Cold Hard Wonk

No sentiment but politics

Mixed News, At Best

Posted by JJ in Federal Elections, Golden Tacks (Wednesday October 18, 2006 at 9:10 pm)

The old line is only half true: the only poll that matters is the one on election day. As the Chilly Wonk recently pointed out, the most significant nationwide numbers aren’t necessarily relative standings. Much more significant is the absolute vote count, only partly because voter turnout can have as significant an impact on the outcome of an election; and polls taken months away from an election can’t reveal much about those numbers.

Which is why much of the analysis offered by the Globe and Mail on a poll released today isn’t particularly useful. Consider:

Until now, Conservative support has been holding steady at around 36 per cent, prompting some strategists to argue that the party now has a new and higher electoral base from which to work.

Which the article presents to imiplicitly disprove via the polling numbers. But as the Ice-Cold Wonk showed in that previous article, the change in the Conservative base came in 2004, not 2006, with the real merger of the two parties. Moreover, the Tory increase in 2006 didn’t come from tugging at Liberal voters — it could be completely accounted for by the increase in voter turnout. Base/bonus voter arguments aren’t the issue.

What is? Movement. To work with the polls, which only offer relative performance, we’d need to start with the assumption that total votes remain roughly constant — the polls tell us nothing about that. Just as importantly, movement in regions has to be closely compared with the real opportunities for seat gains to appreciate how that movement translates into electoral victory. The Globe and CTV don’t do that. Fortunately, it’s not that hard to do.

Let’s start in Quebec (since the poll, irritatingly, ignores the Maritimes as a separate category). In 2006, a rise in Tory tides resulted in a gain of 10 seats in the province, eight from the Bloc Quebecois, and two from the Liberals. A Liberal gain of 7% in Quebec, as the poll suggests comes directly from the Tories, would result in a transfer of 105,000 votes from one party to the other. That number could be swallowed entirely by a reversal of the shift in votes in the eight seats taken from the BQ, which would gain the Liberals nothing — they trailed the BQ by thousands of votes in each of those ridings, and the BQ’s numbers seem to also have risen in the polls. The Liberals lost only eight seats in Quebec in the last election. Of the six lost to the BQ, only three were lost by less than 3,000 votes. So if the movement in Quebec suggested by the poll is right, the Liberals could still wind up gaining no seats. At the least, it would suggest that the Bloc stands to regain between six and eight seats from the Tories. The Liberals might retake five seats, but it will take more than the movement this poll suggests to grow past those numbers.

The Ontario numbers have shifted primarily because of movement to the Green Party from both Liberals and Conservatives. But the Tories have lost less from the change than have the Liberals; and if the Liberal bleeding comes from continued losses on their leftist flank, it may forebode a poorer Liberal performance in the urban areas where those voters are concentrated. If so, these slight changes are unlikely to bring any real change in seat totals in the most populous province in the country.

The “West” is so diverse a region that polling covering changes in that area can mean many things. The only party shown to have lost support since the election is the Conservatives — to the tune of seven percent. The Greens have picked up five points of that, most likely in BC. The Liberals picked up four points of that (impossible, true, but as the poll shows a total of 99% at the election at 101% on October 18th, that’s the way it is). The Liberal gains probably include both a mild improvement in Saskatchewan and Manitoba and an upswing in BC.

Given those numbers, it might be reasonable to expect the Liberals to reclaim Winnipeg South and possibly retain Desnethe-Missinippi-Churchill River (taken in 2006 by a mere 70 votes). In BC, the change will be harder to predict. While Green strength largely saps the Tories’ strength in BC, only in Fleetwood-Port Kells did the Conservatives win by less than 4,000 votes. Liberal strength might lead to a recapture of Victoria from the NDP, but with a 6,800 vote lead to erase, that could prove difficult. The genuinely competitive three-way races in BC don’t allow any obvious conclusions from such vague polling data.

Which suggests Tory losses of up to twelve seats, dragging their tally down to 113, Liberal gains of up to eight seats, bringing them to 110, and Bloc gains of five seats, taking them to 56. The NDP, with no change to speak of, would remain at their present level of 29. Even if the Liberals take back every seat from the Bloc (which would mean overcoming 3,000, 4,000 and 5,000 vote leads), they would only manage to equal the Tory total, in which case the Conservatives could remain in government, still needing only Bloc support to stand pat.

What’s it all mean? If the Tory slide is due to dissatisfaction with government policies, then barring complete disasters, there isn’t much farther to go — there’s already public disclosure of those policies. That means that a Liberal attack strategy for the election can’t bring them much more than a very weak minority position. Both they and the Conservatives must come up with a strong election message to have any chance of making significant gains.

Just as importantly, it signals that Liberal plans cannot rest on Ontario. The lack of change in that province is perhaps the most significant detail for Liberal eyes. Hopes for future majorities need to rest on broadening efforts across the country. A real effort to take more BC races is crucial, as is a plan to recover lost ground in Manitoba, Saskatchewan, and the Maritimes.

For the Conservatives, it shows that they can’t rely on controlling the agenda. The need to respond to events around them prevents them from running the kind of careful, controlled campaign that brought such success in the 2006 election. While there’s a chance they can retake the momentum, it will be more difficult than taking it originally was. The good news for them is their continued strength in Ontario; but a real breakout in either the Maritimes or BC is crucial to their future hopes — they’ve nowhere else to look to.

Lest we forget, growth in Green numbers is hardly a bad sign for that party, either. What they’ll do with it is a different question.

It’s not an easy way forward to majority status for any party. That’s perhaps the best news of all — it means, whatever else the parties had in mind, a real fight for Canadians to watch.

No comments for Mixed News, At Best »

No comments yet.

Leave a comment


(required but not published)

RSS feed for comments on this post. TrackBack URI