for recasting the paradox of representation. Drawing on Derrida enables her to argue that the paradox is not inherent in the concept itself but, rather, it follows from the binary thinking that, as Timothy Mitchell has observed, produces the distinction between authentic “physical” reality and its “meta-physical” representation in the first place.Footnote 43 Young contends that the “way out of this paradox … involves conceptualizing representation outside [the] logic of identity” which, in the case of democratic political representation, calls on the representative to act as a delegate or messenger for his or her constituency.Footnote 44
Although Pitkin is no follower of Derrida, she similarly rejects such a mimetic account of representation, taking pains both to emphasize that representing is not merely derivative of a standing reality, and that the objects of representation are not simply given (or “present” in Derrida's terms). For Pitkin, constituents are not literally present but “considered as present”; they need not literally act but must be “conceived as” capable of doing so. As she repeatedly affirms, “representation means the making present of something which is nevertheless not literally present”; being represented means being “made present in some sense while not really being present literally or fully in fact”; the “represented must be both present and not present”; “in representing, something not literally present is considered as present in a nonliteral sense.”Footnote 45 I term these formulations Pitkin's “resistance to the literal.” This resistance, which crops up in Concept wherever Pitkin verges on the constituency paradox, marks her refusal of the logic of presence. It is also symptomatic of the fact that she lacks a conceptual vocabulary for her most radical insights.
Concept and Empirical Studies of Representation
Pitkin published Concept in the wake of what James Kuklinski and Gary Segura characterize as “two of the most influential studies on representative democracy, Miller and Stokes's (Reference Miller and Stokes1963) ‘Constituency Influence in Congress’ and Schattschneider's Semisovereign People (1960).”Footnote 46 These two studies, which painted strikingly different portraits of the democratic process, would influence the reception of Pitkin's work. The contrast between them is also significant today in light of the contemporary debate regarding the relationship between political and economic inequality.
Warren E. Miller and Donald E. Stokes' study was influential less for their finding on the mandate-independence question than for inaugurating the use of sampling as a “bold and imaginative” methodology to index constituency influence to the “congruence” between a legislator's votes and his or her constituents” opinions.Footnote 47 Like Pitkin's, Miller and Stokes' work lost nuance as it was appropriated by other scholars. Whereas they emphasized that a finding of congruence meant nothing in the absence of specific institutional conditions that would permit it to be construed as an effect of citizens” influence,Footnote 48 subsequent scholars applied their model reductively, by positing the accord between a legislator's votes and the surveyed preferences of his or her constituents (“congruence”) as a straightforward index of democratic representation, a.k.a. “responsiveness.”
Whereas the reductive reading of Miller and Stokes effectively rendered public opinion the “starting point” of democratic politics and democratic theory, Schattschneider took aim at precisely this assumption with his important concept, “scope of conflict.”Footnote 49 Schattschneider maintained that citizen preferences do not determine political outcomes but, rather, are defined in the course of political contest. This idea, Schattschneider rightly proclaimed, would prompt a “revolution in our thinking about politics” by overthrowing the concept of power based on the “familiar simplistic calculus … of a tug of war of measurable forces”Footnote 50 in favor of a structural understanding that holds the size and force of constituencies to be determined by the “scale of political organization and the extent of political competition.”Footnote 51 The scope of conflict gives preferences a unique status in Schattschneider's work. Although they are not fixed prior to political conflict, politicians do not simply script them; rather, elite competition shapes preferences by framing a conflict in more or less universal terms, and staging it on the terrain of either pressure politics or party politics. That interest group politics tends to be the dominant mode of conflict in pluralism, and cost-benefit reasoning tends to set its terms, accounts for what Schattschneider termed its “mobilization of bias” toward wealthy elites.Footnote 52
With this notion of “bias,” Schattschneider brought to light a political force that is neither a simple demographic fact like either the wealth or the (purportedly) pre-political interests of an individual or social group, nor an exclusively political achievement, like the “enlightened interest” that deliberative democrats posit as the outcome of well-conducted public debate.Footnote 53 Bias is a constraint on political conflict. It is a social and cultural factor with which competing elites must contend, as both a resource for and an obstacle to their efforts to stir up support for competing policy programs. Bias is also an organizational force. In Schattschneider's terms, the “bias” of pluralism inheres in the “forms of political organization” that it introduces.Footnote 54
This was, as the subtitle to Schattschneider's Semisovereign People announced, a “realist's” perspective on democracy. It was important for calling attention precisely to those forms of power that are not explicitly manifest in conflict, as acts of domination or unequal outcomes, but that work prior to conflict to set its terms. Jeffrey C. Isaac later elaborated the realist position as an epistemological critique of pluralists' empiricist understanding of causality that enabled them to erase bias because it was a “structural” and “necessary” rather than “manifest” and “contingent” power relation.Footnote 55 Building on Schattschneider, Isaac contended that analyzing the “bias” of pluralism would require a “relational” analysis of its power relations, analysis focused not on manifest interactions between agents but on the “underlying social relations which structure” them.Footnote 56
Schattschneider's work has a powerful legacy today, particularly among political scientists who have taken up the methods of psychology to study political reasoning.Footnote 57 At mid-century, when political science aspired to the precision of the positive sciences, the work of Miller and Stokes became the empirical paradigm of the moment. No doubt influenced by the ready availability of “cross-sectional measures of constituency opinion,” legions of survey researchers set off matching roll call votes to constituent polling data.Footnote 58 These studies treated interest as exogenous to political conflict, and responsiveness as a straightforward index of citizens' control over the representative process. Pitkin's Concept was assimilated to that dominant paradigm by the readers who seized on “responsiveness” as the key to her argument. Yet I contend that her position was structural and system-oriented, like Schattschneider's.
