Try, Try Again? Are Unsuccessful Leadership Contestants Sore Losers?

ABSTRACT

Elections, even intra-party ones, create winners and losers. A number of recent studies have revealed a ‘sore losers’ effect among a number of party actors. The evidence suggests that those who support a losing candidate in an internal party election are significantly less likely to remain active and involved in party politics compared to those who supported the winner. Much less, however, is known about the losing candidates themselves. This paper explores whether losing leadership candidates also exhibit a ‘sore losers’ tendency. Drawing on an original dataset of unsuccessful leadership contestants in three Canadian parties, results reveal that losing leadership candidates do not exit their party en masse but rather they remain generally committed to their party, often seeking re-election during the next general election. The results provide important insight into the behaviour of leadership candidates and provide nuance to the sore losers debate by examining an understudied cohort of party actors: the losing candidates themselves.

KEYWORDS:

• Political parties
• leadership selection
• primaries
• sore losers

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

Requests for access to the data for the purposes of verifying the findings of this article can be addressed to the author.

Notes

1 At the same time, Pruysers (2023, p. 3) notes that many of the theorized benefits of primaries have been ‘short-lived or overstated’ (see also Pedersen & Schumacher, 2015; Wauters & Kern, 2021).

2 In the lead up to the 2022 Conservative leadership election, for instance, leadership candidate Patrick Brown remarked that it was not immediately obvious whether or not he would run for the Conservatives in the next general election if he lost the party’s leadership election (Tasker & Haws, 2022).

3 Note, however, that others are less convinced by the divisive primaries thesis. Piereson and Smith (1975, p. 562), for instance, write that ‘a candidate’s primary election experience bears little relationship to his success in the general election’ – suggesting that a mass withdrawal of party support (i.e., sore losers) does not happen regularly.

4 Carty (1989, p. 125) concludes that ‘running for party leadership pays well, even for losers’. Such a conclusion is based on the observation that many unsuccessful leadership candidates, where possible, receive cabinet positions and other promotions. Likewise, Cross and Crysler (2011, p. 147) write that ‘even in opposition, losing a leadership contest has its benefits. For example, following Stéphane Dion’s election as Liberal leader in 2006, he appointed the runner-up as deputy leader and another challenger as the party’s candidate in one of the Liberals’ safest constituencies’.

5 Included here are the Reform and Alliance as they merged with the Progressive Conservatives in 2003 to form the new Conservative Party of Canada. We focus on these three parties as they routinely run candidates across the entire country (Pruysers et al., 2020), and tend to capture the majority of both seats and votes.

6 It is important to note that these 150 cases are not unique individuals (although the vast majority are). Some leadership contenders contest more than one election, and are therefore included in the dataset more than once. Their data, however, changes with each entry depending on the nature of the race (i.e., age, ballot finish, selection method, etc.).

7 In rare cases the candidate contested a by-election prior to the general election. In these cases, the candidate was coded as remaining active with the party (i.e., given the same code as if they had contested the general election). The logic here is simple: whether it was a general election or a by-election, the losing candidate nonetheless sought election in the party after losing.

8 This, of course, is an imperfect measure. While uncommon, leadership candidates sometimes come from outside of the PPG and therefore it would not be entirely surprising that they did not contest the next general election. Nonetheless, even this is still telling. After all, these individuals are willing to run for the party leadership but not to be a Member of Parliament.

9 While there is a ‘shadow’ cabinet for opposition parties, there are no clear historical records to reliably code this information.

10 We follow the work of Kenig, Cross, et al. (2015) and Cross et al. (2016) and define a ‘primary’ as any selection method in which party members or supporters have at least 50% of the voting authority. Our empirical analysis combines different types of primaries (closed, semi-open, and weighted) into the same category, and treats traditional and mediated conventions as a single category. We are therefore left with a dichotomous variable: primary vs. convention.

11 Interestingly, a number of individuals have sought and won their party’s leadership after being defeated. This includes Brian Mulroney, Ed Broadbent, Jean Chretien, Joe Clark, John Diefenbaker, John Turner, Michael Ignatieff, Paul Martin, and Erin O’Toole. Importantly, many of these individuals went on to become not only party leader but also Prime Minister.

12 While not quite capturing expectations, the multivariate models presented earlier were conducted again with an additional variable that captured an ‘upset’ where a perceived leading candidate did not win. Here the evidence is consistent with the Bernier example (i.e., perceived frontrunners were more likely to exit their party). Since the data span a century of Canadian party politics, however, this result should be interpreted cautiously as information on media perceptions of frontrunners was often sparse and difficult to obtain.

13 Data supporting this can be seen in the Comparative Manifesto Project (Lehmann et al., 2023). During the 2015 election the NDP moderated its positions (left-right score of –9.67) and became outflanked on the left by the Liberals (left-right score of –20.68).

Additional information

Notes on contributors

Scott Pruysers

Scott Pruysers is an Associate Professor in the Department of Political Science at Dalhousie University. His research focuses on questions of participation and representation in party and electoral politics.