Vote splitting in Canada
Canada's 1993 election took the governing Progressive Conservatives from 169 seats to two, while the right's vote moved mostly in Quebec and its seats moved everywhere.
n 1993 a new right party appeared beside Canada's governing conservatives. What happened to the governing party's vote, to its seats, and to the right's total?
Library of Parliament riding results, 1867–2010
The candidate-level results for every Canadian federal riding from 1867 to 2010, used here for the riding-by-riding arithmetic.
Reform ran no candidate at all in Quebec in 1993. The right's Quebec vote fell from 1,844,279 to 506,683, down 72.5%, and its Quebec seats went from 63 to 1.
Outside Quebec the right's vote rose 3.4% in absolute terms while its seats fell from 106 to 53.
Ontario's combined right share was 37.75% in 1993, 37.95% in 1997 and 38.04% in 2000; the right's Ontario seats went 46, 1, 1, 2 across 1988, 1993, 1997 and 2000.
Adding the two right parties' votes together and comparing with the leading non-right candidate gives the right 101 seats in 1993, short of the 148 needed for a majority.
Inter-Parliamentary Union, 1993
The Inter-Parliamentary Union's 1993 record of Canada's vote and seat totals, standing in for Elections Canada's 1993 synopsis, which is print only.
Unsupported block: pub.terrain.quote
Flanagan (2022)
A short book on the 1993 election by someone who was inside the Reform party.
Unsupported block: pub.terrain.quote
Unsupported block: pub.terrain.quote
MacIvor (1999)
An Elections Canada research paper on proportional alternatives to first-past-the-post, using the 1997 result as its base.
Unsupported block: pub.terrain.quote
Unsupported block: pub.terrain.quote
Thompson (2017)
A doctoral thesis on how Canada's two right parties finally merged, ten years later.
Unsupported block: pub.terrain.quote
Whether a united right would have won in 1993 is modelled, not measured: the model credits the merged party with every vote cast for either predecessor, and assumes no voter would have behaved differently.
Bélanger & Godbout (2004)
A conference paper on why the two right parties did not merge for a decade.
Unsupported block: pub.terrain.quote
Law Commission of Canada (2004)
A Government of Canada law-reform report on the electoral system, written after two more elections of the split.
Unsupported block: pub.terrain.quote
Synthesis
The seat penalty belongs to first-past-the-post. Victoria's count transfers a first preference rather than discarding it.
No peer-reviewed causal estimate of the split's cost was found. The available objects are the additivity model, a simulation of a different electoral system, and the before-and-after observation that the merged party polled fewer votes in 2004 than the two halves had in 2000.
The Quebec collapse ran to a secessionist party with no Victorian counterpart, so that half of the story has nothing to compare to here.
What the post-merger party actually polled is an indication of how much the model overshoots, not a measurement of the split's cost.
Linked notes
/notes/preference-flows
/notes/the-front-republicain
/notes/what-counts-as-evidence
Did you enjoy this article?
Recommend it — Standard Reader surfaces well-loved writing to more readers across the network.