Theoretical Framework

The presence of women, ethno-racial minorities and other marginalized groups in Parliament is considered by many politicians and political theorists to be an important measure of the inclusiveness of democratic political processes. But it is, as Hanna Pitkin long ago pointed out, a partial view of representation. Pitkin described four distinctive dimensions of representation: formalistic, descriptive, substantive and symbolic. She discussed each of these in detail yet left open for analysis and further clarification how to understand the relationship among the multiple dimensions of representation that we might normatively seek. One of the contributions of this article is to examine how this relationship varies contextually. As we see below, particular dimensions of the representative relationship advance and recede in importance in relation to distinctive group-related perspectives and experiences. The relationship among the various dimensions of representation is also structured more broadly by institutional features that are inherent in the Canadian political system.

In the US context, there is now a voluminous body of empirical research examining linkages among the formalistic, descriptive, substantive and symbolic representation of minority interests.Footnote ² Among the findings, American researchers have shown that black and Latino legislators are more responsive in their roll-call voting to the interests and preferences of minority constituents, and that there are substantive effects in terms of policy outcomes for those groups. With respect to more symbolic forms of representation, black and Latino voters have been shown to be more trusting of elected officials who share their racial background and more likely to vote when residing in states with a higher percentage of black or Latino lawmakers. By comparison, most of the focus in Canada has been on descriptive representation in various legislative bodies. Research here has not established strong correlations between descriptive and substantive representation, either with respect to women's interests (Tremblay, Reference Tremblay1998) or ethnic minority concerns (Bird, Reference Bird, Bird, Saalfeld and Wüst2011).

The link between descriptive and substantive representation of minority interests in Parliament has been hard to demonstrate in Canada for several reasons. First, it tends to be more difficult to measure substantive responsiveness to constituency preferences, given the strong party discipline exercised within the Canadian parliamentary system. Second, unlike in the US (or many European countries for that matter), there is not a strong undercurrent of racial resentment that tends to produce distinctive group positions on policy issues like welfare or immigration. It is therefore difficult to identify stable and readily knowable minority interests on many policy issues. Third, it is normatively unclear whether we should want descriptive representatives to be more responsive to ethnic minority concerns or should be more reassured to find that non-minority MPs representing diverse constituencies take the representation of their minority constituents seriously. An overarching question, given the leader-centred parties, strong party discipline and pronounced shift of legislative power to the executive that characterize the Canadian political system, is just how much substantive influence any individual MP (ethnic minority or otherwise) can have in shaping policy directions (Coyne, Reference Coyne2010). Finally, the accumulated evidence on constituency influence in Canada suggests that the ethnic background of an MP matters less than the ethnic composition of the riding he or she represents (Blidook, Reference Blidook2012; Eagles et al., Reference Eagles, Koop, Loat, Gidengil and Bastedo2014; Soroka et al., Reference Soroka, Penner and Blidook2009).

Given these inherent challenges and, quite likely, the relative poverty of the descriptive-substantive representative relationship in the Canadian parliamentary context, it makes sense to turn to more recent theoretical advances in understanding political representation (Castiglione and Warren, Reference Castiglione and Warren2006; Disch, Reference Disch2011; Dovi, Reference Dovi2007; Rehfeld, Reference Rehfeld2006; Saward, Reference Saward2010; Urbinati and Warren, Reference Urbinati and Warren2008). These newer approaches address a variety of shortcomings of classical representational theory that arise in the context of growing social heterogeneity, electoral volatility and the increasing complexity of decision making and manipulation of opinion formation. These factors make it increasingly difficult for citizens to reach effective policy preferences, and in turn they render citizens' interests difficult to decipher for representatives. It thus becomes more problematic to posit a clear set of political interests (on behalf of any social or political collective) that can be readily brought into the representational process and acted upon. Responding to these challenges, the newer theoretical approaches advance a more constructivist and relational conception of representation that takes into account both legislators and constituents. They also problematize the territorial and electoral basis for organizing interest representation, and posit other non-elected, non-parliamentarian actors, such as bureaucrats and civil society organizations, as effective representatives of some groups (Saward, Reference Saward2009; Weldon, Reference Weldon2002).

One element from this new approach to political representation seems especially promising for understanding ethnic minority representation in the Canadian context. Michael Saward's concept of the “representative claim” (Reference Saward2010) leads us to focus on the dynamic, constructivist and contested nature of claim making. Saward theorizes representation as a dynamic relationship of claim making (on part of the representative) and claim acceptance (on part of the represented). Related to this approach, Andrew Rehfeld (Reference Rehfeld2006) has offered a general theory of representation which simply identifies representation by reference to a relevant audience accepting a person as its representative, irrespective of the electoral district in which a person lives. In a similar vein, Lisa Disch (Reference Disch2011) has argued for a mobilization conception of democratic representation. Rather than evaluate the quality of representation in terms of legislative responsiveness to constituency preferences, this conception suggests that a representative process can be judged as more or less democratic insofar as it does more or less to mobilize both express and implicit objections from the represented. This focus on claim acceptance and contestation corresponds with Pitkin's suggestion that inquiring into the issues of representation might require one, among other things, to “concentrate on questions of social psychology,” rationalizing that a “man is represented if he feels that he is, and not if he does not” (Reference Pitkin1972: 9). Moreover, this approach shifts the focus of analysis to the perspectives of the represented.

