Does Canada have a bias in favor of rural areas?
A comment to the post on Canada's study of different electoral systems asked if anyone knew if the Canadian parliament had a rural bias (by making rural seats with fewer people). I asked my friend, J. Paul Johnston, who teaches pol sci at the University of Alberta. Paul responded and gave me permission to post his email.
First, a comment about the "comment" regarding rural over-representation in
Canada. Most of it is actually at the provincial level, that is, in the
allocation of seats in provincial legislative assemblies. At the federal level
it appears, though not in any great degree, in two contexts. First, the rules
regarding redistribution under which the "independent" [by law at the federal
level and to varying degrees at the provincial one] electoral boundaries
commissions [an institution Americans should take a LONG, HARD look at!!, then adopt!] operate allows them to accommodate "remote, sparsely populated areas," such as one finds in Labrador or the far northern parts of Quebec, Ontario and the western provinces, by legally establishing districts that are
over-represented by having populations less than the -25% deviation from the
quota set for the province [population quota vary by province due to various
special arrangements that govern the federal apportionment of seats in the
House of Commons among the several provinces, e.g., the quota used in Alberta for the redistribution just completed was just under 107,000 persons.] For example, in Newfoundland and Labrador,there is a single district comprising
the entirety of Labrador which is 62% [i.e., -62] below the provincial quota;
and Ontario has one that is 43.7% below its provincial quota. Secondly, the
rules in some provinces [e.g., Nova Scotis] also allow them to exceed that
lower limit in order to accommodate certain minorities by setting up what are
essentially minority majority districts [or the Canadian "flavor" of that
idea]. This was done recently in Nova Scotia to accommodate Black communities
in the Halifax areas [a carry-over from colonial days and the Underground
railway later] and the MicMac tribe of native Canadians. To achieve this they
formed two districts that are under-represented [!!] by being 20.4% and 28.86%
above their quota. And New Brunswick has a seat for the Acadians achieved by
having one district 27% above the quota and another 21.0% below it.
At the provincial level, the rural over-representation that exists stems
largely from legislative assemblies writing the rules so that urban and rural
districts are formally and legally distinguished and then, in some cases,
designating how many urban [or, as they are called in Alberta "single
municipality"] districts are to be formed [even naming the urban centers and
the numbers of districts for each of them]. Hence, rural dominated assemblies
have been able to retain their status as such by "fixing" the urban-rural
division in the apportionment stage by writing it into the terms of reference
given to the boundary commission. Even then, however, the degree of
malapportionemtn has declined rather markedly in recent years, largely because
urbanization has over the past fifty years left most of the provinces [other
than in the Maritmes]70-75% urban in population. Indeed, the urban-rural mal
apportionment involved in the Supreme Court of Canada's "landmark" decision on
the matter in "The Carter Case" [1992], involving the Saskatchewan Legislative
Assembly, was less than 10% off what the true apportionment should have been.
[You probably should get a copy of John Courtney's new Book, Commissioned
Ridings, that I reviewed for the newsletter of the Representation and Electoral Systems section of the American Political Science Association. It's indispensable for detailed
information about the redistribution process in Canada and its outcomes.]
Enough of my "Pol. S. 496: Representation and Electoral Systems" lecture.
On to the rest of your inquiry.
You can get the exact district populations for the redistribution just
completed at the website of Elections Canada [www.elections.ca] by just
clicking on the "icon" designation "Federal Representation 2004" at the lower
left corner of the screen. You have to open each of the ten provincial and
three territorial commissions sites separately, but everything you'll want to
know is given there, including district maps that can be downloaded or
printed.
You might note, in particular, that the "population variance" range in
district sizes in B.C. puts 24 of 36 ridings within a -5% to +5% range, 32 of
them within -10% to +10%, and all 36 are contained within a -15% to +15 range,
though the formal limits allow up to -2% to +25%. In Saskatchewan, the range
including all 14 districts is -7.88% to +5.54% and Manitoba's 14 seats range
from -5.49% to 5.54%. These ranges may seem quite "generous" given the "strict
equality' standards applied by American courts, but they represent
considerable "progress" in other jurisdictions, such as Canada, where the
requirements for delivering "effective" representation in practical terms are
taken seriously.
As for the urban/rural composition within districts, you have to go to the
official election returns for specific elections for that information, using
the district level results. Each poll within a district is designated as being
urban, rural, or mixed, and the population contained in it is reported. You
get that information at the same website; www.elections.ca.
Comments
As the Courtney book very nicely outlines, Canada has another form of bias that appears in the apportionment, rather than the districting process. Apportionment of seats in the House of Commons is not strictly by population. The Representation Act of 1985 had a grandfather clause that guaranteed that no province would have fewer seats than it had in the parliament elected in 1986. The Act also guaranteed that no province could have fewer MPs than senators, thus assuring that the number of MPs was not proportional to population even in 1986. All of the provinces are overrepresented in the House except for Ontario, Alberta, and British Columbia. Quebec gets the most extra seats (7) but PEI benefits the most in percentage terms in going from 1 to 4 MPs. As a result, the the ideal seat in Ontario is nearly 108,000 while the size of the ideal seat in PEI is under 34,000. And don't forget that the lightly populated territories are each guaranteed at least one seat.
Elections Canada has a nice discussion of the history of apportionment and the current formula at:
http://www.elections.ca/scripts/fedrep/federal_e/fed_prlmnt_e.htm
I am not necessarily trying to say that the malapportionment among the provinces is bad. The Maritimes would collectively have very few seats without it. John Courtney also aptly pointed out to me that governments would have few MPs available to help provide regional diversity among ministers without this deviation from population equality.
As another colleague once pointed out to me, the necessity of alloting seats in whole numbers to states means that there are substantial differences in district sizes across American states as well despite the strong focus on population equality between districts in the U.S.
Posted by: David Lublin | May 28, 2003 10:47 PM
Some other significant differences between the US and Canadian process:
- Canadian redistribution (and all federal election law) is entirely under federal jurisdiction. One set of rules for voting, one for drawing boundaries... The provinces don't set the rules for federal elections (anymore -- we got rid of that in the 1880s IIRC).
- Canadian ridings (federal and provincial) are all much more geographically compact than their US counterparts. I was absolutely shocked, then appaled, by the grotesque ethnic gerrrymandering and outrageous topological contortions, that go into drawing Congressional districts. In Canada, geography and population are paramount in the redistribution process. I really can't figure out how the US has allowed its 40-whatever redistricting processes to get so out of control. (The single-seat states obviously don't have that problem.)
Posted by: WJM | December 21, 2003 3:56 PM