I do not think this example applies in BC for a number of reasons:
- Before the election started the PCs were leading over the Wildrose Party in the polls and often with a margin similar to the results on election day. Other than during the period the 2010/11 leadership races and immediate aftermath, the BC Liberals have not lead the BC NDP since the 2009 election in any of the polling.
- The polling in Alberta during the election was volatile on the right side of the spectrum but not for the NDs and Liberals. The BC NDP support levels have been consistent in and around the mid 40s.
- Alison Redford was never as unpopular herself as her party seemed to be during parts of the election campaign. The latest polling is showing Christy Clark is personally not very popular.
- The PCs were experienced campaigners and the Wildrose Party were not. Get out the vote campaigns matter a lot in Alberta because of their very low voter turn out. Even though the PC percentage of the vote was down in 2012 by 8.8 percentage points, their actual vote was up. Differing election day get out the vote campaigns could have made a 5-6 percentage point difference in the results. In BC it is not likely that the Liberals will be able to organize as strong a get out the vote as the NDP.
- Wildrose leader Dianelle Smith made some serious campaign blunders which meant that many people shifted at the end of the election by either not voting or choosing to vote PC to stop the Wildrose party. The BC NDP has two elections of taking around 42% of the vote, which is enough to win a large majority this time around. There is no serious danger in BC of a lot of people changing their mind at the end of the election based on the last two elections.
The 2012 Alberta election is not a scenario that BC will see in 2013.
In general, I can not find an example of a government in a Westminster parliamentary system that was consistently been 10-20 points behind for 18 months and then turn it around with three months to go. For many older elections - those more than 25 to 30 years ago - I can not find enough polling data to know what the public mood was over the several years before the election so there might be some examples out there, but certainly nothing anytime since 1990.