In the Canadian context, we do not have a well-developed understanding of what it takes for ordinary men and women to “feel represented.” And we have even less insight into how ethnic minorities experience political representation, including whether they identify and expect to be represented as minorities. One of the earliest and most significant studies addressing issues of ethnic and visible minority participation and representation in Canadian politics was volume 7 of the Royal Commission on Electoral Reform and Party Financing (Megyery, Reference Megyery1996). The chapters in this volume speak to progress in descriptive representation in the House of Commons and in political parties' strategies for recruiting candidates and engaging voters from diverse ethnic groups during election time. Yet these studies address minority representation largely “from above,” whether through the perspectives of MPs, party operatives, or visible minority elites. This emphasis on elite minority representation continues with the counting after each election of MPs who can be classified as visible minority. More generally, the focus on elite perspectives of representation is reflected in the profusion of research on representative roles, constituency service and democratic responsiveness among Canadian MPs (Blidook, Reference Blidook, Gidengil and Bastedo2014; Docherty, Reference Docherty1997; Eagles, Reference Eagles1998; Eagles et al., Reference Eagles, Koop, Loat, Gidengil and Bastedo2014; Franks, Reference Franks2007; Koop, Reference Koop2012).

One of the few studies to approach representation from ordinary citizens' perspective used data from the Canadian Election Study to examine how values, political competency and social background condition representational preferences (Anderson and Goodyear-Grant, Reference Anderson and Goodyear-Grant2005). This study showed that individuals of non-European background were less likely than white respondents to support a delegate model of the representative-constituent relationship, in which representatives follow the will of their constituents; rather, they tended to prefer a trustee approach which permits the MP greater discretion in the representative and policy-making role.

Yet this finding leaves more questions than answers. Do minorities put more trust in MPs' decision making because they doubt their own knowledge and expertise in policy matters? Is it because they fear that the majority's policy preferences would prevail under the delegate model? Or is it because such individuals tend to harbour a less democratic, more authoritarian disposition?Footnote ³ Because the battery of questions on representational preferences was quite limited, it was impossible for the authors of this study to probe the reasons for such preferences. Of course, this is the main drawback to surveys. They allow researchers to roughly gauge representational preferences and their correlates; however, they are poorly adapted to addressing the complexity of political representation and for understanding how citizens think about themselves and their interests and discern good from bad representation.

The present article contributes to addressing this gap in knowledge. Approaching minority representation from the perspective of the represented is useful for several reasons. Unlike the focus on elite minority representation, it considers the nature of ethnic minority representation irrespective of the representative's ethnic background. By talking to diverse subjects, we gain considerable insight into how the various facets of representation are experienced under different contexts. We can also begin to uncover a more nuanced understanding of the reasoning and logic for accepting or rejecting representative claims. This is important information for developing better styles of representation and responsiveness and to ensure that minorities do not feel manipulated by political elites, which in turn should enhance minority engagement and sense of inclusion.

Methods and data

There are few sources of information about public beliefs regarding political representation in Canada, and none that allow us to probe the perspectives of ethno-racial minorities on this issue. I therefore turn to focus groups to provide depth and diversity to this inquiry. Focus groups provide a site for analyzing the collaborative construction of social meaning. The method is consistent with the interpretive turn in political science where the focus is on comprehension of social meaning as a prerequisite of explanation (Adcock, Reference Adcock2003; Yanow, Reference Yanow, Yanow and Schwartz-Shea2006).

Focus groups provide a different window for analysis than individually focused methods. They do not permit covariational analysis between pairs of measurable properties that are assumed to capture “objective” attributes. Whether perspectives on representation differ as a consequence of one's race, age, sex, level of education, political efficacy, settlement experience (and so on) is not the question here. Rather, scholars using interpretive methods argue that one cannot understand what any one person believes without studying the beliefs of others that are embedded in her life in various forms such as media, novels and everyday social interactions. The inherent intersubjectivity of meaning cannot be revealed by the methodological individualism of traditional survey research but rather can be (partially) discovered by exploring discourse and conversation.

Consistent with these epistemological assumptions, there are two basic aims of these focus groups. The first is to observe views about representation in a social or group context, as discussants from broadly similar ethno-racial backgrounds reflect on each other's positions, thus revealing how they experience and strive to make sense of phenomena. The second is to better understand the meaning of representation as a complex whole, which requires simultaneously taking into account its many different aspects and the relationships among them, rather than measuring it along a single dimension.

I conducted nine focus group interviews with ethno-racial minority citizens living in the Greater Toronto Area (GTA). Three groups consisted entirely of blacks, three were with people of South-Asian background, and three were with Chinese-Canadians. These groups were selected because they represent, according to Statistics Canada, the three largest visible minority communities in the GTA.Footnote ⁴ All three groups are descriptively underrepresented; nevertheless, there are differences in the numbers elected from each community. Altogether, there are 345 major electoral offices across the GTA, comprised of 253 municipal and regional seats, 46 provincial and 46 federal representatives. South Asian and Chinese representatives each hold 15 of these seats, while blacks hold four. Figure 1 presents these data on descriptive representation at the three levels, relative to each group's share of the overall GTA population. It shows that South Asians and Chinese have fared far better than blacks in achieving descriptive representation in federal politics. South Asians have also done remarkably well in achieving access to the provincial legislature, while Chinese have made relatively more breakthroughs at the municipal level. Overall, blacks in the GTA have been least well represented both in absolute numbers and relative to their share of the population. These patterns appear to be relatively persistent over time (Siemiatycki, Reference Siemiatycki2011; Siemiatycki and Matheson, Reference Siemiatycki and Matheson2005) and may be reflected in distinctive experiences and perceptions of representation across different groups.

