Election forecast: dance on a razor's edge

Are you ready to stay up late May 2? Do you have good coffee and plenty of snacks laid in? This may be the election with the highest quantity of uncertainty in my adult experience. The NDP’s dazzling polling gains simply have no obvious recent precedent. I’m not sure a national party has ever made strides of this magnitude and nature in such a bewilderingly short time.

Think about the questions you have to ask to estimate the impact, in terms of Commons representation, of a shift like this; you have to form ideas about the sincerity of the polling subjects’ intentions, the efficiency of the resulting gains in various regions, and the pure logistical power of the party to get out its vote, all while taking into account the activity and the relative positions of three or four other parties.

And then, as if all that weren’t enough, some old flatfoot comes along and tells some TV guys about Jack Layton getting naked in a place he ought not to have been naking around in. Nobody knows what will happen on May 2—and I don’t mean that in the usual perfunctory way. This time, really, nobody has any idea. Having messed around with election models, I could tell you plausible stories that involve the NDP winning 120 seats; I could tell you stories of roughly equal plausibility that put them at 55.

Of course, there are limits. I am just about ready to rule out a Diefenbaker-like cross-country rampage by the Conservatives. I am just about ready to promise that Michael Ignatieff will not look happy on Monday evening. (Though even then: how stupefyingly low are expectations for him at this point?) What I can tell you is what how I would bet, if I had to bet. I believe, halfway through the weekend, that the Conservative push for a majority will come down to the wire. And I think they are a little more likely to get there than not.

I’ve discussed Quebec already. The Tories will probably lose about a half-dozen seats there, but they have a modest nucleus of three or four where their leads are just too huge to be overcome even by the thirty-point gain that the latest Ipsos poll gives the NDP, relative to the last election. The game will be won or lost in Ontario—and what is hard to appreciate until you do some modelling, some farting around with numbers, is just how good the NDP will be at electing Conservatives there.

It looks to me as though, throughout the 25%-35% range of vote share the pollsters more or less have the Ontario NDP in right now, uniformly-distributed gains in NDP support at the expense of the Liberals yield more Conservative seats than NDP seats. Stop and read that again if you need to. It is kind of troubling and counterintuitive and may even threaten the reader’s childlike faith in democracy, but within that 25-35 band, most of the benefit of each additional NDP vote will end up in the hands of the Tories. If you hold everything else equal at some reasonable level, and push the NDP from 25% to 35% at the Liberals’ expense, you see the Liberals lose something like 25 seats. And the Tories, without gaining or losing an actual vote for themselves, get 14 to 16 of those.

Moreover, for the Tories, the optimum level of NDP vote share in Ontario would appear to be above 35%, and more like 38%. The New Democrats don’t start taking seats away from the Conservatives until the point at which the Liberals are exterminated outright; reduced to zero. Which makes sense on a metaphorical level. The NDP and the Conservatives have to chew through the Liberal centre completely to get at each other.

Is the uniform-swing assumption, the assumption that votes will migrate in the same proportions from riding to riding, a safe one here? It’s never all that safe. But I would venture that it is safer than usual in 2011. The shift to the NDP isn’t a result of appeals to particular economic sectors or social groupings that might vary from riding to riding. It is a personality-driven shift; a true mass movement. It is, in part, surely driven by universal human reactions to Layton’s courage. He is fighting an election he might easily have avoided.

In fact… if you’ll pardon a digression, I am not sure this is fully appreciated, and maybe it should be said by somebody who wouldn’t willingly let Jack Layton handle the Grade Five milk money. Layton faced a choice: fight an election now, which is a squalid and exhausting task for a healthy person, or take time to recover from cancer and a broken hip in relative peace. This was as a free a choice as can be imagined. Nobody on the face of earth would have blamed him for taking a break. He decided not to, and whether he did it for the advantage of the party or for the interests of the country, the decision boils down to “He did a brutally difficult thing because he thought it needed doing”. If you ask me, it’s pretty damn admirable even if he just thought selfishly that this was his best chance at being Prime Minister.

Anyway, we are experiencing, as some have called it, a Layton-Mania. However large it ends up being, it should be fairly similar in magnitude from place to place. That’s not good news for the Liberals. I can’t find much good news for the Liberals anywhere I look.

I think it is natural to suppose that the NDP will disappoint in Quebec, but match or exceed its not-quite-so-absurd polling numbers in Ontario, where the party has a real organization and where most of its candidates are not missing or imaginary or celebrating their 14th birthdays on the Moon. I see no evidence of major surprises anywhere else. Even the minor surprises I can envision would work in favour of the Conservatives. Ralph Goodale slipping on a banana peel in Wascana? The Tories doing unexpectedly well in Newfoundland without “ABC”? Gary Lunn shooing Liz May into the Strait of Juan de Fuca? Linda Duncan losing Alberta’s “orange blob” in Strathcona because of Layton’s oilsands hostility?

I’m not predicting any of those things; the point is that they are conceivable, yet my guess doesn’t depend on any of them. I’ve gone on too long about this already, given the enormous likelihood that I am completely wrong, but I have the Tories at around 160 seats. The NDP? ‘Bout half that. The Liberals? ‘Bout half that half.

Looking for more?

Get the Best of Maclean's sent straight to your inbox. Sign up for news, commentary and analysis.