Introduction The decennial redrawing of federal electoral boundaries in Canada is a contentious and challenging exercise involving the delicate balancing of often contradictory representational goals. Primary among these have been community of interest and population equality, and the challenge of reconciling these principles has figured prominently in the litigation that has grown out of the process of adjusting electoral boundaries in Canada (Roach 1991, 1992; Knight 1999). However, other specifically geographic factors such as the compactness and contiguity of electoral districts, have also been included among the various desiderata of good electoral districts (Morrill 1981; Lijphart 1990; Butler and Cain 1992). In striking contrast to the American case, to date these geographic issues have attracted neither scholarly nor juridical attention in Canada. In this first systematic, countrywide assessment of the achievement of district compactness in recent federal electoral maps, we aim to redress the scholarly neglect of this issue. In addition, we contend that, in one form or another, such questions will likely come before the courts in future electoral boundaries litigation. The paper provides a baseline assessment of patterns in the compactness of the last two sets of Canadian federal electoral districts (FEDs) adopted after the 1981 and 1991 censuses (as defined in Representation Orders adopted by the House of Commons in 1987 and 1996, respectively). These maps were the first to be drawn following the introduction of the Charter of Rights and Freedoms. As a result, they are the first Canadian maps to have been scrutinized through the process of judicial review. Since the first judicial decisions with respect to electoral boundary practices were rendered in the late 1980s and early 1990s, we would expect that constitutional concerns would be more present in the minds of boundary commissioners in the 1990s. As such, a comparison of the two maps offers an exploratory assessment of the impact of the Charter on the shape of Canadian electoral districts. If the American experience (which we review briefly below) is any guide, we expect that efforts to balance equality of population with other representational goals should have the effect of diminishing the compactness of districts over time. Our analysis begins in the next section with a preliminary discussion of why compactness should matter in districting practice. We then consider the increased significance attached to questions of district shape in the US, and assess the prospects for similar developments arising in the Canadian context. Here we review several factors that incline us to believe that assessments of compactness may emerge in future rounds of boundary-related litigation that are virtually certain to follow the adoption of the new electoral maps. A third section discusses the conceptualization and measurement of compactness. This provides the basis for an empirical exploration of the compactness of the last two sets of Canada's FEDs that we present in the paper's fourth section. Here we document significant interprovincial differences in the compactness of FEDs that in the main seem attributable to the different landscapes of the provinces. With respect to our hypothesis about the decline of compactness scores over the two maps, however, we do not uncover strong support for this expectation, in that the mean scores on compactness for all districts generally or for urban districts in particular show little change. We close with some projections for the next map that will be forthcoming after the 2001 census figures become available. Why Should Compactness Matter? Discussions of redistricting virtually always include compactness as one of the criteria to be considered when evaluating electoral maps. Good electoral maps are those in which district boundaries define highly compact districts. Yet, as we outline below, assessing compactness is not a simple or straightforward process. …
Read full abstract