Figure 1. Profile of Visible Minority Representation in the Greater Toronto Area, 2011

Participants in this study were recruited via an explicit invitation to partake in a group discussion on political representation with people of similar ethno-racial background. For example, the flyer recruiting participants to the black focus groups stated, “We are seeking Canadian citizens who identify as having a black, African or Caribbean background to participate in one of several group discussions regarding the quality of political representation in Canada.” These flyers were distributed to ethno-cultural organizations via social media and were also physically posted at community centres, schools, libraries, religious institutions, grocery stores and other hubs of social activity.Footnote ⁵ While each focus group comprised a single visible minority category, there was otherwise considerable heterogeneity of composition. For example, each group included an age span of at least 30 years (thus including two generations), a mix of Canadian- and foreign-born citizens (of diverse national backgrounds), a diversity of religious affiliations and a range of reported interest in politics. All groups included a balance of men and women, with the exception of one black group consisting entirely of women. One interview was conducted in Mandarin, while all others were done in English.

There are, of course, many ways to comprise groups for discussions of political representation. One concern about recruiting participants into “black,” “South Asian” and “Chinese” groups for purposes of data collection was that the research design itself could lead participants to accentuate ethno-racial aspects of their identity and to assess the quality of political representation through that lens. Various steps were taken to counterbalance and assess the social context effects that are inherent in focus groups (Hollander, Reference Hollander2004). First, respondents were asked explicitly about the need for ethnic group-based representation, thus opening the discussion to individual or non-ethnic dimensions of collective identity and interests.Footnote ⁶ Second, two groups (the Mandarin-speaking and the all-female group) were comprised to assess whether distinctive aspects of identity might become more salient in those interviews. There was a greater attention to linguistic identity among the three Chinese groups but no perceptibly greater emphasis among the Mandarin-speaking group. Nor was there any greater emphasis on gender identity among the all-female group. This suggests that, despite the artificially generated context of the focus group discussions, there is nevertheless something salient and meaningful in the relationship between ethnic identity and the understandings of representation that arise in these interviews.

The interviews themselves took the form of semi-structured conversations among five to ten participants. Each session lasted approximately 90 minutes. Following introductions, ground rules and a brief “ice-breaker” activity, participants were asked a series of questions addressing their views about representation, including whether they felt that representatives generally do a good job responding to needs in the local community and whether they trust them to represent their views when doing legislative work in government. They were asked whether it makes a difference to have people from their own ethno-racial group elected to office and were invited to discuss a number of better-known politicians, including some from their own community. At the end, they were asked to tell the rest of the group what they thought was the most important topic they had discussed (see appendix for the full interview protocol). Interviews were audio-recorded, and subsequently translated and transcribed. The resulting approximately 180 pages of text (roughly 9,600 words per interview) form the empirical basis of the analysis.

Findings

On the whole, the participants of this study were well informed and engaged in rich and wide-ranging discussions around the themes of the interview. This contradicts general findings across the literature, which point to lower levels of political knowledge, lower rates of participatory action, and higher levels of political apathy among ethno-racial minorities (Delli Carpini and Keeter, Reference Delli Carpini and Keeter1996; Soroka et al., Reference Soroka, Johnston, Banting, Banting, Courchene and Seidle2007). One possible explanation is that the individuals self-selected into this study are atypical with respect to education, political interest and civic engagement.Footnote ⁷ Another explanation is that people do possess “reasonably sophisticated, politically useful knowledge about current problems that confront them” (Graber, Reference Graber2001: 64). Indeed, many who scored low on the pre-interview measures of education and political engagement showed keen awareness of the issues discussed. Participants from diverse backgrounds also brought a world-wise dimension to these discussions. They drew on their understanding of Canadian politics, but also on comparative experiences and observations drawn from the United States or other countries of birth or prior residence. These comparisons helped them to contextualize and prioritize particular dimensions of political representation. For example, drawing on his experiences in Jamaica, Ojamu discounted the importance of a representative's racial background, arguing that the person's partisanship and ideology matter far more. “After all, we are from a black country, we are used to seeing black prime ministers.” In the following sections, I discuss the relative importance that participants in this study placed on the substantive, descriptive and symbolic elements of representation, as well as differences in representational preferences and perceptions across groups.

“If we have a problem we should be able to go meet her any time”—Contact and Constituency Service

What distinguished good representation, in the views of these discussants, lay largely outside of how elected members voted or what they contributed in the legislative arena. Rather, participants placed greater emphasis on the quality of contact and communication with their elected representatives in the community. Appreciation of this dimension of representation was amply demonstrated throughout the focus group discussions. More than half the discussants recounted having personally met a locally elected representative. Many spoke of their city councilor's, MPP's or MP's presence at various community meetings and praised them for their approachability and responsiveness to personal problems. Several told stories of how a representative had helped them or someone they knew to resolve a problem, such as a bad housing situation or the acquisition of a visa for a visiting relative. Anjela, who had lived in a public housing project in a primarily black neighbourhood, explained:

Jagvir, a discussant in one of the South Asian groups, expressed frustration in not being able to get a personal appointment with his MP, though the constituency staff did prepare the document he needed. Still, his reflection exemplifies the expectation that a representative's role is to provide local service and address problems in the community: “If we have a problem we should be able to go meet her any time.”

