when you properly adjust for turnout, you get more normal numbers for 2015 regarding the liberal-conservative swing, which is indistinguishable from any reasonable margin of error - a point or two.

what few conservative<---->liberal swing voters existed did not actually affect the outcome of the election.