open Secondary menu

Electoral Insight – Readjustment of Federal Electoral Boundaries

Electoral Insight – October 2002

Community of Interest and Minority Representation: The Dilemma Facing Electoral Boundaries Commissions

Community of Interest and Minority Representation: The Dilemma Facing Electoral Boundaries Commissions

Professor Jennifer Smith
Department of Political Science
Dalhousie University

How is it possible to square the principle of voter equality with the consideration of community of interest in drawing electoral boundaries?

Everyone understands voter equality. It is the idea that one person's vote has the same value as the next person's. Community of interest, on the other hand, is an elastic concept. Economic interests; an historical identity; a group identity in a way of life; traditional practices – these are among the many ways that the concept has been used to justify departures from voter equality. Obviously, the concept lacks defined scope. All communities have interests, especially economic interests, and even the newest will eventually generate a history. A more precise, weighty and well-accepted meaning of community of interest needs to be developed. However, there is currently no such development. If anything, more ideas are being added to the mix – in particular, minority representation, which it is my purpose to discuss.

Background Rules

The Canadian Constitution requires that the distribution of seats in the House of Commons be governed by the "proportionate Representation of the Provinces": each province is to receive a share of seats in proportion to its share of the population as a whole. The principle of representation by population is based on the idea of equality, which logically points to the desirability, if not the necessity, of constituencies of comparable voter numbers. Nevertheless, the logical inference of voter equality is not the same as a clear endorsement of it, and in the revision of electoral boundaries Canadian legislators have not pursued voter equality with single-minded enthusiasm. Instead, they have been inclined to focus on accommodating countervailing factors like community of interest and geography, which is why the standard legislative scheme employs the concept of variance.

Under the variance model, as it is termed here, the legislature fixes the size of permissible departures from voter equality, and then outlines the factors that might justify such departures. Thus Parliament enjoins federal boundaries commissions to pursue the objective of equal population size per constituency, while at the same time requiring them to consider non-population factors, like community of interest, community history and geography. The commissions are permitted to accommodate these factors within a variance of 25 percent above and below the average constituency population (the electoral quotient), except in undefined "extraordinary" circumstances. In such "extraordinary" cases, there is no limit on the extent to which the population of a constituency can depart from the average. Since 25 percent above and below the average in fact constitutes a 50 percent range, the federal model clearly offers considerable scope for the accommodation of non-population considerations.

Boundaries commissions serve two masters, judicial as well as legislative. The courts have defined the meaning of the right to vote as requiring a "relative equality of voting power". In Dixon, the Supreme Court of British Columbia concluded that, while the idea of equality is central to the Canadian conception of voting rights, the idea of absolute or exact equality is not. Deviations from voter equality are permissible, declared Chief Justice McLachlin (as she then was), but "only those deviations should be admitted which can be justified on the ground that they contribute to better government of the populace as a whole, giving due weight to regional issues within the populace and geographic factors within the territory governed."Footnote 1 In her next boundaries decision (Carter), Justice McLachlin, now writing for the Supreme Court of Canada, elaborated her concept of the right to vote by defining it in terms of the purpose of "effective representation". "It is my conclusion," she wrote, "that the purpose of the right to vote enshrined in s.3 of the Charter is not equality of voting power per se, but the right to 'effective representation'."Footnote 2 And the primary condition of effective representation, she continued, is relative parity of voting power, modified where necessary by factors like geography, community history, community interests and minority representation.Footnote 3

The concept of community of interest not being a particularly exacting one, mostly it is used to buttress arguments in favour of leaving boundaries where they are. Not that this is illogical. Leaving boundaries where they are confirms the "community" as it has become, willy-nilly, over time. It is a typically English, typically Burkean idea. By contrast, the idea of boundaries that mark out group identity implies sociological representation. This is new and different and deserves debate, which occurred in 1994 in the House of Commons Standing Committee on Procedure and House Affairs. The committee considered alternative ways to define community of interest: the conservative way being existing boundaries, natural boundaries, geographic features and transportation and communications concerns; the broader way including sociological factors, too. In the end, it took the conservative route, on the grounds that an explicit reference to sociological factors would compel the establishment of minority districts and thereby promote an illiberal politics.Footnote 4

Although the committee's report was shelved, the exercise of seeking to define community of interest demonstrated how significant the concept really is. Undefined, it is a repository for everyone's hopes and dreams and a nightmare for boundaries commissioners. How can the valid use of such a concept be tested in connection with any one map? In other words, what constitutes adequate justification for invoking the community-of-interest claim?

