Strategic evaluations

As a follow-up to the federal election, let's take a quick look at just how effective the Vote for Environment strategic voting site was in worked out both in predicting seat results and in assessing voters' strategic options. And it shouldn't come as much surprise that that the projections used to instruct voters how to cast their ballots often varied wildly from the actual election results.

Let's start with its seat chart of projected Con ridings, which was released to much fanfare just in time for the vote.

Vote for Environment's top-ranked riding for strategic voting intended to stop the Cons was Richmond. There, the Cons took 49.8% of the vote for a 19-point margin of victory - and would have won even if all actual Lib, NDP and Green votes had lined up behind one party (as two independents won some votes as well).

#2 was Ottawa South, where the Libs won 49.8% of the vote for a 16-point victory.

#3 was Newton-North Delta, where the combined Lib/NDP vote more than doubled the Cons' vote - making strategic voting unnecessary to beat the Cons.

In #4 West Vancouver – Sunshine Coast – Sea to Sky Country, the Cons took more than 44% of the vote for an 18-point win.

Which means that #5 Mississauga-Erindale was the first riding on Vote for Environment's list where a strategic vote would actually have had the potential to tip a tight race.

#s 6 through 10 look better based on the site's intention, including only one relatively comfortable Lib win in Vancouver-Quadra to go with Saanich-Gulf Islands, Vancouver Island North, Edmonton-Strathcona and Oshawa where strategic voting could have made a difference (and may actually have in the case of Edmonton-Strathcona).

From there on, the results are likewise a mixed bag at best, featuring a range of ridings from ones like Parry Sound-Muskoka and Blackstrap where the Cons won an outright majority of the votes, to St. John's South - Mount Pearl where the Cons finished a distant third at 12.6% of the vote behind a tight two-way Lib/NDP race.

But what if one looks from the other side to see whether Vote for Environment's main recommendations actually anticipated which ridings would be the closest ones? (See Election Pundit Queries on this page.)

There, one has to go all the way to #2 on the list of the closest actual results to find Kitchener-Waterloo - which Vote for Environment pegged as a safe Liberal seat which wouldn't require a strategic vote. While it's a little ways down the list before another riding completely escaped Vote for Environment's attention, Brampton-Springdale and Vancouver South both ranked among the top 20 closest ridings in the country after being listed as safe seats. And perhaps most damningly, another top-20 closest race in South Shore - St. Margaret's actually included a Vote for Environment recommendation to vote for the Lib who finished a distant third behind the tight Con/NDP battle at the top.

Now, the above isn't to say that Vote for Environment didn't mostly have some basis for its recommendations or non-recommendations (however sketchy South Shore - St. Margaret's may have been all along). But it does provide inescapable evidence that even a relatively sophisticated attempt to direct voters strategically simply can't be expected to anticipate how an election will actually play out. And the lesson is one that voters should keep in mind in deciding whether or not to put aside their actual party preferences when they go to the polls.