In the following section, I re-read Concept to recast Pitkin's conception of political representation in such a “realist” light. This is important for making Pitkin's classic available in a new way to contemporary empirical analysts of the relationship between economic inequality and political inequality. I contend that such analysis would benefit from reconceiving the problem that has been termed differential responsiveness in a structural rather than a mimetic way. I propose that the political power of wealth should be understood not simply in terms of manifest outcomes (e.g., greater influence and access for the wealthy, policy congruence with their preferences) but as a power to determine forms of organization that privilege wealth and lend it mobility while discouraging mobilization by those who would challenge it. To put this another way, I argue that the deficiencies of contemporary representation are best understood not in terms of the discrepancy between citizen preferences and public policy but in terms of the mobilization of bias of party institutions and electoral processes that favor some interests and disfavor others.
What Made Concept So Radical?
Pitkin made her first radical move in Concept by reconsidering the fundamental presumption that political representation can be modeled empirically “as a kind of principal-agent problem,” with the representative standing in the “kind of one-to-one, person-to-person relationship to his constituency or to each constituent in which a private representative stands to his principal.”Footnote 59 She contended that representative “outputs” cannot be fully explained exclusively (and perhaps not even primarily) by constituent demand. First, a politician has a “constituency rather than a single principal.” Second, “in legislative behavior, a great complexity and plurality of determinants are at work”; these cannot be reduced to the single axis that runs between a legislator and his or her constituency. The “legislative decision” will take into account multiple and competing factors including bargains struck with colleagues in the legislature, the agenda and interests of the party, the ideals of the nation as a whole; even the legislator's own “views and opinions.”Footnote 60
It follows that democratic representation should be conceived as a “public, institutionalized arrangement,” one where representation emerges not from “any single action by any one participant, but [from] the over-all structure and functioning of the system.”Footnote 61 Understood systemically, representative democracy need not require vertical (i.e., “one-to-one, person-to-person”) communication along the imaginary axis that is supposed to connect voters to legislators in a democratic system. Provided the system is well-designed, acts of government may be democratically representative even though “many individuals, both voters and legislators, are pursuing quite other goals.”Footnote 62
This was a bold move to cast democratic representation as systemic, impersonal, and anonymous. It led Pitkin to an even more radical one: to question the intuition that citizen preferences are “a bedrock for social choice” and a “principal force in a representative system.”Footnote 63 This Pitkin flatly refuses, referencing the survey research of the day to assert “that, at least in political representation, the represented have no will on most issues, and the duty of the representative is to do what is best for them, not what they latently want.”Footnote 64 Does this claim put Pitkin in the camp of the so-called democratic elitists of her day? Taking as their starting point the premise that the electorate is ill-informed and inattentive, they famously privileged competitive elections over policy responsiveness but also emphasized that leaders and elites would “manipulat[e] and shap[e] voters” preferences … in order to increase their chances of re-election.”Footnote 65 This was not Pitkin's view.
On Pitkin's account, although representing is both “anticipatory” and constitutive, it is necessarily reciprocal.Footnote 66 Pitkin begins by asserting “we assume that if the representative acts in the interests of his constituents, they will want what is in their interest and consequently will approve what the representative has done.”Footnote 67 She exemplifies this principle with respect to the national interest, writing that the “national unity that gives localities an interest in the welfare of the whole is not merely presupposed by representation; it is also continually re-created by the representatives' activities.”Footnote 68 Pitkin's use of the terms “merely” and “re-created” creates a contrapuntal movement that ensures reciprocity by working to check, first, a simple positivism and, then, a radical constructivism with respect to interest.Footnote 69 Pitkin acknowledges that unity is “presupposed,” recognizing that politicians present themselves as speaking for entities they have not made. But by injecting the “merely,” Pitkin denies them the pretense that this unity is simply out there to be found. Yet if its “mere” presupposition would be unearned, its flagrant disregard would be untenable. The “re-created” signals that this unity is “present in a nonliteral sense.”Footnote 70 Like Schattschneider's “bias,” it is tapped by the representatives, not invented from scratch. This is the contrapuntal movement: to refuse on the one hand to treat constituencies and their interests as given while denying on the other hand that they can be scripted by elites.
