on funds raised in Alberta and British Columbia; the party’s Quebec
offshoot, the Créditistes, enjoyed initial but not sustained success at
fundraising by using paid television programming featuring their fiery
leader, Réal Caouette.
Research findings in a 1966 report by a parliamentary Committee
on Election Expenses led to the recommendation to limit the depend-
ence of parties on a few big donors.42 The assumption that campaign
spending influences votingbehaviour led in 1974 to legislation im-
posing a ceiling on the amount of spending permitted during
in Transition: Discourse, Organization, and Representation , ed. Alain-G.
Gagnon and A. Brian Tanguay (Toronto: Nelson Canada, 1989), 67.
9 Robert L. Stanfi eld, “Conservative Principles and Philosophy,” in Politics:
Canada , ed. Paul Fox and Graham White, 376–81 (Toronto: McGraw-Hill
10 Robert Krause and Lawrence Leduc, “VotingBehaviour and Electoral
Strategies in the Progressive Conservative Leadership Convention of
1976,” Canadian Journal of Political Science 12, no. 1 (1979): 121.
11 Perlin, Tory Syndrome , 174
attitudes, demographic characteristics, or other
behaviours. Much of the early research on votingbehaviour was de-
voted to establishing correlations between partisan loyalty and demo-
graphic characteristics such as sex, race, religion, income, or education.
With the aid of advanced statistical techniques of multivariate analy-
sis, behavioural research can become highly sophisticated in its ability
to control for extraneous factors; but without an adequate theory of
human nature it remains basically correlational. And as all statistics
teachers tell their students
well over 40 per cent, so the appeal should be broad-based. In
Ontario, and to a lesser extent Manitoba, the party will undoubtedly do some
targeting, but less than one would think. Votingbehaviour is becoming less and
less predictable. Therefore, unless all three tests - namely, past voting patterns,
current opinion-polling results, and internal party intelligence - suggest there is
no real prospect of success, no riding should be ignored. If at all possible, every
riding should be contested.
Indeed, there are strong reasons for running a good campaign in a
that at the same time as the parties are so influential
and powerful within parliament, they are weak outside it, both in terms of gaining
consistent strong allegiances within the electorate and in terms of generating ideas
and policy proposals. It is a well-known and often criticized fact that the parties are
the most important control over an MP'S votingbehaviour within parliament. Not
so much appreciated is how important party is to the entire career of an MP, from
election, through every aspect of parliamentary tenure, to post-parliamentary life.
At the same
1861.’ Social History / Histoire sociale 7 (1974): 355–65.
– Hopeful Travellers: Families, Land, and Social Change in Mid-Victorian Peel
County, Canada West. Toronto: University of Toronto Press, 1981.
Gagan, D.P., P.J. George, and E.H. Oksanen. ‘Ontario Members of Parliament:
Determinants of their VotingBehavior in Canada’s First Parliament, 1867–
1872.’ Social Science History 9/2 (1985): 185–98.
Garlock, Jonathan. Guide to the Local Assemblies of the Knights of Labor. Westport,
Conn.: Greenwood Press, 1982.
Garner, John. The Franchise and Politics in
it a useful explanatory factor; also, prelimi-
nary analysis suggests that whatever power a measure of prior judicial ex-
perience had in the earlier period, it is no longer a significant predictor of
the justices’ votingbehaviour. Consequently, these two Tate and Sittiwong
variables were not included in the analyses reported in this chapter.
7 Since tables 7.6, 7.7, and 7.8 uncovered no relationship between any of these
three variables and the tendency to support pro-government outcomes in
other public law cases, no model of pro-government behaviour will be
) and the 2011
Canadian Election Study (CES), allowing us to measure information flow
and votingbehaviour daily over a one-month election campaign.2
The results fall squarely in line with recent work on campaigns by
showing that the campaign did matter in these elections, which is to say
that information conveyed during the campaigns had an effect on indi-
viduals’ voting choices. But we use the two-moderator model to show
that the more complete story is that campaign information has strong
effects on only a subset of voters: those with both relatively high