Baloney meter: Do byelection voters punish the governing party?

"We all know majority governments don't typically fare well in byelections; the opposition parties do. And these rounds we don't expect to be any different." Conservative party communications director Cory Hann, May 11, 2014

Prime Minister Stephen Harper last month called four byelections in Toronto's Trinity-Spadina and Scarborough Agincourt and in Alberta's Fort McMurray-Athabasca and Macleod for June 30.

Ever since, Conservative party spokespeople have repeatedly tried to lower expectations by citing the maxim that byelections tend to be hard on the governing party.

But while there used to be truth in that old saw, it hasn't reflected reality for at least 20 years.

Spoiler Alert: The Canadian Press Baloney Meter is a dispassionate examination of political statements culminating in a ranking of accuracy on a scale of "no baloney" to "full of baloney" (complete methodology below).

This one earns a rating of "a lot of baloney" the assertion that byelections don't favour governing parties has not held true since at least 1993.

And it certainly hasn't been true for Harper's Conservatives, who've gained five seats, held on to five others and lost only one previously held seat since 2006.

THE FACTS

Eric Grenier, who analyzes opinion polls and election results for his website, ThreeHundredEight.com, has found virtually no difference between how the governing party and opposition parties fare in byelections.

Since 1945, he says the ruling party has won 49 per cent of byelections, opposition parties 51 per cent.

Opposition parties have retained their seats 69 per cent of the time while the ruling party has hung on to its seats 66 per cent of the time.

However, Wilfrid Laurier University political science professor Barry Kay argues that the win-loss record is less relevant than what happens to the governing party's share of the popular vote in byelections, compared to how they fared in the previous general election in those ridings.

On that score, a study Kay did several decades ago of 147 byelections from 1940 to 1980, found that the governing party did indeed suffer during byelections.

When the Liberals were in power, their vote share dropped an average of 5.65 percentage points in byelections, Kay found.

When the Conservatives were in power, their vote share dropped an average of 8.48 percentage points.

Combined, the governing party dropped an average of 6 points during byelections over that 40-year period.

"There's a logic to the fact that it's a free kick at the government," Kay says, noting that byelections give voters a chance to protest against the ruling party without fear of causing a change in government.

"Remember, byelection turnout is very low and the people who turn out in byelections are frequently more likely to be negative about the government than positive."

Kay's findings appear to have held during Liberal Pierre Trudeau's final term in office, from 1980 to 1984, and Conservative Brian Mulroney's two terms from 1984 to 1993.

Overall, the Trudeau Liberals lost 13 previously-held ridings in byelections, held onto 11 and gained four opposition-held ridings.

Under Mulroney, the Conservatives lost four ridings, held two and gained none.

But Jean Chretien turned the maxim on its head when he came to power in 1993.

During his 10 years at the helm, the Liberals held 16 ridings, gained four and lost only three.

That trend has continued under Harper: his Conservatives have hung on to five ridings, picked up five from opposition parties and lost only one -- Peter Penashue's seat in Labrador.

Even using Kay's preferred vote share measure, byelections during the Chretien and Harper eras buck the pattern he found from 1940-80.

Under Chretien, the governing Liberals' share of the vote actually increased by an average of 1.96 percentage points over 32 byelections.

Under Harper, the ruling Tories' vote share has dropped on average less than one point over 25 byelections.

(Mobile users, see the chart here)

THE EXPERTS

Kay, whose study involved a larger sample of byelections over a longer period of time, was intrigued to find that Chretien and Harper era byelections don't seem to have stuck to the same pattern he found from 1940-80.

"Whether or not that is a (new) trend is open to speculation," he said.

However, Grenier concludes the notion that byelections don't favour the ruling party is "mostly just spin."

That said, he notes that the longer in the tooth governments get, the worse they appear to fare in byelections.

For instance, during the Chretien era, the Liberals lost only three previously held ridings -- all in Chretien's final two years as prime minister.

"It isn't good for majority governments, often when they're on their way out," Grenier says.

On that score, the most recent byelection results might bode ill for Harper's Conservatives.

They lost their first Tory-held seat (Labrador) last year and saw their share of the vote plummet by 12 to 19 points in several previously safe Tory seats.

THE VERDICT

While there was some truth years ago to the maxim that byelections are unkind to the governing party, that has not been the case for at least 20 years.

For that reason, there is a "lot of baloney" to the Conservative party's claim that the ruling party typically doesn't fare well in byelections.

METHODOLOGY

The Baloney Meter is a project of The Canadian Press that examines the level of accuracy in statements made by politicians. Each claim is researched and assigned a rating based on the following scale:

No baloney the statement is completely accurate.
A little baloney the statement is mostly accurate but more information is required.
Some baloney the statement is partly accurate but important details are missing.
A lot of baloney the statement is mostly inaccurate but contains elements of truth.
Full of baloney the statement is completely inaccurate.

SOURCES

Elections Canada
Parliament of Canada
Barry Kay, Wilfrid Laurier University
Eric Grenier