Before I jump in, let me be clear on two points:<\/p>\n
1. This model has evolved over the past 3 years (though Stockwell Day will deny it has<\/em>) and each step in it involved a lot of thought, testing, and consideration.<\/p>\n 2. I\u2019m completely 100% open to criticism and suggested improvements. I\u2019m sure it can be improved.<\/p>\n That said, let\u2019s jump in.<\/p>\n THE PROJECTION FRAME<\/strong><\/p>\n At the basic level, this is a uniform swing projection using regional data (Atlantic Canada, Quebec, Ontario, Prairies, Alberta, BC). If the Liberals go up by 5 points in Alberta, the model swings every riding up 5 points. If the NDP drops 8 points in Atlantic Canada, the model drops every riding down 8 points.<\/p>\n I experimented with geometric swings (including the variation 538 used for the UK election<\/a>) and mixed models, but they simply didn\u2019t work as well on any of the elections in my data set.<\/p>\n The reason for this is simple \u2013 if the Liberals swing up 5 points in Ontario, the geometric swing is going to give them almost all of their newfound support in rural areas, since that\u2019s where there are the most votes available to swing to them. And that\u2019s not<\/strong> what usually happens in reality.<\/p>\n Now, I\u2019ll concede in a WTF election like, say 1993<\/a>, when the political landscape of the country is dramatically changed, uniform swing may not work. I wouldn\u2019t use this model to project the next Alberta election. But uniform swing is simply the best frame to build this model out of, especially since its deficiencies are easy to correct at the simulation stage<\/u>.<\/p>\n THE BASE VOTE<\/strong><\/p>\n Projection models generally use the most recent election as their base. The problem with this is that if a party has a good or bad showing in a riding, the data becomes skewed beyond repair<\/em>. It’s the same reason you use more than 1 year worth of data to project hockey or baseball stats – even though Aaron Hill hit 36 home runs last year, it\u2019s foolish to expect him to repeat the feat.<\/p>\n In the political arena, there are prominent examples of this that stand out \u2013 the Greens won\u2019t match their 2008 totals in Central Nova<\/a> next election and, no matter how bad the campaign goes, the Liberals will exceed theirs. The NDP results from last election in Saanich Gulf Islands<\/a> were obviously hurt by some naked truths<\/a> that (hopefully) won\u2019t be repeated in the next election.<\/p>\n But even on a more subtle level, parties get good candidates, candidates run bad campaigns, local issues emerge. There needs to be a way to smooth these events out<\/u>.<\/p>\n To test this, I used a regression model to “predict” the 2008 vote in each riding, based on the 2006 and 2004 riding results and each riding\u2019s predicted vote based on demographics (click here for the low down on this<\/em><\/a>). In all cases, these numbers were adjusted using uniform swing.<\/p>\n And the results were clear \u2013 the best model used all three predictors (2004, 2006, and the demographic regression)<\/strong>.<\/p>\n So keeping the ratios the same, I\u2019m using the following results to get the \u201cbase\u201d vote for each riding:<\/p>\n 2008 election: 38.7% OTHER FACTORS<\/strong><\/p>\n Incumbency exists. It means something. It should be taken into account. I\u2019m not going to argue the point any further, because every study ever run on Canadian politics has come to this conclusion, as has my own research<\/a>. So, I\u2019m adjusting for incumbency based on the effect it had on the 2004, 2006, and 2008 elections.<\/p>\n
2006 election: 17.4%
2004 election: 7.8%
Demographic regression: 36.2%<\/p>\n