Electoral math II: vote splitting in the Liberal ridings the Conservatives won
May 3, 2011

Electoral math II: vote splitting in the Liberal ridings the Conservatives won

While there was an enormous transfer of seats towards the NDP from the Bloc and to a lesser extent the Liberals, the Conservatives managed to move from a minority to a majority largely due to 26 seats they picked up from the Liberals. Three weeks ago, the NDP was stagnant in the polls before its miraculous surge. The question is, was the Conservative majority significantly due to vote splitting from the NDP's surge? There is a larger background question regarding vote splitting on the left in general which was addressed in Electoral math 1, but for this post I assume a baseline of the 2008 results and consider only the vote splitting from the surge.

Here is a list of the 26 ridings the Conservatives took from the Liberals in the 2011 election. Several of the small number of other ridings that changes hands involving the Conservatives are independent of the NDP surge such as the two ridings the Conservatives picked up from the NDP or, say, the Conservative loss to Elizabeth May which would all be more likely to occur without an NDP surge not the other way around. The NDP won four from the Conservatives - all of them in Quebec - which likely would not have happened otherwise. All else being equal, the Liberals would have needed to win 13 more seats to have prevented a majority, 17 seats if you want to assume the NDP doesn't win those four Quebec seats from the Conservatives.

In these 26 ridings, I considered who would win if we, riding by riding, added x% of the NDP vote to the Liberal vote and compared that with the Conservatives. To accomplish the 13 seats requires 16% of the NDP and 17 seats requires 35% to be transfered to the Liberals. These may sound like large numbers, but it is important to remember that in Ontario (where most of those 26 ridings are) the NDP had a 140% increase in the popular vote while the Conservatives increased their seat share by 20% in the province. If one assumes that NDP and Conservative voters overlap with the Liberals but not with each other, had the NDP surge not occurred and the votes thus stayed with the Liberals there is easily room for the 16% of NDP votes to be taken. The 35% number however is on the cusp and certainly a tiny majority or very strong minority are both within the uncertainty of this crude constant distribution model should the NDP surge not have happened.

Most projections of the result before the elections still predicted a Conservative minority, even as they adjusted for the late game NDP surge. Ultimately, I think it is safe to say that vote splitting, particularly in Ontario, was a relevant factor but does not tell the whole story. The ability of the Conservatives to secure over 40% of the popular vote - with little of this due to with the NDP surging - combined with vote splitting to give them the solid majority they received.

Previously: Electoral math 1

Thoughts on this post? Comment below!

Share this post:

Tweet It! Facebook Add Feed Reddit! Digg It! Stumble Delicious Follow

Post a Comment

Frequent Topics: