TWO APPROACHES: INTEREST AND PROCEDURAL

The two main approaches to evaluating good representation adopt different units of analysis. The interest approach (Burke Reference Burke1790 [1968]; Sapiro Reference Sapiro1981) orients evaluations around the activities of individual representatives by measuring them against a particular understanding of the represented’s interests. This approach assumes that the represented have a coherent and stable set of interests, and it downplays the different and conflicting preferences of various subgroups within the represented.

In contrast, the procedural approach (Disch Reference Disch2011; Saward Reference Saward2010) centers evaluations on collective representative processes. This approach focuses on the institutional setting in which particular claims are constructed. It understands good representation to be as whatever results from “good” representative processes. The procedural approach is intentionally “outcome blind.” I discuss each approach in turn.

The interest approach assumes that one cannot determine whether a representative is a good representative for a group unless one first identifies that group’s interests. The substantive content of the represented’s interests dictates what a good representative should do. Therefore, if equal pay is in women’s interests, then representatives for women should advance and pass equal-pay legislation. In this way, the interest approach assesses individual representatives by what they do.

Of course, representatives sometimes do not do what the represented wants. In these circumstances, Pitkin (Reference Pitkin1967) recommended that a representative explain to the represented why the representative’s understanding of interests is better. Although Pitkin recommended preserving the autonomy of both the representative and the represented, she settled conflicts between them by asking for verbal justifications and persuasive interventions. The language of interests, therefore, remains the theoretical currency in which representatives negotiate conflicts with the represented.

Of course, notions of interests can vary greatly. Celis et al. (Reference Celis, Childs, Kantola and Krook2008) argued persuasively that the concept of “women’s interests” has no agreed-on definition. The heterogeneity of social locations within any represented group means that its members are likely to have different, if not conflicting, interests. Therefore, an interest approach often can lead to irreconcilable disagreements about the represented’s proper interests.

Notice that even when disagreements between the represented and the representative are explicitly recognized, the authority to decide whose preferences will be advanced remains squarely with the representative. However, the possibility that representatives can and often do abuse that authority should give us pause: deliberations about interests are likely to favor those present, and those present in representative processes are likely to be more privileged members of groups (i.e., the wealthy and the healthy). Consequently, how representative processes identify and adjudicate among interests is likely to be biased in favor of the perspectives and priorities of existing representatives.^{Footnote 3}

This leads to the second approach to evaluating representation—namely, the procedural approach. According to this approach, good representation arises when collective representative processes have been structured properly. Instead of evaluating individual representatives to some a priori notion of interests, political science is tasked with assessing the collective processes in which political claims are constructed. For instance, Saward (Reference Saward2010) argued that collective representative processes should include diverse interests and be conducive to contestation.

At first glance, the standards of diversity and contestation might seem uncontroversial. However, people can disagree about the value and meaning of contestation and inclusion just as they can disagree about the meaning of interests. “Open” representative processes can increase the vulnerability of those who are targets of political violence. Contestation may be desirable as long as the represented are not punished for disagreeing with their representatives’ claims. The procedural approach, therefore, does not avoid the disagreements that justified giving up on using “interests” as a standard of evaluation.^{Footnote 4}

Moreover, the insistence of the procedural approach that evaluations of good representation be “outcome blind” is worrisome. By focusing on how claims are constructed rather than implemented and enforced, this approach ignores how outcomes can cumulatively undermine the fairness of representative processes and the autonomy of the represented. Losing a particular policy battle can prohibit raising concerns about those policies in the future. Representatives can weaken the civil and political rights of the represented in response to resistance (e.g., undermine their right to protest peacefully).

Besides, the claims of historically disadvantaged groups are not necessarily completely missing from representative processes; rather, such claims can be made and systemically ignored. The “mobilization of bias” within existing representative institutions can prevent such claims from impacting outcomes. Even when representative processes are structured correctly (whatever that means), resource-rich groups still can have disproportionate influence, epistemic standing, and effective bargaining power that resource-poor groups do not. Strolovich (Reference Strolovitch2008) persuasively argued that political organizations for historically disadvantaged groups often systemically underrepresent groups with multiple oppressions. Htun and Weldon (Reference Htun and Weldon2010) maintained that representatives from historically disadvantaged groups often must align with dominant majorities to achieve certain policy goals. Diversity, equality, and contestation are certainly laudable ideals for representative processes. However, evaluating representative processes according to these standards can obscure how alliances with dominant groups determine who wins certain political conflicts.

The bias of the procedural approach in favor of privileged groups is particularly problematic because it treats the quality of the relationship between an individual representative and the represented as irrelevant to the overall quality of representation. Good representation entails examining not only the deliberative processes necessary for authorizing legislation at any given moment but also the capacity of the represented to hold their representatives accountable and to minimize the harm incurred by certain policies.^{Footnote 5} It matters whether the representative listens (or dismisses) the challenges of the represented. It also matters whether the representative can punish the represented for disagreeing with their claims. Their relationship reflects the extent to which representatives are incentivized to respond to the represented’s concerns.

