Incumbent MPs have something of an advantage in elections, since they have had the benefit of experience and the public
exposure that comes with serving in office. As a result, interesting contests occur in ridings where the sitting MP has decided
not to run again.
Milton Chan runs a unique site
riding-by-riding predictions of the contests in each constituency
for national and provincial elections
This site relies on people submitting their assessment of ridings that
they have local knowledge of. Connect to the site and add your two cents
worth! This site was uncannily accurate in the past, correctly predicting 91% of the seats in 2008 - although the success did drop to 76% for the 2011 election.
Since our electoral system is created around local elections, one would think that local candidates would play
an important role in people's decision about whom to vote for. However, opinion polls have consistently found that
Canadian voters overwhelmingly decide how to vote on the basis of party leaders or platforms. A Nanos poll conducted in March 2011 found that only 12% of Canadians felt that the local candidate was the most important factor in deciding how to vote; 48% said they were most influence by a party's policies, 20% by the party leader, and 10% simply said they traditionally vote for the party. This trend was evident in
an Ekos poll
released at the start of the 2008 campaign, in which less than 9% of
respondents claimed that the local candidate is the main factor in
their decision. An Ipsos
poll conducted Sep 17-18 asked a somewhat different question but still
found only 17% saying that local candidates were their most important
consideration. You can read an interesting academic paper on the
importance of individual candidates:
"Does the Local Candidate Matter?" (pdf file) by André Blais, Elisabeth
Gidengil, Agnieszka Dobrzynska, Richard Nadeau, Neil Nevitte.
Readjustment of Electoral Boundaries
Each decade independent
Boundaries Commissions are set up to review the distribution of seats in the House
of Commons to ensure that the number and size of each province's constituencies
meets the legal requirements. The Supreme Court of Canada has ruled, in
Boundaries Reference case, that Canadians
should have a relatively equal voting power. As a result, the number of
voters per constituency should not normally vary more than + or - 25% from
the average. From time- to time, new ridings are added, and at others
the seats are redistributed among the provinces. The
used to allocate seats among the provinces and to draw boundaries has
changed considerably over the years. The previous
readjustment of electoral boundaries was done in time to come into
effect for the 1997 general elections. Prior to 1985, adjustments
to federal boundaries were achieved by amending the British North America
Act, 1867; since 1985, however, the changes are incorporated into ordinary
51(1) of the Constitution Act, 1867 and
Boundaries Adjustment Act governs the process currently used.
However, riding populations can shift dramatically within short periods of time. You can look up the differences in the populations of each riding, to see how they have changed between the 2001 and 2006 Census. Just in this 5 year period, the population changes range from -7.6% in BC's Skeena - Bulkley Valley to a whopping 52.5% in Ontario's Oak Ridges - Markham. StatsCan has a
sortable table of population and dwelling statistics for all the federal
ridings. You can also download an Excel file with all the
population and dwelling data for
Canada's electoral districts.
The 2011 general election was conducted using the same 308 ridings contested in the 2008, 2006 & 2004 elections;
by contrast there were 301 constituencies in the 1997 and 2000 elections.
The 2006 elections were conducted with the same electoral boundaries used in the
2004, with the exception of two constituencies in New Brunswick: Acadie–Bathurst and Miramichi. Note that a few ridings have new names since 2004, but their boundaries have remained intact.
Because the boundaries of electoral districts changed between the 2000 and 2004 elections it is necessary to transpose the results on a poll-by-poll basis before you can compare the results
on a riding-by-riding basis. You can
see how the 2000 elections results would
have been under the new 308 seat distribution (Adobe pdf file). Elections
Canada also shows what the results would have been on a
riding by riding basis.