Here Pitkin anticipates what more recent scholars of representative democracy call the “entrepreneurial” aspect of political representation, which emphasizes that “‘interests’ or ‘identities’ or even ‘values’” do not “pre-exist the representative relationship” but “become objects of representation through” it.Footnote 71 As Mark E. Warren and Dario Castiglione argue, entrepreneurship casts the commonplace that democratic representation represents “constituents, either as individuals or segregated into groups” in a new light: this assumption can be held as a “normative proposition” or as “ontology.”Footnote 72
As a normative proposition, the assumption holds. Acts of representation must be justifiable to constituents as being in their interest if they are to be considered democratic. As a proposition about what ontologically is represented in such an act, this assumption is misleading. It is not the case that representation represents constituent interest in any literal sense. Constituents—“persons” or their “revealed interests, preferences, and values”—are not the “what” that is represented.Footnote 73 That “what” is created in the dynamics of the representative relationship which “select[s] for aspects of persons, by framing wants, desires, discontents, values, and judgments in ways that [make them] publicly visible, articulated in language and symbols, and thus politically salient.”Footnote 74 As Michael Saward has put it, “at the heart of the act of representing is the depicting of a constituency as this or that, as requiring this or that, as having this or that set of interests.”Footnote 75 These interests have to be “‘read in’” to the social field, politicizing it by “an active, creative process”; they cannot merely be “‘read off’” a putatively given social formation.Footnote 76 This makes representation, in Ernesto Laclau's terms, “a two-way process: a movement from represented to representative, and a correlative one from representative to represented.”Footnote 77
Pitkin's conception of representation is distinctive for approaching this dynamism, decades ago, but then—as I will argue in the next section—ultimately retreating from it. She insists that democratic representation involves one agent (the representative) acting for another (the represented) in such a way that the acts of he representative can be “ascribed” to the represented.Footnote 78 Representatives not only make decisions and commitments that those they represent are bound to honor; they also act in ways that their constituents must own. Pitkin emphasizes that the representative relationship is uniquely and distinctively reciprocal. It requires “a relative equivalence between the representative and the represented, so that the latter could conceivably have acted for himself instead.”Footnote 79 In turn, this “relative equivalence” accounts for what Pitkin calls the “fundamental dualism” of representation: it accords to both the representative and the represented the capacity to affect one another's actions.Footnote 80 In Pitkin's model, then, as András Körösényi has noted, there is “an asymmetric but still two-way or mutual relationship between leaders and citizens.”Footnote 81
This dualism makes representing unlike those activities to which it is carelessly compared: the “agent” or “actor” who acts for another; the “guardian” or “trustee” who takes care of another; the “deputy” or “attorney” who substitutes for another; those such as “delegate” or “ambassador” who are emissaries for another; and the specialist who acts independently but in the interests of another.Footnote 82 Each of these analogies likens representation to a one-sided relationship in which the actions, judgment, or expertise of one of the parties subordinates that of the other. Pitkin emphasizes that representing is a “standing for rather than in place of another,” a relationship in which “the replaced person is somehow present in his replacement or in the latter's action.”Footnote 83
This cannot apply to the deputy or guardian. Whereas the former acts not “for” but in place of the one who deputized him or her, the latter takes care of one who “is not conceived as being capable of saying anything about it or acting for himself.”Footnote 84 As for the agent or delegate, each is “a mere tool” of the other.Footnote 85 The case of the specialist is somewhat more complex because he or she is a consultant whose independent judgment informs the actions and choices of a principal. Yet even this is not representation, properly speaking. Whereas the advice and acts of a doctor or lawyer may be consequential for a patient or client, they are certainly not ascribable to that patient or client. To fall victim to a quack or a hack may kill you but it does not reflect on your good name.
For the “relative equivalence” between representative and represented to be preserved, the represented must not simply be subject to being governed according to their interests as others define them. As Pitkin affirms more than once, the represented must be “conceived as capable of action and judgment.”Footnote 86 This “conceived as” is an intriguing formulation that takes representation out of the domain of the literal. Abstract as this may sound, this is consistent with Pitkin's effort to define political representation in a way that corresponds to what political representatives actually do. In fact, they act without mandate but knowing that their actions will be ascribed to another that can refuse that ascription by voting them out of office. Pitkin quotes these words (spoken in 1961 by a senior member of the House Agricultural committee) by way of illustration: “‘I know that they would vote as I do if they had the facts that I have…. I figure if they knew what I know … they would understand my vote.’”Footnote 87 It does not matter whether this legislator is right or wrong; the point, as Pitkin sees it, is that he acts on behalf of another whom he conceives as “capable of acting and judging for himself,” and that this action takes places in a broader process in which “the represented” can and do act and judge for themselves, at least intermittently through the electoral process.Footnote 88
Miller and Stokes confirmed that legislators' conception of an active, attentive electorate is more central to a working model of representation than is the literal participation of that electorate. For, even though simple party identification (and not complex reasoning about policy preferences and the legislator's record) actually motivates most votes, legislators nonetheless “feel that they and their records are quite visible to their constituents.” Consequently, they act as if “their individual legislative actions may have considerable impact on the electorate.”Footnote 89 Citizens' policy preferences do not need to be the bedrock of government policy. Under the proper institutional conditions, representatives will imagine their constituents paying attention and potentially calling them to account.
As I noted early on, Pitkin's resistance to the literal brings us to what I have termed the “constituency paradox,” an apparent impossibility that emerges when the distinctive reciprocity of representation encounters the actual conditions of democratic politics. For, though reciprocity requires that the represented be accorded the capacity to speak and act, as Pitkin notes, the fact is that most of the time the “people know nothing” and care little about the issues that their representatives are called upon to decide.Footnote 90 This leaves them with no basis for speaking or acting in any literal sense. Thus, on Pitkin's account, acts of representation in politics—as in culture, art, and language—are “constitutive.”Footnote 91 The activity of representing is not a mirroring but an active construction that participates in making the groups for which it purports merely to stand.
