Weighted Voting System Didn't MatterThe weighted vote system used by the Liberals for their February 26th vote made no difference to the outcome.
The detailed spreadsheet showing how my calculations are made is available for download by clicking here. The calculations would not be possible without key data obtained by the Vancouver Sun's Chad Skelton who did an outstanding job providing an analysis of the numbers behind the BC Liberal leadership vote. His graphs included the actual absolute number of votes for each of the 85 constituencies for the third ballot. Those data aren't available anywhere else. The Liberals released those figures by accident and Skelton picked them up and made them public. They can be used together with the figures the Liberals did release to calculate the absolute vote count for each of the 85 ridings.
Because voters cast one ballot on which they had to mark at least their first two preferences, we know that the total number of voters on the third ballot had to be the same as the number on the first two ballots except for any voters who marked George Abbott and Mike de Jong as their first and second choices (or reverse order). Hence with only that room for error, it is possible to multiply the percentages by riding, which were released by the Liberal party, by the absolute numbers obtained by Skelton, and get a reliable estimate of what would have happened without a weighted vote. Of course, the Liberals, and probably the leadership candidates, have those numbers. This calculation should come as no surprise to them.
On round one of the vote, the percentage distributed with the weighted votes was: Abbott 24.6%, Clark 37.8%, de Jong 9.3% and Falcon 28.4%. My calculation of the distribution (raw numbers in parentheses) with absolute votes was: Abbott 20.1% (10,986), Clark 38.2% (20,860), de Jong 10.6% (5,766) and Falcon 31.6% (16.993). On the first ballot, the weighted vote made little difference to Clark. It gave Abbot 4.5%, and it cost Falcon 3.2%.
On round two of the vote, the percentage distributed with the weighted votes was: Abbott 22.8%, Clark 42.1% and Falcon 30.2%. My calculation of the distribution (raw numbers in parentheses) with absolute votes was: Abbott 23.1% (12,613), Clark 43.8% (23,945) and Falcon 33.1% (18,070). Clark got a 1.7% advantage from the weighted vote on the second ballot; Abbott was about even but had to drop out, and Falcon had a 2.9% loss due to the weighted system.
On round three, by the weighted vote the final result was Clark with 52.0% (4420 points) and Falcon with 48.0% (4080 points). For the absolute vote count Clark had 52.1% with 28,488 absolute votes, while Falcon had 47.9% with 26,183 votes.
There is more to learn from the actual vote counts. Surrey - Newton had 3,210 voters while Peace River South had only 110 voters. A future article will discuss those differences for all 85 ridings. I learned to double check my calculations before Tweeting.