How the NDP can win British Columbia

The upcoming May provincial election in British Columbia, Canada, looks like it will involve a competitive contest between the incumbent BC Liberals and their centre-left NDP.

Two obstacles stand in the way of the NDP: they need an approximate 3 percentage point swing province wide to hope to pick up the 9 new seats they'll need to win a majority, and they may, like in 2013, be polling better than they will actually perform on election day.

Last time, in the popular vote, the Liberals won the popular vote over the NDP 44-40, which was enough to give then 49 of 85 available seats.

Looking at some work done on the Canadian election blog Blunt Objects about the new riding boundaries, for the NDP to win at least 44 of the now 87 seats, they will need a favourable uniform swing across the province of at least 2.7 percentage points changing from the Liberals to the NDP to pick up these necessary. Key target seats will include Delta North, the new seat of Surrey-Guildford, Fraser-Nicola and Vancouver-Fraserview.

However, a big problem will be that there is little room for error here. Beyond these 9 seats, the current electoral map only has 3 more seats reachable with a province-wide swing of 3.4 percentage points. After that, to pick up any more the NDP will begin to see a province-wide swing of more than 6 percentage points to pick up any new seats. This makes targeting key ridings a critical exercise for the New Democrats if they're not going to benefit from a large provincial swing towards them. (When that happens, local targeting is less important-- wave elections can see candidates with no real campaign at all win election.) It also means that, assuming a uniform swing, the NDP will have to win the popular vote by about 1 percentage point over the NDP province-wide to win a majority government.

The second problem: polls. The latest poll from Mainstreet Technologies put the NDP 5 points up of the Liberals, which would be enough -- if that held -- to give the New Democrats a very narrow majority government. However, there are a few danger signs for the New Democrats. First, undecided voters still make up a large percentage of poll respondents. Undecided voters don't necessarily break in any reliable pattern. Sometimes they break for the opposition, with voters feeling nervous to commit to them until the last minute, and sometimes they have broken sharply for the government, as seemed to be the case in Alberta in 2012, when the Wildrose lost to the long-incumbent PC Party, despite leading every single poll in the campaign.

Second, the Greens and the Conservatives are also both polling very strongly. Smaller parties often poll much better than they end up performing on election day, because they often lack the same brand-loyalty as the larger parties and don't have similar resources to ensure that voter turnout from their supporters is high. Will both of these smaller parties hold their vote? A particularly disastrous scenario for the New Democrats would be for the Green vote to stay loyal, but for the Conservative support to collapse by almost half in the final days, with their voters returning to the Liberals.

A bigger polling question looms for the New Democrats: will they again suffer from shy Tory syndrome? In the 20 polls taken in the 2013 BC provincial election, the New Democrats led in every single one, suggesting margins from anywhere between 2 and 17 percentage points. Some were so confident that the NDP and its leader Adrian Dix would win that the headline "If this man kicked a dog he'd still win the election" ran.

There are lots of reasons for polling failures similar to this: Liberal supporters might be less likely to participate in polls, some voters might feel that it's socially undesirable to admit to voting for a long-serving incumbent government, or voters almost ready to vote for change might simply get cold feet and change their mind at the ballot box. (Various studies have shown that last-minute deciders make up a healthy portion of the electorate.)

How bad was the failure? The three polls in the field in the last few days before election day suggested a popular vote margin of more than 7 percentage points for the NDP. Election day saw a 4 percentage point lead for the Liberals, a dramatic 5.5% swing to the Liberals between the polling results and the vote count. If this was repeated again, the recently-polled 5 percentage point lead for the NDP would turn into a 6 percentage point popular vote victory, easily preserving the Liberal majority government.

Need a guide to Dutch electoral politics in advance of their parliamentary election on March 15?

Have you ever wondered what would happen if a by-election had only one voter but two candidates?

Check out the provincial electoral history of René Lévesque, or search among the thousands of candidates, elections and parties currently in the database.