This is not to say that constituencies are determined by the representative system, for if this were so there would be no paradox at all; democratic representation would merely be a contradiction in terms.Footnote 92 The paradox comes into view only when we rid ourselves of the misconception that democratic constituencies exist spontaneously. Insofar as we recognize that they emerge—or fail to emerge—as a consequence of the representation process (broadly understood as the forms of organization to which policy legacies, rules, modes of campaign finance, and other structural forces give rise), only then does contradiction become paradox.Footnote 93 And the paradox is this: democratic representatives are obligated to act in their constituents' interests, whereas constituencies and their interests come into focus only as popular identities and public policies are contested through the representative process. This paradox puts the normative conditions of democratic representation at odds with its temporal sequence; constituencies and their interests, which must be normatively prior to democratic representation, cannot temporally precede it. This gives the represented an ambiguous status in democratic representation. Neither present nor absent but “not literally present,” constituency interest exacts a deference that is normatively significant—acts of representation must be justifiable in its terms—even if that interest is not independent of but endogenous to the representative process.Footnote 94 This is the most radical insight of Concept: that democratic representation must defer to an origin to which it cannot literally lay claim.
If Concept Were So Radical, then Why Did Nothing Change?
A few of Pitkin's early readers registered the radicalism of her argument. As I have noted, Prewitt and Eulau alerted political scientists that understanding representation in Pitkin's terms would mean “re-think[ing] many of our conventional assumptions.”Footnote 95 Robert Weissberg emphasized that, following Pitkin, there would be “no historical or theoretical reason to limit analysis to dyadic representational relationships.”Footnote 96 James H. Kuklinski and John E. Stanga recognized that Pitkin called scholars to abandon their static, preference-based model of democratic representation to conceive it as dynamic, an “emergent property of a complex system of interactions among citizens and government officials.”Footnote 97 Edward N. Muller suggested indexing representation to the “expressive” and non-instrumental measure of “support” rather than to that of explicit demand satisfaction.Footnote 98 Paul E. Peterson advocated that researchers take up a deliberative approach to assess the rationale for constituent support, assessing the “quality of the reasons the representative gives.”Footnote 99
Despite their acuity, these readings had little impact on the way that Pitkin's text was received. Few readers appreciate the extent to which Pitkin called into question the orthodoxy that she is typically charged with originating. It is worth asking how a text that was so widely read can have its most important contributions so thoroughly fail to register. Academic reading habits probably had something to do with it. At least some of Pitkin's readers probably focused most closely on the two chapters of the book that advertised their relevance to debates in political science, sparing themselves the intricacies of her exhaustive survey of the analogies for representation or her critique of Edmund Burke. As this was where Pitkin did most of her unconventional re-theorization of the term, they took her concepts out of the unconventional context in which she presented them.
If Pitkin's audience can be faulted for reading selectively, I have noted that Pitkin herself did not help matters when she summed up her unconventional argument with the very term that it most challenged. It is even more curious that Pitkin presents “responsiveness” almost triumphantly as leaving “room for independence on the part of the representative but [being] incompatible … with manipulation or coercion of the public.”Footnote 100 It is on precisely this point that Pitkin runs into trouble. Given that neither the interest nor preferences of the represented can serve as a benchmark, just how it is possible to tell the difference between a properly independent leader and a manipulator?
Pitkin resolves this issue by definitional fiat. She falls back on what she has termed the “basic meaning” and “correct definition” of representation: “re-presentation, a making present again.”Footnote 101 As her italics suggest, Pitkin derives this meaning from the “etymological origins” of the word, which she reads off the prefix—the “re” in representation.Footnote 102 From this, Pitkin distills a directional protocol that makes it possible to distinguish between the properly independent leader and the coercive manipulator. She writes, “As the ‘re’ in ‘representation’ seems to suggest, and as I have argued in rejecting the fascist theory of representation, the represented must be somehow logically prior; the representative must be responsive to him rather than the other way around.”Footnote 103 Although there is a trace of Pitkin's resistance to the literal in this formulation “somehow logically prior” (as opposed to a temporal or ontological priority), it is overwhelmed by the directional protocol she imposes in the last phrase: “the representative must be responsive to [the represented] rather than the other way around.” In short, the representative is only a representative, properly speaking, when he or she respects the “etymology” of the word. Pitkin has thus fused representation conceptually to democracy, as Andrew Rehfeld has argued.Footnote 104 Consequently, her dynamic, systemic understanding of representation has become bilateral and simply unidirectional.Footnote 105
Concept is thus a text at cross-purposes. The ambiguity with which Pitkin imbues representation by what I term her resistance to the literal is contravened by her “etymological” definition of representation as “re-presentation, a making present again.” By her resistance to the literal, Pitkin signaled an important insight, that the processes of representative democracy play some part in creating the groups and interests that the prefix to the word casts as existing beforehand. Whereas this should disqualify “responsiveness” as a normative measure of representative democracy, Pitkin's insistence on etymology and definition resurrects it at its most linear and unilateral.
Just how inconsistent this was with the rest of Pitkin's text is inadvertently disclosed by Eulau and Karps in a criticism that they, ironically, aimed not at Pitkin herself but at what they counted as an oversimplification of the congruence paradigm, conjoined to a misreading of her text. They observed that the mere finding of similarity or congruence between a legislator's votes and the policy preferences of his or her constituency proves nothing about the “direction of responsiveness.”Footnote 106 Why assume that influence runs “from the represented to the representative”?Footnote 107 Moreover, any attentive reader of Pitkin should see that this ought not be: “‘Responsiveness’ and ‘response’ are not the same thing. On the contrary, a representative whose behavior is purely reactive … is the very opposite of a politically responsive person in Pitkin's sense.”Footnote 108 Their closing salvo, targeted at scholars of constituency influence, scored a direct hit on Pitkin: “a causal model of representation that draws its arrows only in recursive fashion from the represented to the representative cannot capture … the meaning of responsiveness in Pitkin's sense.”Footnote 109
Eulau and Karps were correct in principle but mistaken in fact. Certainly, “response” does not capture what political “responsiveness” must entail if—as Pitkin maintained—there is often nothing explicitly formulated to respond to. Pitkin repeatedly emphasized that political representation does not require the represented to be literally organized and agitating for an “express popular wish.”Footnote 110 It should follow, as Eulau and Karps argued, that responsiveness cannot point in only one way. The relationship is not linear but iterative; the represented rely on their representatives to “give direction to the relationship” and the representatives in their turn pursue what strikes a chord with the represented.Footnote 111 Yet Pitkin herself insisted that it be unidirectional, as if she lagged behind the moves of her own argument.