This finding complements elite-focused research suggesting that diverse riding electorates tend to receive more attentiveness from MPs, in terms of trips home to the constituency, the amount of time spent on constituency matters, and the percentage of MPs' budgets allocated to the constituency office (Eagles, Reference Eagles1998; Eagles et al., Reference Eagles, Koop, Loat, Gidengil and Bastedo2014). The smaller geographic size, ease of travel to and from Ottawa, and electoral competitiveness of GTA ridings are further structural characteristics that likely motivate MPs to assume a more constituency-focused legislative role (Franks, Reference Franks2007; Heitshusen et al., Reference Heitshusen, Young and Wood2005). Along similar lines, research in the US has shown that black city council members, more than their white counterparts, perceive their largely black districts to be more interested in service than issues, and respond by spending more hours and holding more meetings with constituents (Thomas, Reference Thomas1992).

Scholars have argued that MPs' increased focus on work in the constituency is a symptom of the growing centralization of executive power in Canadian Parliament. Some have questioned whether this orientation signals a fundamental weakness of representative democracy in Canada, in that MPs' role as “local fixer” has supplanted their more serious duty to initiate structural and policy reforms that might obviate the daily needs that constituency offices have come to serve (Loat and MacMillan, Reference Loat and MacMillan2014; McLeod, Reference McLeod2014). While there is validity to this argument, it is nevertheless important to recognize that minorities, in this study at least, expect considerable local contact and service, and that positive experience along this dimension appears to counteract cynicism about politics, and produce greater overall satisfaction in the quality of representation.

“Sometimes the community is wrong”—Trustee versus Delegate Conceptions of Representation

Contact and responsiveness in the local constituency is an element of substantive representation, which more generally concerns “acting in the interests of the represented, in a manner responsive to them” (Pitkin, Reference Pantoja and Segura1967: 209). But as we shift from the representative's role in the constituency to her role in the legislative arena, this notion of “acting for” the represented becomes ambiguous. A longstanding question in representative theory is whether MPs should act as delegates or trustees, that is, whether they should be bound by the wishes and opinions of their constituents, or be entrusted to do what they believe to be in their constituents' best interests. Participants in these focus group discussions were strongly oriented towards the independent trustee model of interest representation. This is consistent with survey-based findings of visible minorities in Canada (Anderson and Goodyear-Grant, Reference Anderson and Goodyear-Grant2005), and of racial minorities in the US (Carman, Reference Carman2007).

Focus group data suggest that this trustee orientation is based on three principal factors. The first is structural-institutional and relates to both the geographic size and complexity of the country as a whole and to the nature of a Westminster-style parliamentary system. For example, discussants explained that the vast size of the Canadian polity made policy responsiveness an unrealistic expectation. As Shuguang in one of the Chinese focus groups noted, “It's really hard to ask MPs to come all the way from Ottawa to their respective ridings every time the Parliament is going to vote on something. It's not realistic.” Cynthia, a participant in the same group, responded: “There are 30 million people in Canada and how do you consult those 30 million people for advice on issues? So generally, they just discuss those matters within the party. It's impossible for them to consult the general public on such matters.” Several participants recognized implicitly that the Westminister-style parliamentary system constrains the quality of representation in Canada. Ahmed, a South Asian participant, lamented, “They [MPs] care about their party's high command. They care about winning the election. They don't care about the people.” But all in his group agreed that this was a “normal” outcome of Canada's political system. As Tarek saw it, “It's just the nature of politics and power in Canada. Unlike the US where you have to face election to win your seat, in Canada the party elects you.” Linda in the Mandarin-speaking group made a similar point. “Even if a particular MPP or MP would like to address certain issues that are relevant to the voters, he may not be able to do this, if that certain issue contradicts with their party line.”

Two other sources of the trustee-style orientation appear to be central to the experiences of ethno-racial minorities living in diverse, urban environments. First was the consensus that it was generally impossible to translate the preferences of an ethnically diverse constituency into a clear mandate that a legislator could be expected to follow. As a small minority within most constituencies, blacks were especially reluctant to endorse a model of representation wherein the will of the majority would dictate legislative decision making. This is apparent in Janine's reflections on how she felt about an MP's responsibility to consult constituents prior to an important vote in the legislature.

Several other participants in the black focus groups referred explicitly to “tyranny of the majority” as a potential outcome of a delegate-style approach to representation. By contrast, some discussants in the Chinese and South Asian groups were more confident that they had the numerical and political capacity to ensure that representatives listened to their community. As Afzel, one of the South Asian participants put it, “We are a formidable vote bank.” Blacks, on the other hand, recognized that their community lacked such influence. In Adriel's words: “All the MPs care about is who is going to vote. And they don't see black people coming up to them and asking them questions.”

Second, many discussants were concerned about illiberal tendencies and lack of political knowledge within their communities, and thus were more willing to trust their MPs' judgment on legislative matters. This consideration was most apparent among the South Asian and Chinese focus groups. For example, Mei-ling, a discussant in the Mandarin-speaking group, reasoned that MPs should

Huang agreed. “Politicians are generally better informed than us since they have all the resources and information available on hand. That's why I agree … politicians should have the vision and audacity to act according to what they think is right.” Other discussants shared the view that ordinary citizens lacked the necessary depth and breadth of understanding on most topics. As Jagvir explained, “They just hear or read something about a particular issue; they don't go deep into the subject.” Consequently, he reasoned that an MP “should talk and listen to the people in the constituency, but rather than completely depending on their decisions, he should be able to decide in the national favour.”

Consensus on the trustee model of representation broke down on some issues, most notably over same-sex marriage legislation. Here a number of discussants criticized their MPs for failing to adequately consult with communities where there was strong opposition to same-sex marriage. For instance, Afzel felt that on this issue MPs had a duty to “articulate the desire and opinions of the majority of South Asians to the decision makers in the party.” Likewise, Kamran complained that MPs did a poor job of representation. “They didn't listen…I think most of them ignored the opinion of their constituents; they went ahead and did what their party directed them to do.”

