Sunday, October 17, 2010

An essential caveat

Before anybody makes too much out of the study on partisan loyalties highlighted by David Akin, it's worth reading through to the paper itself for some rather important limitations.

The analysis carried out by Cameron Anderson and Laura Stephenson is based solely on results from the 1984 Canadian Election Study as the last source of data on parental partisanship. And that would seem to create major problems to the application of the study's conclusions - not just due to the passage of time and changes in political parties generally, but particularly due to the fact that partisan loyalties at a single point in time surely have plenty to do with the political realities of that day rather than long-termm ternds. Indeed, one would figure that a weak Lib party was likely relying more on on historical attachments rather than favourable views of its leader or policies, while relatively strong PC and NDP efforts in the 1984 election were based on their doing better than usual in the persuasion department at the time the data was gathered.

So while I'll look forward to seeing some follow-up to test whether the conclusions actually do apply to other elections where the parties' relative strengths would seem to have been different, it's far too early to draw conclusions as sweeping as Akin's based on the single point in time that was actually tested.

No comments:

Post a Comment