Pitkin recognized this herself. In an elusive footnote to the introduction, she reveals that she had already begun to move beyond the approach to “language and philosophy” that organized Concept.Footnote 112 She reports that Wittgenstein “so much altered my thinking about language and philosophy” that Concept “would be a different book” if she “were to write it over again now.”Footnote 113 Specifically, Pitkin was in the process of reconsidering what she had taken to be the core of a linguistic approach to conceptual conflict, the idea that definitions of concepts are like “flash-bulb photographs of [a] structure taken from different angles”: when the views conflict, the “different photographs together can be used to reconstruct” the structure “in complete detail.”Footnote 114 She has concluded “on the basis of reading Wittgenstein” that the photography metaphor is “in some respects profoundly misleading” because it belongs to an outmoded “tradition which conceives of language as reference, as our way of referring to things in the world.”Footnote 115 Pitkin affirms that a word cannot be conceived as she had treated “representation,” as a “label” for a “presence … being referred to,” a “constant” to be derived from putting its various contradictory meanings in perspective.Footnote 116 She has come to see matters from a Wittgensteinian perspective, one where words do not “stand for things” but are “used to do things.”Footnote 117 From this perspective, diversity of meaning cannot be adjudicated or ultimately resolved in terms of an “essential characteristic” shared by all members of the “class of things for which a particular word stands”; the diverse meanings are not “features” of an underlying object but actions, or moves, in different language games “for whose consequences the speaker is responsible.”Footnote 118
It is worth noting that a significant portion of the project of Concept collapses upon this admission. Recall that she had framed the book as a response to the frustration of political scientists who wanted both an operationalizable definition of representation, and a normative standard that could break out of the perpetual oscillation between mandate and independence. She has now come to emphasize how a word is used over how it is defined. Whereas she can provide an overview of its uses, doing so actually prevents her from making good on the promise to deliver the “essence” of the concept.
This collapse is fortuitous, because this idea of being held responsible for the consequences of one's linguistic moves opens up a promising alternative to the compromised notion of responsiveness. Working with the much-read section of the essay “Other Minds” where J.L. Austin interrogates the paradoxes that arise from claiming to know, Pitkin develops the norm of responsibility as a response to those paradoxes. As Pitkin explains, Austin regards claims to know as exemplary paradoxes because we make them even though we cannot literally mean them. Being human, that is, “we cannot be absolutely sure that [our knowledge claims] will never throughout all eternity turn out to be false,” despite the fact that never being false is precisely what we take this term “knowledge” to mean.Footnote 119
Following Austin, Pitkin explains that the paradox of knowing will seem merely contradictory so long as we hold a referential conception of language, supposing that “a word like “knowledge” must be a label for some (class of) phenomena” that cannot “have contradictory characteristics, to be both X and not X at the same time.” Reprising the very move she made in Concept, Pitkin recommends seeing this puzzle as arising not from what knowledge is but, rather, from the “duality of its grammar.”Footnote 120 Austin explains that saying “I know” is equivalent to saying “I promise”: it is a claim redeemed not by a congruent or matching set of facts but, rather, by the relationships it generates. Although knowledge claims are not indifferent as to facticity, they are like promises in transmitting a trust: “When I say ‘I know,’ I give others my word: I give others my authority for saying that ‘s is P’.”Footnote 121 In saying “I know” we are not referring to a cognitive feat we have already achieved; we are doing something in the saying, authorizing those to whom we speak to reiterate our claim in the first person, that is, to say “I know” in their own voice. The knowledge claim effects a transmission of authority that, Austin explains, will have “bound myself to others, and staked my reputation, in a new way.”Footnote 122 This “transmission” of authority obliges me not to be absolutely right but to take care: “if I say it lightly, I may be responsible for getting you into trouble.”Footnote 123 Claims to know are at once constative and generative: they are assertions about the state of the world that engender relationships that alter that world.
If a knowledge claim were a factual proposition first and foremost, then the way out of the paradox—claiming to know while knowing that we can be wrong (and thus that we do not know)—would be straightforward. Pitkin observes that, being infallible “we ought not, strictly speaking, to claim knowledge at all, but at most belief.”Footnote 124 But such advice misses the generativity that she, following Austin, holds to be distinctive to knowledge claims. This generativity makes them different in use to assertions of belief. Whereas belief statements merely describe “our state of mind,” knowledge claims do more. In Pitkin's suggestive formulation, “I know” opens up a “‘gap’ between what we offer in speaking and what seems ours to give”: “we offer, or claim, infallibility, yet obviously we are not infallible, and no one supposes us to be.”Footnote 125 This gap is not to be closed by an empirical proof (which would effectively reduce “I know” to an assertion of certainty); it remains open to be bridged by trust rather than fact. Pitkin affirms, in saying “I know,” the speaker takes “on responsibility for guaranteeing that the thing will never turn out to be false,” enabling the hearer to proceed on the strength of that guarantee.Footnote 126
Exactly what do I make myself responsible for in claiming to know? For Austin, I am implicated in what you go on to say by virtue of that claim. Pitkin gives Austin's notion an explicitly political twist, affirming that saying “I know” gives rise not simply to speech but to action: it grants others a “kind of freedom to act” in “a world beset with unforeseeable and uncontrollable events.”Footnote 127 The question is, does the case of knowledge claims illuminate the case of claims to represent?