Still, the vast majority of discussants felt that it is better that an MP lead than follow community preferences, particularly on moral issues and matters concerning minority rights. On the question of same-sex marriage, many participants indicated that they trusted the moral compass of their elected members over the collective judgment of the community. Jaskiran, a young woman in one of the South Asian groups, put it most bluntly. “Sometimes the community is wrong.” Abid argued that “when it is fundamental to Canada's character, an MP should vote their conscience, vote their opinion, because that constitutes leadership. They should vote their opinion because you don't vote for a polling machine, you vote for an individual with vision.” The same logic regarding an MP's moral leadership is evident in the comments of Sufei in the Mandarin-speaking group.

Michael, a discussant in one of the black groups, reasoned that politicians needed to take the lead on same-sex marriage rights, as they had in advancing civil and voting rights for blacks in the US. “There are times when the politicians lead. I think that on the question of same-sex marriage, they do a bit of educating.” Rahim recounted the leadership of a particular South Asian MP on this issue within his own community.

This statement clearly manifests the trustee orientation towards representation as described by Pitkin (Reference Pitkin1972: 152). “He is to act—not they; he is to act for them—not they for themselves.”

In summary, group discussions revealed a three-pronged logic of the trustee orientation towards representation. It is undergirded by a basic awareness of the challenges to meaningful consultation in a country as large and complex as Canada, and by some understanding of the constraints of party discipline within the Canadian political system. But this is intensified by two other factors embedded within the particular experiences of these minority citizens. One is the vast social heterogeneity of their communities and the political vulnerability that stems from one's status as a minority. The second is the expectation that elected representatives, especially at the national level, should educate and persuade newcomers and the less liberal-minded in their communities toward liberal democratic principles. Both of these factors render delegate-style representation problematic from a minority perspective. An alternative hypothesis, that a more deferential attitude toward representative government arises from an authoritarian disposition among some immigrant minorities, is not borne out in discussions among these participants.

“Don't betray our trust”—Trust and Disillusionment in Minority Representatives

The trustee model of representation, while widely accepted, presents a unique paradox with respect to the relationship between minority representatives and constituents. On the one hand, many discussants expressed feeling greater confidence and trust in candidates and representatives from a minority ethno-racial or cultural background. As Jaskiran explained, “Not that many South Asian or Muslim people, or Sikh people run for office compared to white Canadians. So just if you have the fortitude or the will power to run for office, I will be impressed.” Across all groups, discussants disclosed that while they were not predisposed to vote for a candidate just because they shared the same ethnic background, they were more likely to notice such a candidate and listen to what he or she had to say. These observations are consistent with research in the US context, showing that minority candidates have a positive impact on levels of political trust and political efficacy among those of the same racial background (Bobo and Gilliam, Reference Bobo and Gilliam1990; Tate, Reference Tate2004).

On the other hand, discussants worried about minority politicians' inclination to exploit their community's trust. This concern came to the fore in conversations about Wajid Khan, a former Liberal MP of South Asian background who had crossed the parliamentary floor to join the minority Conservative government.Footnote ⁸ Discussants unanimously condemned Khan as a representative who lacked integrity. His actions were considered a “betrayal” of his partisan supporters. “It's like if you go and buy something that is advertised in the store but when you take it home, it's a totally different thing.” They were especially concerned by his violation of the trust placed in him by his ethnic community. Yasmin, in the South Asian group, remarked: “He is frowned upon by the community—his own community and the Pakistani community in that riding. They are very disappointed.” Afzel noted that “what [Khan] did made the South Asian community trust the South Asian candidates less.” Another participant, Abid, worried that the affair had tarnished the image of the community as a whole. “Overall, what happened is all South Asians in the political area are looked at as a bunch of opportunists.”

The Wajid Khan incident suggests an inherent pitfall of the enhanced trust that tends to characterize the descriptive representative relationship. In expecting blind support from his ethnic community, Mr. Khan evidently lacked “the virtue of critical trust building” that is expected of good representatives (Dovi, Reference Dovi2007: 91). The voters promptly removed Mr. Khan from elected office. However, as these discussions suggest, such a betrayal of trust contributes to disillusionment and a broader ambivalence regarding the merits of descriptive minority representation.

Diverse Perspectives on Descriptive Representation

A key issue explored with discussants was whether it mattered to have representatives who shared their racial or ethnic background. Interestingly, the deepest ambivalence was found among the black focus groups. Blacks expressed profound disappointment in the accomplishments of Barack Obama. As Neville explained, “In the United States, the most powerful country in the world, a black president was elected and he has made no difference in a practical sense.” A number of discussants viewed descriptive representation as an illusory and potentially harmful deception. For example, Noelle argued that to look at a problem like gun violence as a “black issue” to be addressed by black MPs was problematic. She explained:

Many discussants were reluctant to agree that black representatives necessarily had the life experiences or legitimacy to speak on their behalf. They were apprehensive about the tokenistic character of black representation in Parliament and expressed doubt about the merits of those who achieved political leadership roles. Discussing the appointment to the Senate of a black pastor known for his social conservativism in addressing problems of violence among black youth, James considered that the government gained a tactical advantage in having a black representative deliver its message.

A black city councilor in the city of Toronto was likewise criticized for comments he had made justifying police practices of racial profiling. Another noteworthy black representative, former Governor-General Michaëlle Jean, elicited little enthusiasm. Many of the discussants did not recognize her, while others suggested she was more closely linked to the French-speaking Haitian community in Montreal and had little connection to Toronto's black population.

