Skip to main content
andrew steele

A couple walks up to a polling station in Saanich-Gulf Islands, near Sidney, B.C., to vote in the federal election on Oct. 14, 2008.Deddeda Stemler

Polls are one thing, but elections are another. Mandates are made and broken when voters cast a ballot, not when they answer a survey.

But - as with polls - there is a temptation to read too much into a relatively small check-up by the electorate.

South of the border, it's election week in the United States.

However, this is an "off-year" so the number of major races is very small. The main action on Nov. 3 sees votes for governors of two states, Virginia and New Jersey.

Both states could see the party in power switch, meaning the Democrats lose. This is likely in Virginia and too close to call for New Jersey.

If the two state-wide offices are lost by the Democrats, there will be much hooting and finger-pointing by the Republicans. The GOP is desperate to show the end is nigh for Barack Obama and another 1994-style landslide in store for Congress next year.

However, the reality is off-year elections tend to have little predictive power for election results.

Since 1978, Virginia has elected a governor of the opposite partisan stripe as the President, every single time. Not once has a President successfully retained Virginia after winning his own election.

New Jersey is a similar, if slightly less direct, negative correlation.

The first elections held under Presidents George H.W. Bush, Bill Clinton and George W. Bush all saw the Governor's mansion change hand. Before that the record is less direct but still saw a similar flip away from the Republicans when Eisenhower became President.

The lesson is simple: don't read too much into off-year elections in the United States.

Closer to home, the Harper government faces four by-elections across the country.

Just like in the U.S., none should be taken as a referendum on the government.

The British Columbia by-election is getting the most attention, with some calling it a referendum on the Harper government's initiation of a harmonized sales tax with the B.C. Liberals.

However, the party attempting to make this an HST referendum is the same party that held the seat, the NDP. This tactic of pretending the maintenance of the status quo is some type of massive repudiation doesn't really hold water. If the NDP vote experienced a massive and inexplicable boost, then perhaps there is some juice in the HST issue. The more likely situation is the opposition is better able to rally its vote when the governing party's supporters aren't worried about who actually runs the country.

Cumberland-Colchester-Musquodoboit Valley is another unique circumstance. Bill Casey's resignation may lend some gas to the opposition, but the ground is stony. This seat has only gone for someone other than a Tory once, in the nightmare year of 1993. Other than that single Liberal MP, the record is unbroken Conservative representation.

Montmagny-L'Islet-Kamouraska-Rivière-du-Loup has been a safe Bloc Quebecois seat since 2004. This is the most Catholic, most French-speaking and most French-as-mother-tounge riding in the country. Hochelaga is a similarly strong Bloc seat, one of its best in Montreal. The seat, and those that it was formed from before the 2004 election, were solid Bloc right back to 1993. In fact, part of the seat includes Laurier-Ste Marie, where Gilles Duceppe first won for the "ad-hoc temporary rainbow coalition."

This is not exactly ground where one expects a true test of public opinion. One seat has changed hands between the Conservatives and NDP recently, while the others are dictionary-definition safe seats.

Basically, it is entirely reasonable to expect nothing to change. And even a shift to the Conservatives in B.C. could mean only a few hundred votes going right instead of left.

As such, any pronounciations of grand impacts on the Harper administration due to maintenance of the status quo are almost definitely overblown.

What will be more telling are the detailed breakdowns of the vote. Who got their vote out and who didn't? What trends were present across the country?

As with so many things, the truth is in the details, not the topline result.

Interact with The Globe