Electoral Districts After Redistribution - Canada 308

Justifying Departures from Voter Parity in Favour of Community of Interest

For boundaries commissions, the lesson to draw from Charter jurisprudence on electoral boundaries is the need for justification. Individuals or groups who are dissatisfied with the way that commissioners draw a particular electoral map can contest it in court. As a result of this development, the reasoning behind decisions about boundaries revisions, previously submerged in the minds of the commissioners, has been forced to materialize in the clear light of day. Or it has been concocted – by someone.

The issue of justification is proving to be somewhat troublesome. So far, it appears that the courts’ insistence on justification has put an end to the older practice of electoral maps that conform only to custom and partisan calculation and to no external standards at all. Presumably there will be no repetition of the MacKinnonFootnote 5 case in Prince Edward Island in 1993, in which the provincial government found itself constructing arguments to defend an electoral map that had remained largely unchanged for one hundred years. But there remains the habit of some boundaries commissions of supplying no written reasons for their decisions. The Alberta map at issue in 1993 in Reference re Electoral Divisions Statutes Amendment Act was accompanied by very little in the way of written justification, prompting the appeal court to write: "[W]e simply do not know, with precision, why the many variations were permitted. In this case, what was needed was a riding-by-riding justification. It has not been offered."Footnote 6

In addition to the problem of no justification, there is the problem of adequate justification, which is especially difficult when an all-purpose phrase like community of interest is used. Finally, there is the issue of justifying community-of-interest departures from voter equality that fall within a stipulated variance like the +/- 25 percent rule. Is anything within the variance valid on its face? Fortunately, there is further jurisprudence in P.E.I. on that point. The province's century-old electoral map, mentioned above, was felled easily in MacKinnon. The province's subsequent offering was drafted by the legislature, which used the variance of +/- 25 percent of the electoral quotient per district. The legislation accompanying the map prescribes the use of a boundaries commission for future revisions and instructs the commission to take into account the usual aspects in reviewing boundaries, namely, geography, population patterns, community of interest, municipal boundaries and anything else it deems relevant. So the model in place for future revisions is a version of the federal model.Footnote 7 Meanwhile, the new map authored by the legislature was unaccompanied by a report explaining and defending what was done. When Charlottetown and some nearby smaller communities contested the map in court, the government's counsel had to generate a rationale in response to the contention of the applicants that the 27 electoral districts, which ranged from +21.10 percent above the average population to -19.28 percent below it, failed to achieve the standard of relative voter parity; that the districts over-represented the rural areas of the province to the detriment of the urban areas; and that they failed to conform to the new boundaries of Charlottetown and Summerside, described as communities of interest.Footnote 8

The idea that the two cities were communities of interest was a throwaway. The real beef of the applicants was that the map failed the test of voter parity because it over-represented rural areas at the expense of urban ones. The government conceded the rural over-representation, and then presented an array of arguments to defend it, including a novel community-of-interest claim in connection with rural Kings County, citing statistical indicators to demonstrate that the county is a distinct and disadvantaged region of the province. However, Chief Justice MacDonald would have none of it.

The irony is that Chief Justice MacDonald upheld the map, while accepting none of the arguments offered to defend the very rural over-representation that caused most of the population discrepancies among the districts in the first place. He queried P.E.I.'s decision to choose +/- 25 percent and asked if a lower variance might not be more appropriate – say +/- 10 percent. It is not necessary to explore his reasons for landing on this figure. The important point is what came next. His idea of a valid reason for departing significantly from voter equality arises in his discussion of compactness. The government had argued that districts should be compact, and that compactness requires some latitude with the numbers. Chief Justice MacDonald countered that compactness is an ideal that ought not to be pursued at the expense of voter equality – except, however, in the district of Evangeline—Miscouche, in which much of the Island's Acadian community resides. On the existing map, the Acadian voters are combined with a smaller number of English-speaking voters, and he thought this unnecessary. "[T]here is no reason," he wrote, "why this district could not be given special status and allowed to go below the 25 percent variance. By doing so, the Acadian population would have their exclusive district and a minority English population could be attached to an English district."Footnote 9 Even the applicants, he noted, were prepared to give Evangeline—Miscouche dispensation from the principle of voter parity.

It is worth sorting out what happened here, especially since the decision was upheld on appeal, the majority of the appeal court taking essentially the same position as the lower court.Footnote 10 First, the explanations for the departures from voter parity were put together after the fact – again – and the courts did not find them convincing. Second, the courts signalled clearly that a minority cultural community that is geographically compact can fill the community-of-interest bill. In this model, the minority is a community of interest and as such constitutes a valid reason to depart from the standard of voter equality. However, there is another, different model in effect, too, which takes us to Nova Scotia.

