September 13, 2010 3:52 PM

Since the PCs seem to be holding on to a significant lead in the last couple of days of polling in the Telegraph-Journal, a few observers will probably be tempted to dust off some adjectives such as "historic" or "unprecedented," just in case they need them on election night. This is not to say the PCs have the campaign in the bag. But in case they do remain on track, let's review three patterns that could be relevant to any discussion of a Tory victory.

1. New Brunswicks never defeat a one-term government. This is true, at least post-Confederation. But watch your terminology: there have been one-term premiers who have died in office, retired or moved on to Ottawa. It's the party that wins a mandate and forms a government, and in those terms, no party has been ousted from power after a single mandate since Confederation. A Liberal defeat this month would be the first such occasion.

2. The kid always wins. For more than five decades, when a party in power has been defeated at the polls, it has been at the hands of a leader under the age of 40. Louis Robichaud, Richard Hatfield, Frank McKenna, Bernard Lord and Shawn Graham were all in their 30s when they defeated incumbent governments. They defeated premiers who were always older than 40, but it's the age of the challenger, not the incumbent, that has set the pattern. (In other words, in any scenario in which the challenger was older than 40, the incumbent has won.) Alward would be the first 40-plus leader to defeat a sitting government since Hugh John Flemming in 1952.

3. The party seeking a second term loses support. We haven't had a chance to deploy our vast team of Spin Reduxit researchers and interns to fully verify this, but Bernard Lord, among others, has pointed it out, and it sounds like he's right when you look back five decades. Robichaud's Liberals, Hatfield's PCs, McKenna's Liberals, and Lord's PCs all saw their popular vote go down in their second victory, relative to their first victory.

Lord was talking this up because there was a rough tie in the popular vote last time, suggesting Graham can't lose ground this time without losing power. What Lord forgot is a drop in the popular vote for the governing party doesn't preclude a drop in the popular vote for the second-place party as well. That's what happened in 1991, when McKenna's vote dropped and the PC numbers dropped as well. (Despite that, the PCs emerged with three seats, up from zero in 1987, showing that popular vote fluctuations can have unpredictable results.) Graham's popular vote can go down in this election, but if the PCs' goes down further, the Liberals can still win.

So rules no. 1 and 2 mitigate against a PC win, while rule no. 3 is practically meaningless. Of course, precedents themselves don't mean anything. In 1988, when Vice-President George H.W. Bush ran for president, much was said about the fact that no VP had been elected directly to the presidency since Martin Van Buren. It didn't stop Bush from winning.
-- Jacques Poitras

[an error occurred while processing this directive]