I was having a perfectly nice Monday morning doing what most normal people do – blog about obscure electoral statistics. After posting about the minimum threshold historically needed to secure a majority in Canada, ink-stained wretch Vaughn Palmer entered the conversation on Twitter to make things more complicated.
My point was that no majority government has ever been formed in Canada over the past century with less than 38.5% of the popular vote. Fairly straightforward, but Vaughn wanted to belabour it.
So, fine, let’s get into it – when Jean Chrétien won a majority with 38.5% in 1997, he had some help. The right was hopelessly splintered. Despite a low popular vote, the Liberals had a 19-point margin over the second place Reform Party, the sixth-largest margin-of-victory, in terms of popular vote between 1921 and 2015. Plus, the Liberals annihilated the opposition in Ontario. They won virtually every seat. Let’s also remember the NDP was in the serious doldrums nationally in the 1990s. It was easy street for the Chrétien Liberals. Ridiculously easy.
Of course, Vaughn couldn’t leave it at that. He had to consult his groaning book shelves for more statistical peculiarities.
By extending this barely-read Twitter thread, Vaughn was making me think I needed to do a deeper statistical dive.
And I did. Is there a pattern between polarization and majority/minority governments? After a pile of work, the answer is… not really.
Here is a chart that shows the combined amount of the top 2 federal political parties (popular vote) from 1921 to 2015. The blue dots represent majority governments and the black dots represent minority governments. Some majorities happen when there is low polarization and some
The ‘extreme polarization’ occurred in 1925 and 1926 when William Lyon Mackenzie King and Arthur Meighen waged battle, and in 1930 when R.B. Bennett prevailed over Mackenzie King, peaking at 93% (combined votes of Liberals and Conservatives). In spite of the polarization, Mackenzie King and Meighen both failed to win a majority, with the Progressives holding the balance of power.

William Lyon Mackenzie King: “Majorities are hard”, he might have said. He finally got one on his 5th try.
Extreme polarization flared up again in 1958 when the Progressive Conservatives and Liberals combined for 87% of the vote (mainly PC). That’s the last time any two parties combined for over 80%.
The Liberal – PC oligopoly held between the 70% to 80% level from 1962 to 1988. In the 1990s, all hell broke loose when the PC coalition shattered with the Bloc Québécois going on a five election run of 10% to 13% of the national vote, and the Reform Party devouring the PC’s starting in western Canada. For six elections between 1993 and 2008, the top 2 level ranged from 58% to 66%. Very low polarization with many parties receiving double-digit popular vote amounts.
In 2011, the top 2 level rose above 70% and was 71% in 2015.
While this is kind of interesting (to me) about federal polarization, it doesn’t really say much about likelihood of minority and majority governments.
Vaughn then helpfully recounts how the BC NDP did better in the popular vote – losing – than when they actually won. True, Glen Clark with 39% and Mike Harcourt with 41% won majorities, while Bob Skelly at 43% was crushed. Difference was that Skelly faced a dominant Social Credit party while Clark and Harcourt faced a split opposition.
So, I looked at it further, putting aside family time and personal wellness, to deal with Vaughn’s haranguing.
In the past 29 elections, there were 12 where the #1 party won by about 12% of the popular vote or more. All of those were majorities.
Then there’s a set of 8 elections where the winning party had a popular vote edge of about 7.5% to about 11.5%. Half of those were majorities, half were minority governments.
Finally, there is a set of 9 close battles where the party with the plurality of seats won the popular vote by 7% to minus 4%. Huh? Yes, three governments in the past century lost the popular vote but won the plurality of seats – Mackenzie King in 1926, John Diefenbaker in 1957, and Joe Clark in 1979. (I should add that Meighen won the popular vote and the plurality of seats in 1925, but Mackenzie King hung on with support of the Progressives, ultimately leading to the King-Byng Affair). Of those 9 elections, there was only one majority: R.B. Bennett in 1930.
Moral of the story: in #elxn43, the margin between the two parties appears to be pretty close. The public polls indicate a way lower spread than 7%, at this time. History tells us that there is strong likelihood of a minority government if it is a tight race, especially if third parties have strongholds where they have a greater chance of winning.
I think we all knew most of that already, but Vaughn has succeeded in sparking a tour through dusty old election results. Ah, it wasn’t so bad.
***
See below for stats:
Table 1: Results of top 2 parties (1921- 2015); sorted by difference in popular vote between party with plurality of seats, and second place party (pop vote)
Top 2 | Margin | Plurality | 2nd | Majority | |
1993 | 59.9% | 22.55% | 41.24% | 18.69% | y |
1940 | 80.6% | 22.08% | 51.32% | 29.24% | y |
1984 | 78.1% | 22.01% | 50.03% | 28.02% | y |
1958 | 87.4% | 19.92% | 53.67% | 33.75% | y |
1949 | 78.8% | 19.50% | 49.15% | 29.65% | y |
1997 | 57.8% | 19.11% | 38.46% | 19.35% | y |
1953 | 79.5% | 17.41% | 48.43% | 31.02% | y |
2000 | 66.3% | 15.36% | 40.85% | 25.49% | y |
1935 | 74.5% | 14.84% | 44.68% | 29.84% | y |
1968 | 76.8% | 13.94% | 45.37% | 31.43% | y |
1945 | 67.4% | 12.16% | 39.78% | 27.62% | y |
1980 | 76.8% | 11.89% | 44.34% | 32.45% | y |
2008 | 63.9% | 11.39% | 37.65% | 26.26% | n |
1921 | 71.1% | 11.20% | 41.15% | 29.95% | n |
1988 | 74.9% | 11.10% | 43.02% | 31.92% | y |
2011 | 70.3% | 8.99% | 39.62% | 30.63% | y |
1963 | 74.3% | 8.68% | 41.48% | 32.80% | n |
1965 | 72.6% | 7.77% | 40.18% | 32.41% | n |
1974 | 78.6% | 7.69% | 43.15% | 35.46% | y |
2015 | 71.4% | 7.58% | 39.47% | 31.89% | y |
2004 | 66.4% | 7.10% | 36.73% | 29.63% | n |
1925 | 85.9% | 6.39% | 46.13% | 39.74% | n |
2006 | 66.5% | 6.04% | 36.27% | 30.23% | n |
1972 | 73.4% | 3.40% | 38.42% | 35.02% | n |
1930 | 93.3% | 2.29% | 47.79% | 45.50% | y |
1962 | 74.2% | 0.25% | 37.22% | 36.97% | n |
1957 | 79.0% | -2.00% | 38.50% | 40.50% | n |
1926 | 88.3% | -2.45% | 42.90% | 45.35% | n |
1979 | 76.0% | -4.22% | 35.89% | 40.11% | n |