In the context of a democratic political system, a claim to represent is, on my reading of Pitkin's account, similar to a claim to know. Both are anticipatory rather than retrospective. Each puts the reputation of the claimant at stake. And neither can be redeemed simply by what it lays claim to, constituencies and interests in the case of the claim to represent, literal knowledge in the claim to know. Each is proven only in the relations and actions to which it gives rise. This makes them “quasi-performative.”Footnote 128
Quasi-performativity is not Austin's term but Pitkin's, and its meaning is not entirely straightforward. She apparently coins it to separate out from those phrases that bring about a change in or by being said (i.e., performatives), phrases that “issu[e] a certain kind of guarantee.”Footnote 129 In other words, such phrases are not guarantees in themselves, as are performatives such as “I pronounce” or “I christen” or, even, “I threaten.” Instead, Pitkin writes that they perform “an action, for whose consequences the speaker is responsible,” conveying the power to act by provisionally affirming a relationship.Footnote 130 A claim to know or to (democratically) represent provisionally forges two kinds of links, that between the claimant and the group that is mobilized by the claim, and those among the individuals who, in taking up the claim, constitute themselves as a group. This gives quasi-performatives moral overtones that performatives (which can only succeed or fail) do not.
To clarify, to argue that claims to represent are “quasi-performative” is not to assert that either democratic representatives or their constituencies emerge merely in the claiming. One cannot make oneself representative of someone or something just by claiming to be; nor can one simply conjure the constituency that such a claim presupposes (either of these would be a misunderstanding of performativity in any context, linguistic or political). By adding this prefix, “quasi-,” Pitkin underlines the contingency and risk involved in claiming to represent in a democratic context. In other words, at what Saward calls the “heart” of the act of representing, lies not just “depicting … a constituency as this or that”; there is also being held responsible for that claim.Footnote 131
To be sure, there is at least one significant difference between a claim to know and a claim to represent. If the claim to know, in Pitkin's terms, grants a kind of “freedom to act,” the claim to represent takes that freedom. As Rehfeld has argued, speaking of political representation generally, a claim to represent has two folds. It is a claim to “stand in for another in order to perform a specific function.”Footnote 132 The function “describes the substantive activity a Representative is to do when ‘standing for’ the Represented.”Footnote 133 As I have noted, because hers is not a general but a specifically democratic theory of political representation, Pitkin emphasizes that the represented must be somehow “made present” in the standing for.Footnote 134 In her terms, the function must be “ascribed” to the represented so that the represented is made present in the activity but not eclipsed by it.Footnote 135
Democratic representation raises a critical question that Rehfeld's explicitly non-normative “general” theory does not (and probably need not) address: what gives rise to this function? On the conventional account, the function or substantive activities of democratic representatives should be arise out of constituent demand. I have argued that contemporary scholarship and even Pitkin's theory suggests that democratic representation quite frequently works “the other way around.”Footnote 136 The activities of representatives do not follow from constituent demand but solicit and shape it. Thus, in democratic representation, to claim “I represent x” and to exercise a specific function in light of that claim is typically not to take up a grant but, rather, to perform a function that one imputes or ascribes to a constituency in a bid for support. The representativity of such claims turns not so much on their correspondence to pre-existing demands as on the response they solicit: Do they succeed in rallying a constituency? And on what terms of identification?
This is the normative pay off of recognizing that claims to represent are quasi-performative: the idea that such claims and the functions that follow from them are not just responsive to their constituents but responsible for them. They constitute and mobilize them by the “biases” they tap: the identifications they activate, the conflicts they prioritize, and even the stereotypes they call forth. From this perspective, it is as crucial to examine how bias forecloses serious challenge to the interests of dominant groups—what Peter Bachrach and Morton S. Baratz famously termed the “dynamics of nondecisionmaking”—as it is to analyze the conflicts that the system permits.Footnote 137 The idea of responsibility lends to the normative analysis of democratic representation a new critical edge by shifting its narrow focus on assessing the match between what representatives do and what constituents say they want (i.e., responsiveness) to a much broader consideration of the role that systemic features, such as campaign finance and the legal status of broad-based mediating institutions like unions, media ownership, etc., have on fostering or foreclosing political conflict.Footnote 138 That is, responsibility redirects attention from congruence toward what I have called “constituent effects.” I propose this notion of constituent effects to open for analysis the impact that a system of representation has on the encouragement or discouragement of constituencies and on the constitution of political conflict.
Hanna Pitkin and the “Bias” of Contemporary Pluralism
The concept of constituent effects and the turn from responsiveness to responsibility is highly pertinent to the growing body of what Lawrence R. Jacobs terms “critical studies of American democracy.”Footnote 139 Refuting the notion that “politics and political power [are relatively autonomous] from economic forces and private power,” this work frames economic inequality as a political problem in the US and document its effects on democracy.Footnote 140 It provocatively identifies the contemporary US political system as a democratic oligarchy, a regime where a wealthy elite dominates policy not in “all facets of politics,” but on issues particular to “wealth defense.”Footnote 141 This scholarship advances towards but does not quite reach a structural analysis of democratic politics. It does important critical work in providing evidence of connections between economic and political inequality that pluralist ideology persistently disavows. Yet it seems to equate wealth with power, and to cast the problem with the democratic system as its “differential responsiveness” to wealthy constituents.Footnote 142 Put simply, this work remains caught up in the normative apparatus of Pitkin's Concept, judging democratic oligarchy as a failure of democratic responsiveness. I argue that this framing undercuts its normative edge, for, as I have demonstrated, responsiveness affords little purchase as either critique or remedy in a context defined by the constituency paradox.