Against this critical position, other participants countered that the marginalization of black youth made it particularly important to have role models from the community in elected office. For example, Marsha argued:

Michelle added, “If you are raised and you see a diversity of people, and even black people in politics, you begin to expect that's possible for you, where if you don't see that…. Expectation can be very, very powerful.”

The black focus groups all struggled with the profound representational dilemma they saw facing their community. These groups saw the politics of presence—the idea that a group's presence enhances the possibility that its interests get substantively represented—as problematic and insufficient. They expressed concern about descriptive black representation as “window dressing” that served to evade concerted action on systemic problems facing the black community. They were distinctive from other groups in their apprehension of the neo-colonialist undertones of black representation. As Odion described it, “Our politicians are not full politicians, the way a white person can be. They come in with token issues and orders about what their role is. They are in a box.” Some remarked that the political system worked well for immigrants, especially those who brought with them economic wealth and high educational attainments, but that black needs and interests were poorly served.

Outside of the black focus groups there was less evident angst over the value of descriptive representation. Some insisted that an MP's ethnic background was unimportant. “No, I'd rather have the best qualified candidate rather than having someone of my ethnicity.” “Emotionally it looks great” was another person's response, “but practically it has zero benefits to the community.” Others were less prepared to discount descriptive representation, and thought it added some value. Many discussants felt that MPs from their ethnic community could better understand and give voice to the community's concerns. For example Ahmed argued that “[non-minority] MPs would not have the same understanding of our issues, our problems, our socio-economic challenges. So unless we have our own representatives, it is hard.” LiQing suggested that “a so-called ethnic MP cannot be responsible to represent all so-called racialized groups, but maybe can serve to counterbalance dominant interests.” Sufei in the Mandarin-speaking group suggested “an MPP or MP with a Chinese background could explain this cultural heritage to Westerners and serve as a bridge to communicate differences… he could bring some explanations to the mainstream society.”

These latter ideas are consistent with the argument advanced by Jane Mansbridge (Reference Mansbridge1999), that the virtue of the shared experience that lies at the core of descriptive representation is that it tends to enhance the quality of communication. Mansbridge argues that a more diverse parliament improves communication both “vertically,” that is between representatives and their constituents, and “horizontally” among deliberating legislators. Discussants in this study were less convinced that ethnic minority MPs exercised influence horizontally. For example, Tiffany, a participant in the English-speaking Chinese group expressed doubts on this count.

Yet for most, the quality of vertical communication between ethnic MPs and constituents was the key concern. This was particularly the case in the Mandarin-speaking group where discussants differentiated between MPs who could communicate with constituents in their native language, and those who could not. The language barrier in the latter case was seen to diminish an MP's representative capacity. By comparison, “Olivia Chow has worked hard to keep her Chinese origin, to keep the language. This makes us feel comfortable. For new immigrants, they like her a lot. People really feel closely connected.” In one of the English-speaking Chinese groups, participants discussed the responsibility of Chinese representatives to communicate to diverse constituents in multiple languages. As Qiao put it,

Another in that group agreed. “There is a much bigger burden for the ethnic MP than for the Anglo-Saxon one.”

In summary, we see substantive differences across groups in their views of descriptive representation. Among black groups, this topic generated intense debate and a fairly radical critique. This finding echoes empirical results in the US showing that black constituents place less emphasis than whites on descriptive representation (Gay, Reference Gay2002). The response in other groups was more muted. While there was some ambivalence, most discussants were supportive of having more members of their ethnic group in Parliament. Few felt that better descriptive representation would lead to different policy outcomes. Rather, most focussed on accessibility and a sense of comfort with same-ethnic elected officials and on the two-way communicative benefits that came from having an MP from one's own ethnic community. Chinese discussants, in particular, felt that the ability to engage with an MP in one's mother tongue helped to break down barriers to participation, while that MP's ability to connect with voters outside of their ethnic community helped promote cross-cultural understanding.

“It's a song-and-dance routine”—Perspectives on Symbolic Representation

How important is symbolic representation from the perspective of minorities? MPs in diverse ridings frequently use language or participation in religious or cultural ceremonies to signal “I am one of you” or to indicate their recognition and respect for a particular community. But as Pitkin explains, “for a political leader to represent symbolically, like any symbol he must be believed in.” A key question then is whether an ethnic group “either believes or does not believe in the symbolic representation enacted before it” (Pitkin, Reference Pitkin1972: 104-05). There was awareness among focus group discussants that their representatives engaged in a good deal of symbolism and performance. For example, an MP for a riding with a large Tamil population was seen as having “a song-and-dance routine” that he performed for that community: “You can see right through it. He has one routine where he brings out the Charter of Rights and Freedoms in Tamil, and he makes a big show about it. But it's not a deep understanding or articulation.” In contrast, discussants were far more impressed with Jack Layton, who had famously learned Mandarin to communicate to his Chinese constituents.

This was contrasted to Adrienne Clarkson's failure to evoke meaningful signals of attachment and shared values with the Chinese community. The former Governor General is, like Chow, a Chinese-descended immigrant from Hong Kong. Yet, as one discussant remarked, Clarkson “really tries to make it invisible, as though she is not Chinese at all. So there is a balance between how much you play it up and how much you don't play it up. I am suspicious of those who play it down a lot, rather than those who are just who they are.”

A discussion in one of the South Asian groups turned spontaneously to the government's apology for the historic Komagata Maru incident, and debated the meaningfulness of this performative act.Footnote ⁹ For various reasons, discussants did not accept that the government's apology signalled a genuine acknowledgement of this historic wrong. One in this group asked rhetorically, “And that's an important gesture?” Another admonished: “Don't forget, words are cheap.” Most considered that the apology was issued as part of a broader bid to gain support for the Conservative party among the South Asian community. One discussant argued that the government had undermined the credibility of its apology by delivering it at a largely political event in British Columbia, rather than from the floor of the House of Commons, where it would be inscribed as part of the nation's historical record.

