The general tone of media coverage of polls and the upcoming federal election is that the Liberals can win no better than a minority. The lead for the story the Globe story on Saturday March 27 on the poll was: “Prime Minister Paul Martin's leadership is providing little help for Liberal support still mired in minority-government territory…” This is a reasonable guess at the outcome. However, my own seat calculation actually has the Liberals winning a small majority so I think the Globe’s estimate should be more qualified.
I have three ways of translating poll numbers into seats, and all yield, in the case of the Reid poll, similar but not identical numbers. My core estimate is as follows:
Liberal 157
Conservative 78
NDP 18
BQ 53
A bare majority would be 155 seats. Given the errors inherent in polling & seat estimates, 157 would be right on the line between minority and majority. A very slight shift in opinion would either yield a comfortable Liberal majority or a weak minority.
Some further observations:
1. I have the impression that there may be a good deal of latent anti-Liberal sentiment that might not be fully captured by any of the polls. Any long-established incumbent government tends to accumulate voter grievances, and the Martinites efforts to distance themselves from the Chrétien era simply can’t completely overcome this phenomenon.
2. One problem facing the analyst like myself who must rely on published polls is that there are relatively few, and most are from a single source: Ipsos-Reid, the company that polls for the Globe and Mail. In general, I think the Reid polls are ok but I do think they overestimate Green and “Other” support. For example, the Globe poll reports 5% for the Greens and 4% for “Others”. The NDP will be running on a very Green type platform. You can’t really tell for sure but it is possible that Reid is underestimating NDP support. I am guessing that they include the Greens in their ballot question. In 2000 the Greens received 0.8% of the vote and “Others” received 1.5% so this poll is estimating a huge increase in these two categories. I am skeptical.
3. I think the Conservatives are a great deal weaker overall than is the impression conveyed in media reports. When the Mulroney PC coalition split up between ’88 and ’93, it divided into three parts, not just two. Add Bloc support to Conservative support and you get 37% compared to the Liberals’ 38%. If the Mulroney coalition could be recreated you would get a race that is genuinely competitive. But the Mulroney coalition is still in two distinct parts, so for the Conservatives to be really successful they must win over large numbers of Liberals.
4. This poll was taken too soon after the budget and the election of Stephen Harper to really capture the impact of either.
That is enough for now. I haven’t been posting much partly as a consequence of problems with my back and neck.
Comments on Canadian & American politics, economics, polls, elections, and media.................
Tuesday, March 30, 2004
Wednesday, March 10, 2004
Minority or Majority?
I have been otherwise preoccupied for awhile, in part reconstructing my forecast model with data from the transposition of the 2000 results to the new electoral boundaries.
I do have a simple version of my forecast model up and running, although it requires more tweaking.
When I input a weighted average of the last three nationall polls (the last two Reid polls plus the Ekos poll) I get a Liberal minority government based on winning 152 seats (155 needed for a majority) with 74 for the Conservatives, 26 for the NDP and 56 for the Bloc. Just the most recent Reid poll gives the Liberals a majority.
Despite the media focus on the Conservatives, it appears to be the Bloc that represents the greatest threat to the Liberals. Quebec also appears to be Martin's best opportunity to forge a comeback. To accomplish this he must take the attention of Quebecers off the scandal and focus it elsewhere. Quebec political trends are always difficult to devine and the current moment is no exception but it appears to hold the key to the next election.
The Liberal slide was significant in the West but only in relation to recent trends. The Liberals are still running ahead of their 2000 showing west of the Manitoba-Ontario border. The new Conservative Party is currently doing well only in Alberta and even there they are weaker than they were in 2000.
Majority or minority? One thing seems certain. It is the Liberals we are talking about not the Conservatives.
I do have a simple version of my forecast model up and running, although it requires more tweaking.
When I input a weighted average of the last three nationall polls (the last two Reid polls plus the Ekos poll) I get a Liberal minority government based on winning 152 seats (155 needed for a majority) with 74 for the Conservatives, 26 for the NDP and 56 for the Bloc. Just the most recent Reid poll gives the Liberals a majority.
Despite the media focus on the Conservatives, it appears to be the Bloc that represents the greatest threat to the Liberals. Quebec also appears to be Martin's best opportunity to forge a comeback. To accomplish this he must take the attention of Quebecers off the scandal and focus it elsewhere. Quebec political trends are always difficult to devine and the current moment is no exception but it appears to hold the key to the next election.
The Liberal slide was significant in the West but only in relation to recent trends. The Liberals are still running ahead of their 2000 showing west of the Manitoba-Ontario border. The new Conservative Party is currently doing well only in Alberta and even there they are weaker than they were in 2000.
Majority or minority? One thing seems certain. It is the Liberals we are talking about not the Conservatives.
Subscribe to:
Posts (Atom)