This article should be read as an ongoing dialogue between Suzanne Dovi and ourselves about a common concern: the quality of representation in general and, in particular, the good substantive representation for women (SRW). We strongly share Dovi’s concern that democratic institutions and processes can favor those in positions of power and can be used to dominate and oppress. We also are persuaded that for democracy to function well, a specific type of representative is required (Dovi Reference Dovi2007). The key difference between us is that Dovi’s focus (2002; 2007) is on the individual representative’s characteristics and qualities, whereas we turn our focus to the level of representative processes. Representation is a process of advocacy and deliberation taking place within and outside of formal political institutions, where differences in political perspectives are advocated for and deliberated over (Saward Reference Saward2010; Urbinati Reference Urbinati2000; Williams Reference Williams1998). SRW in formal institutions such as parliaments—we argue in this contribution—should meet specific “quality-control” criteria. We defend our preferred conception of good representation as procedural but, as we show herein, Dovi’s values of the good representative well may be important prerequisites for the good processes that we envisage.
Our claim for a shift away from the actors (i.e., women/feminist Members of Parliament) and content of SRW (i.e., legislative and policy outcomes) toward a focus on the processes of representation stems from two sources. First, we must fully acknowledge that women are a highly diverse group with varying and even conflicting interests. Second, we should reject an elision between SRW and feminist substantive representation. In our view, good SRW does not occur when the interests of only a limited group of women are represented. Neither does it occur when only a specific feminist understanding of gender equality is articulated. Instead, good representational processes take seriously the heterogeneity of women’s interests while accepting that not all women share feminist ideals.
Hence, we judge SRW processes using three criteria. The responsiveness criterion assesses the representative claims that are evidently in play and judges them in terms of their representative relationship with the represented—that is, the extent to which they are responsive to women in society. Following Severs (Reference Severs2010), responsiveness is established when either representatives make claims congruent with prior stated interests of the represented or the represented a posteriori agree with the claims that representatives formulate on their behalf. The inclusiveness criterion addresses the representational processes as a whole and is particularly concerned with representative claims that might be excluded. The inclusion of all relevant voices is necessary to establish the meaning of what is in women’s interests and to counter within-group inequalities (Weldon Reference Weldon2002). The third criterion points to the relative status of representatives’ claims. It demands more than the mere ability to equally articulate one’s interests; voices should receive equal respect and consideration and be able to generate an effect (Severs Reference Severs2012).
Dovi (Reference Dovi2015) is critical of our procedural approach. It risks, in her view, being too inclusive and therefore ultimately agnostic about outcomes. Dovi is correct to state that there can be no privileging of feminist outcomes in our approach, as there can be in her exclusive-content–based approach. However, we are far from agnostic about outcomes. Indeed, our preference for a procedural approach has everything to do with what we see as an unworkable conception of what counts as “anti-women” that would serve as the basis for exclusion. Dovi’s content approach returns us to the place from which we feel gender and politics scholars needed to retreat: a subjective, a priori definition of women’s interests.
That said, Dovi’s critique prompted us to think further about the requirements of democratic representatives and to reconsider whether our preferred conception of good SRW as a responsive, inclusive, and egalitarian process depends on the nature of individual representatives who populate our democratic institutions. As our discussion of Dovi’s three values of the good representative demonstrates, we fully recognize the virtue in her good representatives. We could not agree more that a representative must be fair-minded and engage in critical trust building and good gatekeeping. Indeed, we contend that the presence of representatives who embody these values is likely critical for good SRW processes. We outline how our preferred institutional process is likely interdependent with representatives embodying particular qualities and characteristics (Celis, Childs, and Curtin Reference Celis, Childs and Curtin2016). We also importantly contend that these qualities are more realizable collectively than individually. Nevertheless, the institutional implementation of our ideal process requires subsequent “institutional experimentation” (Williams Reference Williams1998, 236). The criteria we outline here enable the evaluation of existing arrangements, processes, and procedures (Celis, Childs, and Curtin Reference Celis, Childs and Curtin2016) and should inform institutional reform that seeks to enhance the quality of SRW.
THE VIRTUE OF FAIR-MINDEDNESS
For Dovi, a fair-minded representative advances public policies that foster civic equality: the equal political standing of citizens. Accordingly, these representatives advance not only the policy preferences of their constituents but also the civic equality of all (Dovi Reference Dovi2007, 101). Only by seeking to increase civic equality can the legitimacy of democratic institutions be safeguarded. Such legitimacy is necessary for citizens to seek to resolve their conflicts through democratic institutions. Citizens consider democratic institutions to be fair when there is equality in public policy making. All of this implies that a fair-minded representative mediates and accommodates disagreements among citizens rather than represents particular interests (Dovi Reference Dovi2007, 118–19).