These discussions reveal subjects to be astute adjudicators of symbolic and performative representations. They focused on deciphering and deconstructing representative claims to distinguish what was meaningful from what was manipulative. Discussions thus shed new light on the symbolic dimension of representation. Quantitative research in this vein has focused on whether minority citizens experience higher trust and political efficacy when represented by a member who shares their ethnic background. However, such a static and reductive operationalization fails to capture this element of representation on at least two levels. First, leaders from historically marginalized minorities may have the potential to evoke deeper faith and trust in government among group members. But they may equally provoke suspicions about strategic manoeuvring to emphasize or downplay their group identity for political gain (Collet, Reference Collet2008). Likewise, leaders from the ethnic majority can incite in minorities a greater sense of trust and inclusion, for example by speaking and listening to them in their native language. Or they may be derided for their tokenistic oversimplification of complex communities of identity and interests. Properly understood, the symbolic dimension of representation thus concerns any MP's ability to behave in a way that is sincere and meaningful to a particular community.

Second, it is clear that symbolic representation can have both negative and positive implications. Pitkin was largely concerned with the former. Writing as she was in the post war era, she saw symbolic representation as closely connected with fascist leadership techniques, as “a power relation, that of the leader's power over his followers” (1967: 108). Certainly representatives do engage symbols in an effort to placate their constituent minorities, yet this very act also opens up a space for deconstructive critique and potential resistance. Indeed, we cannot understand symbolic representation without taking into account of this kind of reflexive mobilization and engagement by group members in response to representative claims by various actors (Disch, Reference Disch2011: 104). This is precisely what South-Asian discussants were doing in assessing the government's apology for the Komagata Maru incident, and what the Colour of Poverty activists were doing in critiquing the political instrumentalization of their communities by the Conservative party.

Summary and Conclusions

This article has looked in detail at perspectives on various dimensions of representation among three visible minority groups within the diverse, urban environment of the GTA. A central finding is that members of minority groups have contextually derived preferences regarding political representation, and that the dimensions that are important and that incite feelings of being represented vary. It is thus problematic to focus the investigation of political representation on any single dimension; rather, citizens' assessment of the quality of their representation is to be found at the nexus of the multiple dimensions of representation that come together in complex configurations. In some contexts, formalistic representation (having elected the representative and being able to hold him or her to account in the next election) is essential to citizens' feelings of being represented. In other contexts, particularly where language is an issue, having a descriptive representative who shares one's ethno-cultural identity matters most. In other instances, symbolic representation involving credible signals of recognition and respect by representatives to the represented incites stronger feelings of being represented. The relative importance placed on these various dimensions of the representative relationship from one instance to the next depends upon how that relationship is constituted by and within institutions, political conflicts and collective actions (Castiglione and Warren, Reference Castiglione and Warren2006: 11). In the context of this study, we can understand minority citizens' perspectives on representation as constituted by the rules and norms that define representative roles within the Canadian political system, and by the historical experiences, collective resources and political capacity of particular ethnic groups. Further, these perspectives appear to vary in relation to the local geo-spatial context. Finally, while they are not explored in this study, differences in representational preferences may also be related to individual factors such as one's level of education and political sophistication.

Turning to the concrete findings on minorities' perspectives of political representation, participants in this study were not particularly concerned with policy responsiveness in the legislative arena. They placed greater emphasis on the quality of local service, contact and communication with their representatives, and demonstrated relative satisfaction with this aspect compared to other facets of the representative relationship. Furthermore, participants did not expect their MPs to act as delegates who follow the will of the constituents in legislative matters. This is the case despite choosing their representatives via the candidate-centred single member district system and despite visible minorities comprising a numerical majority in many GTA constituencies, both factors that would be expected to promote a more constituency-based focus of interest representation. This finding is consistent with other survey-based findings on the representative orientation of visible minorities. However, the focus group discussions help to identify the reasoning and logic that underlies this expectation of representation. It is rooted in minorities' recognition of the intrinsic impossibility of identifying any clear mandate for legislative action on behalf of socially diverse constituencies. For blacks, in particular, it stems from their sense of vulnerability to majority interests. A “liberal expectancy” perspective toward less informed or less liberal-minded members of their communities was also an important source of the trustee orientation, as many felt that MPs in particular play a vital role in integrating newcomers into a liberal-democratic value system.

While a more descriptively inclusive City Hall and House of Commons were seen as generally desirable, opinions diverged on what practical difference it made to have more members of one's ethno-racial group in elected office. Some saw benefits in terms of reducing barriers of contact and communication between MPs and their constituents, and enhancing understanding across diverse groups. This was especially true among Chinese participants, who were most concerned about linguistic barriers to communication and political participation. Many expressed more inherent trust and interest in listening to what a same-ethnic candidate had to say. The exception was black discussants, who demonstrated considerable apprehension about the tokenistic and colonial undertones of black representation, and who questioned the possibility of achieving political equality within existing institutions and power relations. By comparison, Chinese and South Asians were less cynical about politicians and more satisfied with the performance of both descriptive and non-descriptive representatives.