Legislative Definition of Community of Interest

In Nova Scotia, the legislature has issued instructions about the accommodation of minorities in the drawing of electoral boundaries. Since Parliament issues no such instructions, and instead leaves the federal commissioners to judge the issue of community of interest, it might be supposed that the Nova Scotia case has no relevance for the federal arena. Yet it does, because of the way in which the legislature has flagged minority representation – thereby supplying a precedent for all to see – and conceptualized it.

In 1991, a select committee of Nova Scotia's House of Assembly set the terms of reference for the province's first independent boundaries commission. In its report, issued after the Carter decision, the select committee affirmed that relative parity of voting power in the form of constituencies of equal population is a "primary" factor of "paramount importance" in ensuring the right to effective representation, and that departures from voter parity are permissible only to the extent that they can be shown to contribute to the better governance of the population as a whole. It also listed five other primary factors that might be involved in effective representation: geography; community history; community interests; projections of the rate of population growth; and minority representationFootnote 11. Singling out minority representation, it instructed the boundaries commission to pursue the "advice, support and co-operation" of representatives of the African-Canadian community, the Acadian community and the Mi'kmaq people. Finally, the committee declined to establish a variance rule, emphasizing instead that the commission was "not to be governed by a predetermined population deviation factor or by a predetermined split between urban and rural ridings".Footnote 12

Since the legislature accepted the report of the committee in its entirety, the boundaries commission was faced with the challenge of producing a provincial electoral map that met the legislature's requirements regarding minority representation, as well as its stricture on the paramount importance of relative parity of voting power. In response, the commission devised an "entitlement" model to be used in conjunction with a concept of the "protected constituency". The details of both the model and the concept are in the commission's report.Footnote 13 Briefly, the entitlement model results in a map based entirely on the principle of equality. It shows the number of constituencies to which any area of the province is "entitled", in terms of its population size. Having determined the kind of map that the factor of population alone would generate, the commission turned to the non-population factors in the terms of reference.

As has already been noted, the legislature pressed the representational cause of specified minority communities. The commission needed to judge (1) which areas of the province, if any, are distinguished by communities of interest and history, or minority communities or geographic isolation, and (2) the extent to which the population in these areas requires protection in the form of discrete and undersized electoral districts or, indeed, can be protected in that way. Such judgments are not always easy to make. Moreover, if members of a particular community are scattered far and wide, it is difficult to accommodate their representational needs within the confines of a territorially based electoral system. The situation of Nova Scotia's Black community is a case in point, since members of the community reside throughout the province. However, a significant number live in a series of small communities in the Preston area, which is where the commission chose to establish a protected constituency. Although the population of the new constituency was only half the average population of the non-protected constituencies, the Black communities still comprised only one third of it. Thus the constituency could not be described accurately as a designated "Black seat". On the other hand, the literature on the competition between the candidates of political parties in single-member constituencies suggests that a percentage in this range is enough to provide members of a minority community with a reasonable prospect of electing a member from among themselves. That prospect materialized when the Preston constituency elected the first Black member to the House of Assembly in the election using the new electoral map.

There remains the question of how protected constituencies are tied to the entitlement model. Altogether, the boundaries commission established five protected constituencies: three in Acadian areas of the province, one in the Preston area, and one in a sparsely populated part of rural Cape Breton. Since the total number of seats in the legislature was fixed at 52, the decision to depart from relative voter parity in five constituencies meant that the population average or population entitlement for the rest of the constituencies would necessarily change. In an effort to be absolutely clear about this consequence, the commission decided to omit the protected constituencies entirely from the calculation of the population entitlement of the remaining 47. As a result, the adjusted entitlement for them was about five percent higher than it would have been in the absence of protected constituencies. The five percent difference was the "penalty", as it were, that residents of the 47 constituencies paid for the effective representation of the residents in the smaller-sized constituencies. It was a penalty that the legislature had determined to be appropriate and publicly acceptable – and still does.

The province is undergoing a review of electoral boundaries at this time. In November 2001, a select committee of the legislature reviewed the terms of reference of the boundaries commission and decided to retain the emphasis on minority representation. As before, it ranked voter equality as the first of the primary factors of representation, followed by geography, community history, community interests, and minority representation, in particular, representation of the Acadian and Black peoples of the province. However, its model now more closely approximates the federal model, with a variance rule of +/- 25 percent. It also links minority representation to the notion of extraordinary circumstances: "The Provincial Boundaries Commission is to be governed by the general principle that a constituency should not deviate by greater or lesser than 25 percent from the average number of electors per constituency, except in extraordinary circumstances. Extraordinary circumstances are the desire to promote minority representation by Nova Scotia's Acadian and Black communities."Footnote 14


The experience of the two Atlantic provinces leaves federal boundaries commissioners with two points to ponder about community of interest. The first is the weight of the precedents that are being set. For a second time, the Nova Scotia legislature has issued clear instructions on minority representation that the boundaries commission once again can be expected to implement. Then there is the reasoning of the courts in P.E.I. that minority representation is a clear example of a valid claim of community of interest.