Placing this contemporary scholarship alongside the “bias of pluralism” tradition from the radical political science of the 1960s provides one corrective.Footnote 143 The argument held that, contrary to what Schattschneider ironically termed the “pluralist heaven” (where interests are cross-cutting, organize groups are many, and power is open at multiple access points), “only a small minority of citizens, mostly from upper socioeconomic-educational brackets, participate actively in the political parties and interest groups of contemporary politics,” and that minority benefits disproportionately from its participation.Footnote 144 These kinds of arguments, notably by Bachrach and Baratz, William E. Connolly, Peter Euben, and Steven Lukes, were important precisely because they did not equate wealth with power, and did not assume that legislators are simply more responsive to wealthy constituents.Footnote 145 In fact, this work joined Pitkin's in defying the standard “behavioral” model of responsiveness as an index of the power that a constituency exercises “over” its representative by virtue of its capacity to deliver votes, its wealth, or some other quantifiable resource.Footnote 146 As a “realist” critique, it located the “bias” of pluralism not in the simple fact that interest group struggle privileged the wealthy but in the forms of political organization that intensified the political value of wealth while undercutting that of broad-based collective action. Its focus, then, was not simply on what the representative system was responsive to but on the forms of organization that it could be held responsible for.
By virtue of its focus on forms of political organization, Jacob Hacker and Paul Pierson's Winner Take All Politics belongs in the tradition of “realist” critique although, as Fred Block, Frances Fox Piven, and Brian Waddell have noted, Hacker and Pierson neither acknowledge their debts to earlier critics of pluralism nor fully take those arguments into account.Footnote 147 Their central thesis is that scholars and analysts have been so captivated by “‘politics as electoral spectacle,’” a preoccupation characterized not only by an individualistic focus on candidate personalities and the electoral horse race but, more importantly, by a misplaced belief in “election-driven change,” that they have missed “‘politics as organized combat,’” where the truly consequential victories occur.Footnote 148 This terminology, although it sets up a neat contrast, is misleading insofar as it suggests that the authors' central intent is to deflect the focus from voters and elections to interest groups.Footnote 149 Whereas the book lends itself to being read as “a modified interest group analysis that stresses the advantages of organization in winning political battles,” I think their analysis is both more original and more profound.Footnote 150 Political parties remain very much in their picture, not, however, as vehicles for the expression of voter preferences, but as “organizers” functioning within a system of constraints.Footnote 151 Thus, Hacker and Pierson date the rightward turn of US politics not to the “electoral spectacle” of 1968, which reaped the first benefits of the “southern Strategy,” the Republicans' successful bid to become the vehicle for the emergent preferences of disgruntled whites, but to the late 1970s, with the redefinition of the “organizational landscape of American politics” by the advent of money-intensive politics.Footnote 152
The conventional view would hold that Republicans' move rightward was enabled by charismatic politicians who gave the party an electoral golden goose, a populist base bought off by symbolic politics on social issues and insufficiently attentive to register the beating it was taking from Republican economic policy. In pursuit of the median voter, that story would continue, the Democrats merely followed suit, tailoring their social initiatives to a public suspicious of big government and allergic to identity politics.Footnote 153 This conventional view rests on three assumptions. The first is that election outcomes reflect the “popular will and appeal of particular leaders.”Footnote 154 The second is that parties and their candidates are vehicles for the expression of that popular will; consequently, and this is the third assumption, changes in the ideological positions of the parties reflect changes in the ideological positions of the electorate.Footnote 155 These assumptions add up not merely to an electoral-focused account of political change, as Hacker and Pierson emphasize, but to a traditional conception of democratic representation as responsiveness—grounded on constituent preferences, expressing what the constituency already knows itself to want, and involving first and foremost the legislator-constituency relationship. Thus, in breaking with politics as “electoral spectacle,” Hacker and Pierson not only break with superficial, individualistic, candidate-centered political analysis, they also take up Pitkin's most radical lines of argument about democratic representation.
In place of this candidate- and electoral-focused account of political change, Hacker and Pierson focus on the changing forms of political organization that accorded unprecedented prominence to campaign finance, to its vehicles (PACs, direct mail, and now the Internet), to its legal regulation, and to a centralized, disciplined national party committee. They explain the predominance of the Republican Party or its agenda since the late seventies in terms of this “new ecosystem,” which gave the Republican Party an early strategic advantage.Footnote 156 Not only were they traditionally closer to the business interests that gained a new centrality in this organizational form, that form also suited their party structure—a centralized organization that had traditionally prioritized fund-raising over voter mobilization.Footnote 157 Most important, Hacker and Pierson suggest, was that unlike the Democrats, Republicans could deepen their alliance with business without becoming “cross-pressured” by a disjuncture between their donor base and their voter base.Footnote 158
Having laid out the terms of this organizational transformation, Hacker and Pierson go on to direct attention to what I term its constituent effects. They emphasize that over the past generation, the Republicans have become the party of “winner-take-all” while the Democrats have become the party of “drift.”Footnote 159 As Republicans single-mindedly and uncompromisingly pursued tax cuts and financial deregulation, Democrats did nothing to ensure that policies “designed to limit inequality and economic insecurity” would keep up with the pace at which a rapidly changing economy introduced new forms of risk and vulnerability.Footnote 160 In addition, they did little to oppose, and frequently abetted, key Republican initiatives (e.g., on stock options, deregulation of the banking industry, capital gains taxation) that accelerated the pace of economic innovation and exacerbated its risks.
