1 Courtney, J., “Reinventing the Brokerage Wheel: The Tory Success in 1984,” in Penniman, H., ed., Canada at the Polls, 1984 (Durham: Duke University Press, 1988), 201.Google Scholar

2 Blake, D. E., Carty, R. K. and Erickson, L., Grassroots Politicians: Party Activists in British Columbia (Vancouver: University of British Columbia Press, 1991), 99.Google Scholar

3 “P.C. Leadership Convention,”s CBC Newsworld, February 1991.Google Scholar

4 Courtney, J., The Selection of National Party Leaders in Canada (Toronto: Mac-millan, 1973).Google Scholar

5 For a good review of the debate (but one unsympathetic to the reform position) see Perlin, G., “Leadership Selection in the PC and Liberal Parties: Assessing the Need for Reform,” in Thorburn, H. G., ed., Party Politics in Canada (6th ed.; Scarborough: Prentice-Hall, 1991), 202–20.Google Scholar

6 Latouche, D., “Universal Democracy and Effective Leadership: Lessons from the Parti Quebecois Experience,”Google Scholar and Woolstencroft, P., “‘Tories Kick Machine to Bits’: Leadership Selection and the Ontario Progressive Conservative Party,”Google Scholar in Carty, R. K., Erickson, L. and Blake, D. E., eds., Leaders and Parties in Canadian Politics: Experiences of the Provinces (Toronto: Harcourt Brace Jovanovich, 1992), 174–202 and 203–25Google Scholar, respectively.

7 The record starts with the Manitoba Liberal leadership convention in April 1961 and goes through the Alberta Progressive Conservative selection of December 1992. Though several of those conventions chose a woman as party leader, of them, only British Columbia's Social Credit leader Rita Johnson became subsequently a provincial premier. It should be noted that at the time when the data analysis was done at the end of 1992 several of the new party leaders had still to be tested at the polls.

8 The one exception is the Prince Edward Island New Democratic party. With only 154 members in 1989 it is difficult locali it a major party.

9 Note that we do not have systematic data on the direct causes for the leadership changes in terms of the reasons why the outgoing leader was giving up office. For the purposes of our argument it is not necessary to assume that the outgoing leader was driven out in the conviction that a convention boost would follow. Certainly the politics of leadership exit deserves a study in its own right. By our preliminary count, of 69 leaders of major (government or opposition) parties who left their position between 1960 and 1990, 3 (4%) died; 16 (23%) suffered both personal and party electoral defeat; 21 (30%) suffered a party, but not personal seat, defeat; 13 (19%) were more or less pushed out by caucus or party executive; and the other 16 (23%) left for myriad personal reasons.

10 Wiseman, N., “From Jail Cell to the Crown: Social Democratic Leadership in Manitoba,” in Carty et al., eds., Leaders and Parties, 147–69.Google Scholar

11 Courtney, J., “Leadership Conventions and the Development of the National Political Community in Canada,” in Carty, R. K. and Ward, W. P., eds., National Politics and Community in Canada (Vancouver: University of British Columbia Press, 1986), 98.Google Scholar

12 The decision rules of party leadership conventions are fundamentally similar. Winners require an absolute majority (50% + 1 ) of the votes cast and a series of successive ballots are taken until that result is obtained. On each ballot the candidate with the fewest votes on the preceding ballot is automatically dropped and other candidates are permitted to withdraw (no new candidates can be put for ward). Parties do vary in their rules for eliminating candidates and in some circumstances more than one candidate may be forced to withdraw. In practice, poorly placed candidates often rush to withdraw after the first ballot, though the membership vote-convention system used by the Ontario Liberals led candidates to stay on the ballot.

13 For the record see Courtney, J., “Provincial Party-Leadership Conventions and Votes: 1961–1991,” in Carty et al., eds., Leaders and Parties, 227–35.Google Scholar

14 The 1991 British Columbia New Democratic party is something of a case in point. In the election of that year, under their new leader, Mike Harcourt, they won a smaller vote share than they had in 1986 (when they formed the opposition) but they still managed to win a majority of seats and form a government.

15 The one exception was the Manitoba New Democratic party selection of Gary Doer noted above. He declined to take office as premier, preferring to await the judgment of the electorate. As a result he has yet to be premier.

16 The quotation is from J. Neilson, Social Credit ex-cabinet minister and some-time leadership candidate, in The Globe and Mail, April 1, 1991, A1.Google Scholar

17 Of the 47 minor-party leaders chosen, 29 did not get elected to the legislature. That represents a 62 per cent failure rate.

18 We note, however, that their existing seat share made no difference to the outcome for opposition parties.

19 Clarkson, S., “The Dauphin and the Doomed: John Turner and the Liberal Party Debacle,” in Penniman, ed., Canada at the Polls, 99.Google Scholar

20 The only major-party leaders chosen more than three years before the next election, who then managed to win it, were Mike Harcourt (British Columbia New Democrat) and Roy Romanow (Saskatchewan New Democrat). Both had to wait until a very unpopular government had run out of time before the election was called. As noted above, both also were exceptions in that they boosted the success rate of acclaimed leaders.

21 However it is true that the legislatures in the small Atlantic provinces are relatively small so seat turnovers will translate to larger percentage change than similar turnovers in the larger provinces.