Fair-mindedness, although evidently about content, is for Dovi also about how outcomes are produced. Policies should seek to “reduce and compensate for inequalities of political resources” (Dovi Reference Dovi2007, 111). Citizens’ capacities “to interact as equals” should be protected and “shared political status and standing” secured (Dovi Reference Dovi2007, 105), not least by “supplementing the political resources available to the worst-off citizens” (Dovi Reference Dovi2007, 113). Concerned by citizens’ access to decision makers, a fair-minded representative reaches out to those who hitherto have been marginalized by political processes.
Like Dovi, what we conceive of as the ideal representative process considers conflict about policy preferences and requires granting equal and fair consideration to divergent interests.
It perhaps goes without saying that our preferred SRW processes would profit from these fair-minded representatives. Like Dovi, what we conceive of as the ideal representative process considers conflict about policy preferences and requires granting equal and fair consideration to divergent interests. We could not agree more with Dovi’s requirement for good representatives to mediate and accommodate disagreement among citizens. However, there exist points of differential emphasis and ongoing concerns. First, what constitutes civic equality too easily can be presented as a given (Dovi Reference Dovi2007, 108–19). We are concerned—as Dovi is—that what constitutes civic equality must be more than formal inclusion as part of the demos. She argues that formal, threshold gap approaches should be balanced by a good democratic representative (Dovi Reference Dovi2007, 116–17). In our view—and arguably to a greater extent than acknowledged by Dovi—the equal political standing of women is realized only during the very processes of political representation. This is precisely because what counts as civic equality on the ground is often open to debate, as she acknowledges. Hence, any tendency to conceive of it in a generalized way, or to use it as a standard, seems rather inadequate to the task of securing good SRW.
Second, Dovi makes a trade-off between “advocacy representation = political efficacy” and “fair-minded representation of all = democratic efficacy,” in which the latter is privileged. She holds that democratic efficacy refers to the “effect” of a policy on civic equality rather than a simple “obeying the policy preferences” of the represented or the majority (Dovi Reference Dovi2007, 104). In our view, the absence of strong advocacy representation may very well undermine both political and democratic efficacy. Moreover, in our approach, the apparent trade-off dissolves. The advocacy for particular interests should include the interests of all affected, and the subsequent deliberation should be inclusive of all these interests. This results in both political and democratic efficacy: representatives are establishing partial political efficacy for their own group (in our case, women) and, as a collective, also establishing democratic efficacy.
A high level of responsiveness is another important benefit and a key ingredient of good representation, engendered by our process approach and constrained by Dovi’s fair-mindedness. Responsiveness is about having one’s interests represented in a focused way by one’s representative. That said, once representatives have articulated the interests of their constituency, they should—as part of the subsequent deliberation—listen to one another and give equal consideration to all opinions. This is when democratic efficacy returns. When what is “in the interests of women” is being constituted (Urbinati Reference Urbinati2000; Weldon Reference Weldon2002), representatives should be willing to let go of their original opinion. Overall, then, the representative process of deliberation should deliver democratic efficacy.
The third distinction we draw is that Dovi is foremost concerned with citizens’ equal access to representatives, whereas we foremost are concerned with equal inclusion of citizens’ claims in the representative process. As we understand it, Dovi sees a unique role for elected representatives in that respect. A fair-minded elected representative informs herself about the multiple and conflicting interests and opinions that exist in the citizenry and makes up her mind about what best serves civic equality. She then represents this view in the representation process. In our approach—in contrast to Mendez’s contribution to this symposium—the claims of the represented do not have to “pass through” representatives to be included in the representative process. Our criterion of inclusiveness also would allow for non-elected representatives to advocate for their interests as part of the representation process. We further contend that the value of fair-mindedness works differently for the non-elected (cf. Urbinati Reference Urbinati2000): as advocates, their “right” to be partial would be self-evident. Furthermore, recognizing that strong fair-mindedness is not demanded of all representatives shifts the responsibility for enacting fair-mindedness to the aggregate level: good SRW requires guarantees at the system level that the process of representation is inclusive of all and that all claims are treated with equal consideration by the collective of representatives brought together in the representational process. In our view, it is predominantly through such a fair-minded process of representation that disagreement among representatives about what is in the interests of the represented can and should be mediated.
THE VIRTUE OF CRITICAL TRUST BUILDING
The second virtue of a “good representative” refers to the representative’s capacity to engage democratic citizens in representation in a way that increases their critical trust. How to foster this? Acknowledging that invitations to participate can be “manipulative” rather than a source of “self-governance,” good representatives must be sufficiently informed of the costs and benefits of particular policies, must be kept accountable by peer or horizontal accountability, and must constitute good role models (Dovi Reference Dovi2007). When these conditions are fulfilled, citizens can assess when their active participation is needed and when, in contrast, they can rely on the democratic institutions to settle political conflicts (Dovi Reference Dovi2007, 134). However, even under optimal conditions, vulnerable groups of citizens might lack the capacity or the material resources to participate in the representative processes, thereby remaining dependent on potentially abusive others to interpret their interests (Dovi Reference Dovi2007, 142).
