Polls and BC Election

Now that the final numbers are in I can begin some B.C. election post-mortems. On the polls I have compared the final vote shares province-wide to the predicted numbers from Mustel, Ipsos Reid and the Strategic Counsel.

The final vote percentages were Liberal - 45.8; NDP - 41.52, Green - 9.18 and Others 3.5.

And the closest pollster was the Mustel Group with a total net error of 5.64 percentage points, followed closely by Ipsos-Reid at 6.04, both being good performances. The Strategic Counsel poll for the Globe trailed substantially with total error of 14.04. Their estimate of the NDP vote at 36% was outside their claimed margin of error ± 3.1%. The poll had also suggested a 13 point win for the Liberals.

My forecast model predicts a Liberal win of 49-30 compared to the actual 46-33 using the actual vote shares. The model adjusts outcomes based on changes from the 2001 result, which was likely out of the ordinary pattern of B.C. politics. The results are interesting and I will post on them at a later date. It is useful I find to compare them to the 1996 election, which more closely resembles 2005 in terms of its overall outcome than 2001.