A batch of new polls popped up just as the government was falling Friday and the country's 41st general election was about to get underway. The polls are unanimous on this issue: Harper and the Conservatives are out in front; the Liberals are second, followed by the NDP and then the Greens. In Quebec, the Bloc Quebecois is the heavy favourite.
And yet, each pollster — Ipsos Reid, Ekos, Angus Reid, and Leger — all have substantially different actually numbers. Let's take a look:
Why so different? I'm no professional pollster (and they are probably best positioned to answer that) but I would say that, though they purport to measure the same thing — i.e. voter intention — they each used substantially different methodologies to do that. EKOS, for example, was out in the field from March 17 to March 24 while Ipsos was just March 23 and 24. One thing, then, to consider is: All of Ipsos survey respondents would have had information about the federal budget March 22 while only some of Ekos respondents had that information. Would that have influenced voter intention? The other two polls were also out in the field over slightly different time periods.
Each pollster contacts survey participants differently. Some use an online panel, some use old-fashioned phone calls; some use a mix.
The precision of these numbers will, I think, be more important closer to polling day on May 2.
But for now, one way that it seems useful to me to “interrogate” these numbers is rather than compare pollster-to-pollster, what might we learn if we compared each firm's previous poll to each firm's current poll? In between polls, a lot might have happened that one might think would influence the numbers. The Bruce Carson scandal, contempt of Parliament hearings, the federal budget are just three.
Now I don't have (or can't find) a comparable Leger poll but let's take a look at the other three firms, all of which were in the field at the beginning of March for polls released around March 10 or so.
Well, look at that. All three pollsters found that despite a budget and some heat over ethics issues, Tory support remained unchanged. One pollster found Liberal support unchanged; one found a bump; one found a slight drop. The Liberal changes are all within the margin of error so probably safe to say Liberal support is largely unchanged. Same with the NDP and Greens.
In other words, NOTHING happened through the middle of March that seemed to get a voter angered/excited enough to change their mind about who to vote for!
Let's see if these election campaigns will change that.