Why did they do this? Because competing with the Republicans in this changed organizational landscape meant a wholesale reorientation of party practices from broad-based mobilization to cultivating large donors, and a reorientation of the party's agenda from the material concerns they had pursued in partnership with labor unions to those favored by social liberals. In this context, drift was not inaction. It was an active party strategy that Democrats pursued to manage the cross pressures of the new organizational form. Drift enabled them to satisfy the preferences of their affluent contributors, and to do so without explicitly mounting policy initiatives that would provoke opposition from their traditional base.Footnote 161 In sum, the redefined landscape of money-intensive politics focused the Republican Party on an agenda—“redistributing resources through deregulation and high-end tax cuts”—that put them on the right side of Mancur Olson's “logic of collective action”: it provided “concentrated benefits to attentive constituencies” while diffusing the “costs broadly … across the public.”Footnote 162 At the same time, they were virtually ensured that Democrats would not mount a vigorous opposition, for to advocate serious economic reform would mean “going after concentrated economic interests” on which they had come to depend to keep up financially with the Republicans.Footnote 163
Hacker and Pierson tell this story to emphasize that the rightward movement of economic policy and subsequent polarization of contemporary US politics does not reflect a “growing polarization of voters.”Footnote 164 On the contrary, as Benjamin I. Page and Lawrence R. Jacobs have demonstrated, polling data reveals a widespread consensus among Americans for reducing inequality that cross-cuts party differences, and broad agreement across party lines that “taxes are necessary to fund essential government programs.”Footnote 165 Instead, Hacker and Pierson deem elite movement rightward to be symptomatic of what I term the constituent effects of the organizational shift to money-intensive politics. By reorienting the Democratic Party toward large donors, both in its agenda and in its recruitment strategies, that shift diminished the party's “capacity to reach out to voters on economic grounds.”Footnote 166 By making “drift” visible as an active political strategy of the Democratic Party, Hacker and Pierson effectively hold the Democrats responsible for the constituent effects of their strategic choices. Their analysis suggests that the Democrats are not victims of changed popular priorities but, rather, of their own failure to mobilize constituencies and frame conflict in terms of fairness.Footnote 167 As Andrea Campbell has argued, one reason that citizens have not acted as an effective “counterweight” to winner-take-all politics is that the Democratic Party and interest groups on the left are neither helping them understand the complex dynamics of upward redistribution nor helping them make the connections between those dynamics and growing inequality.Footnote 168 It follows, as Jacobs has emphasized, that research on “organizational combat” not only requires attention to “concrete and observable” instances of elite influence; it also requires “accounting for the absence of concrete action.”Footnote 169 Hacker and Pierson do not pursue this line of argument. Lacking both a concept of constituent effects and a normative vocabulary for speaking about them, they sum up their diagnosis in conventional terms, blaming the dominance of “winner-take-all” politics on the “declining responsiveness of American politicians to the electoral middle.”Footnote 170
My aim has been to derive an alternative to the responsiveness model from the work of Hanna Pitkin, one that affords greater critical purchase on the bias of contemporary pluralism. I suggest that the fundamental problem with the contemporary US system is not that it is disproportionately responsive to wealthy interests—though it has undeniably advantaged the wealthy.Footnote 171 Analysts need to step back from their emphasis on the influence that the system registers to analyze the dynamics of group- and conflict-formation. The fundamental democratic deficiency of US politics is that it has taken on an organizational form that gives wealthy interests a significant strategic advantage while at the same time making it increasingly unlikely that opposition would mobilize through electoral politics. Again, I aim to shift the normative assessment of democratic representation from the preferences to which the system responds to its constituent effects. This means paying attention to the question whether affected parties recognize themselves as such and, so, mobilize to demand a response in the first place. Various empirical scholars of inequality have begun to study the dynamics of group- and conflict-formation through such concepts as the “mobilizing and informational effects” of interest groups, the “majority-limiting features” of the US political system, and the “submerged state,” which conceals the role of the state in the upward transfer of wealth and, thereby, keeps its contestation off the political agenda.Footnote 172 This work calls attention to constituent effects insofar as it underscores a systematic bias against political mobilization to challenge inequality.
It follows that democratic representation in the US is at least as distorted by the conflicts it keeps off the table and the groups whose mobilization it forecloses as it is by the groups to which it disproportionately responds. Responsiveness, then, is not only a simplistic but even a complacent measure of democratic representation: it keeps the focus of analysis trained too narrowly on legislator-constituent relations when it needs to take in the broader systemic dynamics that affect the scope and bias of conflict. The radical strand of argument that I have pulled out of Pitkin's Concept by means of her turn to Wittgenstein affords a critical purchase on those dynamics that theories of responsiveness cannot. The concepts of responsibility and constituent effects offer a new normative vocabulary to catch up with the insights emerging from critical studies of American democracy. These concepts make it possible to argue that inequality compromises democratic representation without having to invoke constituencies and their preferences as the starting-point of democratic politics.
