For years scholars and commentators have lamented the inability of Members of Parliament (MPs) to speak with their own voice or to represent their constituents’ preferences when they are at odds with their party and its leader. Party unity in Canada, in fact, exceeds what is found in comparable Westminster democracies (see Christopher Kam, Party Discipline and Parliamentary Politics, 2009). We have never really had a good explanation for why this is the case, and it is here where Jean-François Godbout’s Lost on Division makes its central contribution. He does so by bringing Parliament and elite behavior back into central focus.
Godbout begins by identifying a puzzle: party unity was once much lower in Canada’s Parliament, and it rapidly escalated in the decades after Confederation (see fig. 1.1, pp. 6–7). He makes a complex, multistep argument for why this was the case. Catholic and Francophone MPs could be found in both the Conservative and Liberal Parties immediately after Confederation, which “represented a challenge for party leaders…because language and religious questions always had the potential to divide the parties internally” (p. 184). Indeed, that was soon the case, with tensions flaring due to the inability of the federal government to protect the French language outside of Quebec (e.g., the Manitoba Schools Question). Macdonald’s Conservatives bore the brunt of Catholic wrath, and through a process of replacement “virtually all French Catholic MPs came to be associated with the Liberal Party, while most English Protestant members came to be associated with the Conservative Party” (p. 8).
Godbout argues that this newfound party cohesion allowed party leaders to adopt rules to curb the power of private members, such as by limiting private members’ business, opportunities for amendments, and speaking time (p. 186). These rule changes, in concert with the reforms to limit the distribution of patronage, reduced the incentive of ideological minorities—like Quebec nationalists and prairie populists—to continue working within the Conservative and Liberal Parties (p. 207). The collapse of the first party system was then “not so much the result of fundamental realignment of the electorate… Rather, it was the incapacity of the existing parliamentary structure to accommodate the interests of certain members” (p. 208).
The argument is exceptionally ambitious. To sustain it, Godbout makes use of an impressive new dataset of more than 13,000 recorded divisions from 1867 to 2015 in the Canadian House of Commons and Senate, paired with information on the nature of these divisions and the background of the MPs who cast the votes. The dataset alone is nothing short of a massive contribution to the field of Canadian politics. It allows Godbout to measure both party loyalty at the level of the individual MP and unity at the party level, which he uses throughout the rest of the book as dependent variables.
Godbout first casts doubt on alternative explanations of the development of party unity, such as electoral competitiveness, franchise expansion, and MP socialization (chapter 5). Instead, he finds that the type of division matters the most, with unity decreasing on private members’ bills and on cultural questions that divided MPs by religion (chapter 6). But in what Godbout calls his “cumulative analyses,” he finds nonlinear increases in party unity over time that are not accounted for by the other variables in his models (see figs. 5.5 and 6.3).
This motivates Godbout’s deep dive into two particular eras crucial to his argument. First, he draws our attention to the period just after Confederation, using scaling techniques to show that MPs were divided by religion and language during instances of party disunity (see fig. 7.4), which were disproportionately on votes related to religious and ethnic issues (p. 168). This religion-based disunity largely disappeared from the scene after a rapid realignment of Francophones and Catholics to the Liberal Party in Parliament that Godbout shows preceded, rather than followed, similar sorting in the mass public (chapter 7, pp. 155, 163).
The implications of these findings are considerable. They tell us that parliamentary behavior matters in Canada. It is not the case that party elites are completely at the mercy of the whims of the mass public. They have the power to restructure voter coalitions. And this has considerable relevance for the continued development of Canadian politics, as party elites have polarized along ideological lines (Christopher Cochrane, Left and Right: The Small World of Political Ideas, 2015), which may have trickled down to influence the attitudes and identities of the mass public (Anthony Kevins and Stuart Soroka, “Growing Apart: Partisan Sorting in Canada, 1992–2015,” Canadian Journal of Political Science, 51 [ 1], 2017).
Next, Godbout describes how political leaders adapted to this realignment by reducing the power of private members in the early twentieth century (chapter 8, pp. 188–96). He highlights the lower levels of party loyalty found among Western and Francophone MPs after World War I (see fig. 8.4) and uses historical evidence to link these legislative rule changes to the decision of these MPs to exit the party system (pp. 199–207). In doing so, he highlights the potential of parliamentary rules not only to influence the behavior of political elites but to fundamentally reshape party systems as well.
Godbout concludes his empirical analyses with a chapter illustrating the remarkable growth of party unity in the Senate (chapter 9; see fig. 9.3). Again he highlights the important insights that can be gleaned from this new dataset, indirectly illustrating the salience of ideological cohesion in fostering party unity, because party leaders have considerably less leverage over unelected senators.
Godbout artfully combines historical inquiry with rich quantitative description and analysis. Lost on Division is an engaging and highly accessible read that raises plenty of questions for future research. For instance, how much of these changes in unity were driven by party cohesion rather than discipline, which are observationally equivalent (pp. 79–80)? Under what conditions can legislative rule changes—or parliamentary behavior more broadly—reshape electoral politics and party systems? We know that third parties have emerged sporadically in Canada for reasons unrelated to parliamentary behavior and legislative rule changes (e.g., the Reform Party).
The observational nature of these data makes it difficult to pin down causality, as Godbout acknowledges to his credit (pp. 92–93). For example, was the increasing internal cohesion engendered by the “Catholic Sort” crucial to the implementation of parliamentary rules curbing the power of private members (p. 196)? It appears that disunity drove the implementation of the cloture rule change by Prime Minister Borden, for instance (p. 196). And escalating demands on government time surely mattered in precipitating a desire for more agenda control throughout this period. To put it another way, in a counterfactual world where the major parties remained internally divided by religion, would private members have maintained more of their power? Is the Catholic Sort a necessary or sufficient condition for the legislative rule changes we observe? We can ask similar questions regarding Godbout’s contentions that elite sorting in Parliament influenced voter behavior and that legislative rule changes caused the breakup of the first party system and the subsequent increase in party unity.
These are all very tough questions to answer, but Godbout does an admirable job making his case. Lost on Division is a must-read for scholars of political parties and party systems, and indeed anyone interested in the marriage of historical analysis and quantitative data.