The second point is the development of competing ways of conceptualizing minority representation. One way is the idea of a minority as a community of interest in and of itself, an idea played out in the P.E.I. courts, where some judges suggested that the minority's community of interest was the only valid reason for departures from voter parity. Just how valid, is apparent in the comments of the dissenting judge on the appellate court, who found no good reasons for the province's electoral districts to be so unequal in population, except for Evangeline—Miscouche. It is the type of district, he wrote, for which "there is no limit to the amount of deviation allowed provided it is justified as necessary for more effective representation."Footnote 15 The other way of conceptualizing minority representation is the Nova Scotia select committee's categorization of minority representation as an extraordinary circumstance. This recalls the federal rule that enables boundaries commissions to move past the variance of +/- 25 percent in extraordinary circumstances and the British Columbia rule that permits the same thing in "special" circumstances.

Are these models analytically distinct from one another? Arguably, yes. If minority representation is conceived in terms of the concept of community of interest, then it must compete with all other notions of community of interest. However, I suspect that it would compete very well with these other notions simply because the claim of cohesiveness in a significant dimension of life – cultural identity – seems more compelling than, say, patterns of transportation. On the other hand, if minority representation is thought to be a special or extraordinary circumstance, then it moves to a different conceptual plane. It becomes a separate category altogether and, by implication, an unusual one. Given the terms of reference under which they operate, including the undefined concept of community of interest, federal boundaries commissioners can go either way in assessing a claim of minority representation – as a matter of community of interest or extraordinary circumstance. One thing is certain – the claim of minority representation is not going away.


Footnote 1 Dixon v. British Columbia (A.G.) (1989) 59 D.L.R. (4th), 267. It is important to emphasize that Justice McLachlin included the idea of justifiable deviations as part of her concept of the right to vote under section 3 of the Canadian Charter of Rights and Freedoms – justifiable deviations are not separate from the definition of the concept. This is apparent in her discussion of the relationship between section 3 and section 1, the opening clause of the Charter, which stipulates that governments may impose reasonable limits on rights if they can show that the limits are "demonstrably justified in a free and democratic society". She wrote "that every citizen is entitled to an equal vote, subject only to such exceptions as may be justified on regional or geographic grounds as providing better government. The question under s. 1 is whether breach of that standard can be justified on the basis that the breach is reasonable and demonstrably justified in a free and democratic society. In other words, can population discrepancies incapable of being supported on the above grounds be upheld under s. 1 of the Charter?" Ibid, 270.

Footnote 2 Reference Re Provincial Electoral Boundaries (Sask.) (1991) 81 D.L.R. (4th), 35.

Footnote 3 Ibid, 39.

Footnote 4 House of Commons Committee on Procedure and House Affairs, Report, 25 November 1994, 33: 5-11.

Footnote 5 MacKinnon v. Prince Edward Island (1993), 101 D.L.R. (4th), 362.

Footnote 6 Reference re: Electoral Divisions Statutes Amendment Act, 1993 (Alberta) (1994), 119 D.L.R. (4th), 14.

Footnote 7 Electoral Boundaries Readjustment Act, Stats. P.E.I. 1994, c. 13.

Footnote 8 Charlottetown (City) et al. v. Prince Edward Island et al. (1996), 150 Nfld. & P.E.I. R., 98-99.

Footnote 9 Ibid, 101.

Footnote 10 Charlottetown (City) et al. v. Prince Edward Island et al. (1998), 169 Nfld. & P.E.I. R. 188; 521 A.P.R. 188. The appeal court's decision in turn was appealed to the Supreme Court of Canada, which declined to hear the appeal.

Footnote 11 The commission pursued the idea of a Mi'kmaq seat as instructed by the legislature, consulting with Mi'kmaq leaders about it. There was no consensus then among the Mi'kmaq about the appropriate way to fashion such a seat. However, the legislature chose to recognize a 53rd seat belonging to the Mi'kmaq, thereby issuing a standing invitation to them to take it up.

Footnote 12 Ibid, 7.

Footnote 13 Nova Scotia, Provincial Electoral Boundaries Commission, Effective Political Representation in Nova Scotia: The 1992 Report of the Provincial Electoral Boundaries Commission (March 1992), 20-30.

Footnote 14 Provincial Electoral Boundaries Commission, Information Package, March 2002.

Footnote 15 Charlottetown, 1998, 213.


The opinions expressed are those of the authors; they do not necessarily reflect those of the Chief Electoral Officer of Canada.