By David Zylberberg
Following the census, Canada’s federal electoral districts are redrawn every decade. On Monday, Ontario’s proposed new ridings were announced, the last province to do so. You can look at the details of the proposed new ridings or the process of consultation, here. The proposed changes have led me to think about the origins and rationale for electoral districts. In particular, I will be discussing the importance of communities of interest for designing effective ridings in our particular system and how these priorities are reflected in proposed ridings for Saskatchewan and northeastern Ontario.
Canada is one of only two countries to use the Westminster system of government with first-past-the-post balloting. The English House of Commons developed medievally as an elected body of taxpayers with the authority to approve all new taxation. Not all men had the vote and ballots were not secret, but there were geographically specific constituencies. An essential principle of this system was that each region or town had unique interests and got a say in the taxes being levied upon it.
Two types of constituencies existed: counties and boroughs (aside from graduates of Oxford and Cambridge Universities, who until 1945 elected representatives in addition to their local votes). Boroughs were towns, often on the coast, that at one point had been important enough to be given their own courts and parliamentary representation. Neither counties nor boroughs were divided on the principle of an equal number of voters in each district.
Constituencies were not redistricted between 1660 and 1822. In 1832, significant reforms were brought into the parliamentary system in order to reduce corruption. Following the Reform Bill, a number of very small constituencies were abolished and new ones created in growing manufacturing towns, including Manchester and Sheffield. Stephen Thompson wrote a very good piece about the role of the 1831 census in that redistribution. Despite the redistribution of seats, there was no still no suggestion that constituencies should have an equal numbers of voters. Every county and town had unique concerns and was entitled to parliamentary representation to address them. Manchester’s 227,808 residents did not merit greater representation than Southampton’s 19,324. It was not until 1945, that constituencies were designed on the principle of having comparable populations.
Out of the British tradition, Canadian ridings have always been based upon the notion that each electoral district has important economic, social and infrastructure interests that need to be represented in Parliament. This is one reason that Canada has not adopted proportional representation and why ridings are designed around ‘communities of interest’ where feasible. With Canada’s vast distances, geographic, cultural and linguistic diversity, regional representation is essential. However well intentioned, officials in Toronto or Ottawa will rarely understand the culture or concerns of places like Terrace Bay or Oyen. Proportional representation assumes that the divisions inside communities are greater than those between them. As well, party lists in such a system disproportionately include residents of major centres. That is partially why many Canadians disagree with such a system and why central Toronto was home to the only ridings to support it in Ontario’s 2007 referendum.
Canadian redistricting is done by committees of non-partisan officials appointed by the provincial chief justice. Canada has generally avoided the problems of explicit partisan redistricting found in the United States. The current Canadian redistricting uses the 2011 Census and designs ridings with the intention that ridings in each province be of similar size to each other, within an acceptable standard of variation. However, officials are instructed to also:
- respect communities of interest or identity (for example, communities based around language or shared culture and history)
- respect historical patterns of previous electoral boundaries, or
- maintain a manageable geographic size for districts in sparsely populated, rural or northern regions of the province
Most of the coverage of the current redistricting in Canada has focused on the thirty proposed new ridings. These were proposed so that parliamentary representation would better represent population numbers, and the boundaries of these new ridings do not appear controversial. Aside from South Cowichan-Juan de Fuca, all of the proposed new ridings are in the country’s six major cities or their suburbs. These have been the main areas of population growth in the last decade.
However, the more interesting decisions involve redrawing riding boundaries to reflect population changes in regions that were not awarded new seats. While changes are proposed in every province except Prince Edward Island, I will be discussing the proposed new ridings in Saskatchewan and northeastern Ontario. The changes in Newfoundland are also very significant and, in having the rural ridings reflect the highways, should create constituencies that are easier to campaign in and represent, along with a greater community of interest.
Since the 1960s, electoral boundaries in Saskatchewan have been designed so that each of the two cities was divided into quadrants. This resulted in eight mixed urban-rural ridings with one quadrant of the city, large swaths of farmland and some small towns over one hundred kilometres from the city. With the growing cultural and economic polarization between rural and urban Canada, such ridings no longer constitute a community of interest. Under the proposed redistribution, Saskatoon will have three seats and Regina 2, with one rural-urban riding that encompasses north-east Regina and the Qu’Appelle valley. By separating the province into rural and urban ridings, MPs should be better able to address and advocate the specific urban concerns of Saskatoon residents and the rural ones of Rosetown.
Changing from rural-urban to separate rural and urban ridings will affect the seats each party wins in Saskatchewan and herein lies some controversy. Currently, New Democratic support in the province is concentrated in the cities and despite getting 32 percent of the province’s votes in the 2011 election, the NDP did not win any seats. There is some irony here, as the rural-urban seat model was adopted at a time when NDP support was still concentrated in rural Saskatchewan.
Northeastern Ontario is probably the region that presents the greatest problems for designing electoral districts. The distances are large and the population small, but Sault Ste Marie and Greater Sudbury are large enough for their own ridings. One difficulty is that the populations of Sudbury, North Bay, Timmins and Sault Ste Marie have remained steady while the smaller mining and lumber towns have been hemorrhaging people since the 1970s. The greater difficulty relates to creating six coherent ridings, since the natural communities of interest create five ridings that include the four cities. In the last redistribution, putting all of the areas that did not naturally fit in the five communities of interest created the most geographically- incoherent riding in Canada, Algoma-Manitoulin-Kapuskasing. The riding has two main areas of population along the north shore of Lake Huron and Highway 11, but these are unconnected by roads. One consists of two Francophone mill towns of ten thousand people while the other region is Anglophone and includes the uranium-mining turned retirement town of Elliott Lake and Manitoulin Island, with its small family farms and cottage economy. Putting Kapuskasing and Hearst in Timmins-James Bay as part of the 2012 redistribution will give them communities of interest, and create ridings that are easier for campaigning and for the Member of Parliament to represent them efficiently.
Attempting to create six ridings of even population in northeastern Ontario has still created some absurdities. The most glaring of these involves Wahnapitae. Wahnapitae is a town of a couple thousand people located twenty-five kilometres from Sudbury at the eastern edge of the City of Greater Sudbury. The main road in town is Highway 17, which goes to Ottawa. It has traditionally been in the Riding of Nickel Belt, with the numerous mining towns that surround Sudbury. While it has a distinct identity, most residents commute into Sudbury and its community of interest will always be with the rest of the outlying areas in the City of Greater Sudbury. The proposed redistricting will move the southern periphery of Sudbury from Nickel Belt into Algoma-Manitoulin-Killarney. The dividing line between this new riding and Nickel Belt-Timiskaming in the east will be Highway 17. Unless it is revised, this will divide Wahnapitae. To a lesser extent, it will do the same to nearby Coniston, as Coniston is primarily south of the highway. If ridings are to be culturally coherent, towns of less than five thousand people should not be divided. To do so goes against Canadian electoral districting traditions and would have disturbed every Father of Confederation.
Overall, the proposed new ridings are the product of conscientious and informed work by a team of dedicated commissioners. Trying to design geographically coherent ridings with similar populations and ‘communities of interest’ is a very difficult task. This is more difficult in some regions than others. Communities of interest and population have both changed in recent decades, but not always in complementary manners. In trying to balance the multiple priorities in electoral districting, communities of interest have occasionally been sacrificed, to the detriment of parliamentary representation. In Saskatchewan, the new ridings better reflect communities of interest and will hopefully provide better local representation. In northeastern Ontario, they will generally do the same but problems like that of Wahapitae need to be addressed before the final boundaries are approved.
David Zylberberg is a PhD Candidate in the Department of History at York University.