The value to our approach of Dovi’s concept of critical trust building lies in its likely effect in increasing responsiveness and inclusiveness. The two-way exchange of information between represented and representatives should increase the former; the envisaged participation of non-elected fosters the latter. We also share Dovi’s concern that vulnerable groups still may be limited in their participation. Although we accept that our process approach does not have a full solution to this problem, our commitment to the egalitarianism criterion goes some way to address this, augmenting Dovi’s critical trust-building virtue. Dovi also is highly convincing in describing the qualities, form, and prerequisites of good information and participation. But our approach extends her analysis. We conceive of our approach as having—in effect—operationalized critical trust building. Our criteria of responsiveness, inclusiveness, and egalitarianism work to identify the group or groups (in our case, different groups of women) affected by a particular issue: those whose participation is required. Our preference for a process approach further requires that representatives engage in information giving and responsive-related dialogue with affected groups. These groups themselves may be present in the process of representation. A process approach also guides when and how to hold fellow representatives to account.
THE VALUE OF GOOD GATEKEEPING
The value of good gatekeeping judges representatives by “the company they keep” (Dovi Reference Dovi2007, 145). A good democratic representative’s main purpose is to promote the political inclusion of all democratic citizens (Dovi Reference Dovi2007, 147). To this end, she must expand her relationships beyond her own political base and foster mutual relationships with all democratic citizens (Dovi Reference Dovi2007, 161–2): political opponents and the marginalized. In short, a good democratic representative must actively reach out and seek to obtain full insight in the various and conflicting interests at stake. She is neither delegate nor trustee (Dovi Reference Dovi2007, 149–50); rather, the relationship should be characterized by mutual recognition and trust (Dovi Reference Dovi2007, 152–3). In all of this, citizens should recognize that their fates are linked with fellow citizens (Dovi Reference Dovi2007, 147) and representatives have a role to play: helping citizens to shape and consolidate their particular identities by identifying and articulating their interests; promoting their identification with the representative (thereby increasing ownership of and responsibility for the actions of the representative); and binding them to their democratic institution (through interaction with their representatives, citizens can better understand how democratic institutions safeguard their well-being) (Dovi Reference Dovi2007, 155–9).
Dovi’s third value again speaks strongly to our criteria for good SRW: it cannot be established except by reaching out and ensuring that all relevant voices are heard and listened to. For underrepresented groups, a good democratic representative not only opens the door for their legitimate claims (Dovi Reference Dovi2007, 170) but also acknowledges the complexity of power and oppression. She will “limit and constrain the influence of those that exclude unjustly” (Dovi Reference Dovi2007, 148, 171).
Echoing concerns regarding fair-mindedness, the represented—especially dispossessed and marginalized groups—will likely benefit from representatives who advocate exclusively for their concerns.
In our reading, however, there are qualifications to the value of good gatekeeping. Imagine a representative of the oppressed who sits at the table with the oppressors. Will the oppressed trust that representative? We think that the chances are greater when the representative is more exclusively devoted to the interests of the oppressed. Echoing concerns regarding fair-mindedness, the represented—especially dispossessed and marginalized groups—will likely benefit from representatives who advocate exclusively for their concerns. (See related discussions on the respectability problem by Strolovitch Reference Strolovitch2008 and on the cultural brokerage by Severs and de Jong in this symposium.) Such a representative will likely enhance trust on behalf of the represented; mutual relationships between the represented and the representative will deepen. Such relationships of trust are necessary for representatives to do what Dovi (Reference Dovi2007) requires of them.
The fact that (women’s) interests are not pre-given but rather fully constituted in the course of the representation process through engagement—confrontation, even—with the views of others underpins our second concern with Dovi’s gatekeeping role. Vertical and “bilateral” relationships between individual representatives and those they claim to represent are arguably rather limited in this respect. Individual representatives may not be able to establish the necessary array of mutual relationships with those affected by an issue (Mendez’s contribution to this symposium is a case in point). Hence, our emphasis on inclusion at the system level: the door should be open to all relevant voices. Furthermore, in any case, it is during the process of representation that individual representatives come together as a collectivity to ultimately identify what is in the interests of all—in our case, what is in the interests of women.
CONCLUSION
As this discussion demonstrates, we see much virtue in the values of fair-mindedness, critical trust building, and good gatekeeping for processes of good SRW. Although we agree with Dovi that these are, for the most part, important at the individual level, we contend that they are more realizable by representatives collectively. At the level of representative processes, our qualifications to Dovi’s values and the tensions between them also are somewhat mitigated. For example, in the case of the former, when individual representatives do not privilege the interests of the dispossessed and marginalized but rather seek on an individual basis the representation of all, good outcomes might be more difficult to achieve, and critical trust may well be lost for the represented—especially if they are marginalized groups. Regarding the latter, shifting to the collective level enables resultant conflicts among the three virtues to be both more transparent and capable of being “worked through.” To wit, imagine our processes of representation, characterized by inclusion, responsiveness, and egalitarianism, populated with representatives collectively embodying Dovi’s good representative values.
