A two-purpose survey was conducted among members of the Conservative Party of Canada (CPC) prior to the party’s first leadership convention later this month. One purpose was to measure members’ beliefs, purposes for belonging, and expectations for the future. The other purpose was to forecast the outcome of the leadership contest. Each purpose used a different sample size and different sample structure.

A leadership-focused, n=1254 survey among CPC members was conducted March 9-11, 2004 with at least 4 interviewees by riding in order to take into consideration Conservative Party of Canada (CPC) rules governing the leadership contest.

A subset of n=507 respondents were interviewed at length about how and why they became members, their beliefs, and their expectations.

The n=507 subset was selected to reflect the actual regional distribution of membership, which is disproportionately strong west of the Ontario-Quebec border. The n=507 sample, reported in Part 1, can be deemed accurate to within 4.5 percentage points 19 times out of 20.

The n=1254 survey on the leadership contest, reported in Part 2, can be deemed accurate to within about 3 percentage points 19 times out of 20 with respect to the leadership preferences of CPC members, assuming equal weighting by constituency, i.e. assuming that each riding had the same number of CPC members. By acting on this assumption, our sampling structure parallels the process of CPC voting as required under the party’s rules.

Estimating how respondents’ leadership preferences translate into convention votes does not have a known margin of error. In practice, we offer two estimates of how CPC members’ leadership preferences translate into convention votes—first assuming that all members will have an equal propensity to vote, the other that some will indeed be more apt to actually cast their ballot.

View / Download complete poll in PDF