Ali Kashani, a data-scientist, has run the numbers on Canada's electoral constituencies (called "ridings") and concluded that if the candidates from the NDP and Liberal parties in sixteen of those ridings agreed to one or the other withdrawing, the Conservative Party could not form the next government.
The NDP — a social democratic party — formed the official opposition in the last national election, and went on to take the Alberta provincial election, a surprise upset as Alberta is the Conservative stronghold and had been Tory for more than half a century. Just a month ago, they were leading in the national polls. Despite this, the press continues to treat them as the "third party" and covers the Liberal party and its MPs as though they were the opposition and the NDP were just an anomaly (this may be a self-fulfilling prophecy, of course).
But the Tories have crept up in the polls since, taking the lead, despite the party's disastrous reign, which culminated this week in the finalisation of the Trans Pacific Partnership, whose major selling point from the Tory minister who negotiated it is that it may increase the national GDP by 0.5% over a decade — while costing Canada legal sovereignty, requiring it to extend copyright terms, legally binding it to protect DRM, and weakening its ability to uphold its obligations to First Nations people.
Kashani has calculated that a tactical agreement between the Liberals and NDP in sixteen ridings would hand a majority to a "progressive" candidate in each of those ridings and consequently deny the Tories a chance at forming the government.
As Koshani points out, the Tories have been a disaster. However, I think that it's a stretch to call the NDP and Liberals both "progressive" and imply that whichever one wins, it'll be better than the Tories. The Liberals created Canada's SOPA, bill C-11 (though it was passed by the Tories later, with Liberal support). The Liberals backed the Tories on bill C-51, Canada's mass surveillance bill. The Liberals support TPP.
That said, Kashani's projections show that in half of the sixteen ridings, the NDP has no hope of winning; and in the other half, the Liberals have no hope of winning. A withdrawal in all sixteen ridings would make it such that either party could still form the government and the other would form the opposition, and the two thirds of Canadians who want to get rid of Harper would get something like their wish (assuming that Liberals would turn up to vote for the NDP if it meant voting against a Tory; and vice-versa).
It's true that there are other areas in which the Liberals would be preferable to the Tories, but when it comes to mass surveillance; gutting environmental, labour, and First Nations policies in the name of free trade; and giving legal protection to DRM so that you can't examine or modify your property, nor rely on third parties to tell you about deadly security vulnerabilities in systems with embedded computers, the Liberals hold essentially the same policies as the Tories.
That, more than anything, is the fly in the ointment here. Its true that the Liberals would load up the top civil service positions with people who weren't the rapacious fiduciary-value-cultists that the Tories prefer, and in that regard they would be a better government than the Tories, but compared with the NDP, the Liberals are a terrible party.
Strategic voting is a politically-aware populace's solution to systematic electoral defects. It requires ground-up campaigns to inform and engage the electorate, and there are now various organizations in Canada doing just that. But it also requires reliable polls to guide the voters.
When it comes to tight races, strategic voters face a number of challenges:
Polling results cannot be trusted when races are within the poll's historical error margin.
*The voters find it particularly hard to vote strategically for a less desirable candidate, especially when their favourite candidate is still within striking distance.
*Voters are concerned about throwing their vote away in a failed attempt at strategic voting.
We must acknowledge these barriers when proposing a voting strategy, in order to rally the necessary voter support.
The proposed 16-riding strategy addresses these problems: voters are only encouraged to vote for a less desirable candidate, if their favourite candidate is not within striking distance (even when factoring in polling errors).
This approach only requires the cooperation of 1/5 of one percent of voters (0.18%), and every strategic vote for one party is offset by a vote for the other, thus not changing the overall popular vote outcome. With limited resources for grassroots organization, focusing on a small set of ridings where voters face the least amount of emotional resistance is the most effective way to make strategic voting work in Canada.
There is actually a way to guarantee Harper's defeat. Here's how: [Ali Kashani/Medium]
(via Naked Capitalism)