Understanding good representation as simply the outcome of individual representatives’ activities or as a procedural feature of representative processes does not adequately recognize the vulnerability of the represented to the power of existing representatives. Interest approaches downplay how representatives ultimately decide whose interests are advanced. Procedural approaches wrongly assume that the appropriate audiences can effectively challenge their representatives’ claims. To address both of these problems, I propose a blended approach, what I call the autonomy approach.

THE AUTONOMY APPROACH

The autonomy approach asserts that good representatives must actively preserve the capacity of the represented to challenge and reject their representatives’ claims. Therefore, having open town meetings as well as private meetings with their constituents is critical for hearing dissenting views of a representative’s claims. The nature of the interactions between the representative and the represented is not extraneous to evaluations of representation. Good representation is not simply how representatives construct their claims in concert with other representatives but also how representatives interact with and impact the represented.

Because representative processes inevitably distribute disadvantages unequally,^{Footnote 6} good representatives should structure “the choice landscape” so that the represented can better negotiate their vulnerabilities. Ben-Porath (Reference Ben-Porath2010b) convincingly argued that the state can structure citizens’ choices in ways that support the fact of pluralism while simultaneously advancing particular policy objectives. Building on Ben-Porath’s discussion, my approach holds that good representatives structure choices to reduce the vulnerability of the represented to the coercive authority of the state.

Instead of denying that a notion of interests underlies conceptions of fair, equal, and inclusive procedures, or assuming that the represented have a unified set of interests, the autonomy approach focuses on whether the represented can mitigate or even circumvent the vulnerability produced by their representatives’ claims. Good representatives institutionally empower represented individuals to have the final say on how policies affect their life choices. Thus, good representation is tied directly to the autonomy of the represented.

To be sure, Saward (Reference Saward2010) explicitly stated that the represented must have the ultimate say in claim construction. However, Saward never adequately elaborated on how that say should be institutionally supported. One contribution of this article is its suggestion that good representatives must institutionally empower the represented to resist the claims of their representatives. One way to accomplish that is to institute opt-out options that can alleviate the vulnerability of the represented. These opt-out options aim to sustain the social and economic resources needed for full participation in democratic governance and thereby their political autonomy.

To illustrate how an autonomy approach could work, consider the question: “How should good representatives ‘represent women’ when women possess conflicting and mutually exclusive opinions about the burqa?”^{Footnote 7} Deep disagreements about the burqa suggest that whatever a representative does will go against the expressed preferences of some women. Whether the representative votes for banning or permitting the burqa, the representative will fail to act as a good representative for some women. Instead of assessing good representation by whether the processes for deliberating the burqa were fair, or trying to evaluate whether a burqa is in women’s interests in some “objective” universalist sense, the autonomy approach focuses on reducing the women’s vulnerabilities that occur from either banning or permitting burqas.

After all, to permit the burqa can leave women who do not wish to wear it vulnerable to coercion from other group members. Hence, the autonomy approach would evaluate good representatives by whether they provide adequate support services to women who refuse to wear the burqa and who want to exit their communities.^{Footnote 8} Supporting these opt-out options allows women to become more autonomous because it supports the decisions of those who do not wish to wear the burqa. These women might need effective restraining orders or even witness-protection programs to ensure their full participation in society. Good representation requires that policy outcomes support the capacity of the represented to become full political members, thereby securing their political autonomy.

Similarly, good representatives who support a ban on burqas must provide opt-out options that reduce the vulnerability of women who continue to wear them. After all, an undesirable consequence of banning the burqa is to further isolate vulnerable women. Banning the burqa means that some women will be unable to work, travel, or obtain an education. The autonomy approach maintains that it is not enough that such a decision was made under procedurally fair conditions; it requires mitigating the adverse effects of these policies. The autonomy approach would demand that good representatives subsidize or fund schools that permit the burqa to alleviate the economic and social vulnerability of women who wish to wear it. The autonomy approach demands addressing the vulnerability that can accompany the claims of representatives.

Focusing only on whether the deliberations of representative processes are fair ignores how policy outcomes impact the opportunities of the represented to fully participate and thereby decide how they wish to live their lives. The capacity to challenge claims is not independent from the choice landscape fostered by representative processes: policy “outputs” can constrain the ability of the represented to dissent within their communities, as well as challenge their representatives’ claims in formal representative processes. Thus, fostering autonomy becomes an important prerequisite for properly enacting the virtues of good representation.

The autonomy approach thus favors providing opt-out options that counteract the negative consequences arising from representatives’ claims and that thereby expand the choices available to vulnerable groups. In this way, the autonomy approach discourages good representatives from implementing “one-size-fits-all” public policies. Good representation does not simply voice dissenting opinions during deliberations but rather incorporates the insights about vulnerability into policy outcomes. Good representation should not be judged exclusively by how representatives construct “claims” but also by whether their activities sustain unjust marginalization (Dovi Reference Dovi2009).

My discussion of the autonomy approach is clearly meant to be suggestive, not exhaustive. Its main claim is that assessments of good representation should focus on how representative processes adversely impact the ability of the represented to resist and challenge their representatives’ claims. Assessments of good representation must recognize the importance of institutionally sustaining the autonomy of the represented.