The B.C. Liberals have a lot of ground to make up and not a lot of time to do it. If you ask them, they might say that the polls in British Columbia tend to tighten up as an election approaches. They would not be wrong, but even that provincial pattern is probably not enough to save them.
With fewer than eight weeks to go before British Columbians head to the polls, the B.C. Liberals are in a deep hole. The most recent survey, conducted by Ipsos-Reid for Global News, put Christy Clark’s party 19 points back of the leading B.C. New Democrats. The opposition party managed an impressive 51 per cent support in the poll, their best result under Adrian Dix, who became NDP leader in April 2011.
But that is not the widest margin that has been awarded to the NDP in recent polling: a survey by Justason Market Intelligence gave the party a 22-point edge in a late January poll. Overall, the New Democrats have averaged a 15-point lead over the B.C. Liberals in the last 10 polls stretching back to October.
It has become a common refrain among B.C. Liberal boosters and those who want to keep the race interesting – but also from those who have long memories in British Columbia: the polls tighten up when it comes time to vote. Is there any truth to that?
The following chart tracks the size of the lead the Liberals enjoyed over the NDP in the six months running up to the elections of 2001, 2005, and 2009. The lead that the NDP has held since November is also tracked (in yellow). The chart uses a three-poll rolling average in order to smooth out any statistical anomalies.
The experience of the B.C. Liberals over the last three elections does seem to suggest that the polls in British Columbia do tighten up once the campaign starts. In all three election cycles, the lead the Liberals held over the NDP grew before dropping back down by election day. In 2001, the lead increased in size until about two weeks before the vote, going from over 50 points to 46 a few days before the election, before settling at 36 points.
In 2005, the lead grew to just over eight points when the election was called, before decreasing throughout the campaign (save for one late campaign poll, which turned out to be an outlier). And in 2009, the lead shrank from almost 14 points two months out to four points on election night.
It does give the B.C. Liberals a glimmer of hope – but not much more than that. Though the polls have tightened up during campaigns, the largest difference between the margin in a poll and the election result over the last three elections was 18 points, less than what would be needed to close the gap that was recorded in Ipsos-Reid’s survey (though enough to put the Liberals ahead in six of the last nine polls). But that was from a 1996 survey that put the Liberals at an unrealistic 70 per cent support to only 16 per cent for the New Democrats. That sort of lopsided result is the kind that will almost always be corrected at the ballot box.
More realistically, the largest swings have been closer to around 13 points. That only covers the spread in three of the last nine polls. On average, the margin has closed by about six points over the last three campaigns – not nearly enough to put the B.C. Liberals in a position to win on May 14. In fact, it has been almost two years since the margin was that small.
The 1996 campaign is one that is often cited as inspiration for a B.C. Liberal comeback. Around six months before the vote, Gordon Campbell’s Liberals held about a 30-point lead over the New Democrats. But on election day, that margin had been reduced to only two points – enough to give the NDP more seats despite a vote deficit.
However, a change in the NDP’s leadership (swapping Mike Harcourt for Glen Clark) played a big role in the change of fortunes, and in any case the tightening of the polls took place before March – at this point in 1996, the Liberals and NDP were already neck-and-neck. And most of the gap-narrowing in the four elections between 1996 and 2009 were to the benefit of the New Democrats, not the B.C. Liberals.
The B.C. New Democrats are in a strong position. The lead they currently hold over the B.C. Liberals is larger than the gap that existed at this stage of the race in 1996, 2005, and 2009. And in the last three campaigns, despite a closing of the gap, the party that was leading in the polls ended up winning the election.
Recent history suggests that the B.C. Liberals have good reason to expect the horserace to get a little more interesting between now and May 14. They could also hold out a long-shot hope that B.C. will be another Alberta in terms of poll accuracy (despite, for a myriad of reasons, the odds of that happening in British Columbia being extremely low). More realistically, however, the B.C. Liberals may have dug themselves too deep to have any real expectations of victory.
Éric Grenier writes about politics and polls at ThreeHundredEight.com.