Discussants recognized the pitfalls of placing too much trust in an ethnically descriptive MP. However, the downside of the trust relationship appears to be mitigated by a community's capacity to sanction wayward representatives at the polls (an aspect of formal representation). South Asian discussants felt generally confident in this respect; however, blacks were less certain about their capacity to elect and hold accountable such representatives. Furthermore, while an appointed black Senator or Governor-General might provide an important role model (descriptive representation), such actors were viewed as less accessible and less capable of providing the contact, service and two-way communication that are such a vital part of the representative relationship (an aspect of substantive responsiveness).

This study also suggests that the importance of symbolic representation should not be dismissed. While Pitkin voiced concern about the manipulative aspects of symbolic representation, recent theoretical advances around representative claims making have shed new light on this dimension. Meanwhile, emerging empirical research suggests that symbolic representation may be key to voter mobilization among some marginalized groups (Bastedo, Reference Bastedo2012). Despite the negative gloss on symbolic representation, there are many examples in Canada where symbols have been used to signal recognition and respect, and initiate meaningful dialogue amongst diverse groups. The symbolism of an officially bilingual parliament has been a key element of linguistic minority representation since 1867. More recently, one of the most meaningful acts of minority representation was that of Elijah Harper, the Cree politician who inspired Canadian Aboriginals by blocking the Meech Lake Constitutional Accord while clutching an eagle feather in the Manitoba legislature. In the Aboriginal community, this symbolic representation of a Canada with more than one “distinct society” is viewed as a turning point in asserting and upholding First Nations rights and recognition. As one admirer recently put it, “Elijah Harper was the first public figure who made me proud to be indigenous.”Footnote ¹⁰ In her study of national apologies, Alia Somani similarly argues that while symbolic performances may be intended to placate minorities, it is possible that “with enough conviction, we can ‘blast open the continuum of history’… and find within apologies a sign of Messianic hope, redemption, and possibility” (Reference Somani2011: 7).

At a broader level, this study has produced new insights that complement existing empirical scholarship on visible minority representation in Canada in at least three ways. First, the findings go beyond rudimentary, primordialist approaches based on counting visible minority MPs, MPPs and city councilors, and show that non-descriptive dimensions of the representative relationship may be more important from the perspective of minority citizens. Second, the study contributes to understanding representation from below, thus supplementing empirical analyses that have been largely elite driven, concentrating solely on MPs' views of their representational roles. Bringing both top-down and bottom-up perspectives to bear on the empirical study of representation is in line with newer approaches in representational theory that argue that representational roles are ultimately relational, constituted by both the self-understanding of the representative and by the expectations of those being represented (Castiglione and Warren, Reference Castiglione and Warren2006: 11). Third, the findings can inform particular strategies to enhance the quality of minority representation in Canada. Within the GTA, it appears that blacks have distinctively negative evaluations of the way they are represented, along with low trust and belief in the overall quality of representative democracy. Addressing the history of racism in US context, political theorist Melissa Williams argues that black distrust of political institutions can be “at least partially mended if the disadvantaged group is represented by its own members” (Reference Williams1998: 14). Yet empirical findings from the US show that the perceived benefits of descriptive representation among black constituents are by no means straightforward (Gay, Reference Gay2002; Swain, Reference Swain2006); nor is there much evidence in the present study that blacks in the GTA will feel better served by having more black representatives. Improvements in descriptive representation may not instil trust and can even be counterproductive under some circumstances, for example, where minority representatives are appointed, where the community has little capacity to hold them accountable, or where there is little opportunity for contact and two-way communication between constituents and legislators. Appointing more black Senators will probably not improve the quality of representation in the eyes of the black community, nor is it likely that a more proportional, list-based electoral system, such as the mixed-member proportional (MMP) method proposed in Ontario in 2007, would cause minorities to feel better represented. Such an electoral system may enhance descriptive representation via a more ethnically diverse slate of party candidates, but at the cost of diminished accessibility and accountability of those representatives to the community. Increasing the number of representatives, enlarging constituency budgets, and staffing constituency offices with employees from diverse backgrounds who can interact with residents in a variety of languages may do more to strengthen access, trust and engagement in the political system. Direct strategies to boost civic and electoral participation among smaller and more disadvantaged minorities are also crucial.

While this study has sought to deepen our understanding of how particular minorities feel about the various facets of their political representation, many questions remain. Future research should investigate more systematically how contextual features (such as having a different- or same-ethnicity MP, or living in a majority-minority riding) influence representational preferences. For example, do Chinese residents in the riding of Markham-Unionville, where they comprise some 40 per cent of the population, expect more responsiveness to constituents' direct mandates? Do their representational expectations of Michael Chan, the Hong Kong-born Liberal MPP for the provincial riding since 2007, differ from their expectations of John McCallum, who has held the federal riding for the Liberals since 2000? What kinds of representative claims do such politicians make, whether to the large Chinese and South Asian communities in their ridings or to the general population? And how do constituents respond to those claims? How should we understand the mobilizing responses of various communities when the central party plays its hand in selecting a particular kind of descriptive representative as the local candidate in such ridings? Addressing the issue of intersectionality, we might also investigate differences in representative claim-making and claim-acceptance on the part of male and female, minority and non-minority MPs. Research should also examine how minority citizens' representational preferences vary in relation to individual-level factors (such as gender, age, class, immigrant background, strength of group belonging, level of education, political sophistication, and so on). Some of these questions can be subjected to quantifiable analysis via linear additive regression models and a variety of simplifying assumptions. However, the contextual variability of representational preferences also calls for further qualitative exploration, thick description and closer “soaking and poking” into the complex relational dynamics of minority representation in different settings. Such research is vitally important to understanding and improving the quality of representative democracy in Canada, and for ensuring that minorities feel included rather than manipulated by the political system.