Introduction
Scholars frequently rely on the biographical details of legislative candidates to make inferences about electoral performance, particularly related to gender (Everitt and Tremblay, Reference Everitt and Tremblay2023; Sevi et al., Reference Sevi, Arel-Bundock and Blais2019; Thomas and Bodet, Reference Thomas and Bodet2013; Tolley, Reference Tolley2015; Trimble and Arscott, Reference Trimble and Arscott2003; Trimble et al., Reference Trimble, Arscott and Tremblay2013) and occupation (Mechtel, Reference Mechtel2014; Sevi et al., Reference Sevi, Blais and Mayer2020). The preponderance of studies in Canada has focused on the analysis of biographical information at the federal level. Indeed, recently published datasets have provided researchers with the ground-breaking opportunity to observe the differences in representation over time (Taylor et al., Reference Taylor, Lucas, Kirby and Hewitt2023; Johnson et al., Reference Johnson, Tolley, Thomas and Bodet2021; Sevi, Reference Sevi2021).
This note consolidates publicly available information into a dataset that traces the political trajectories of 7,200 legislators elected to the parliaments of Ontario, Quebec, New Brunswick and Nova Scotia since the establishment of their colonial assemblies in the mid- to late-eighteenth century.Footnote 1 The information included should allow researchers to identify the characteristics of parliamentarians that foster or impede their legislative career by identifying who is being represented. Considering that Canada has an electoral history going back to the mid-eighteenth century, it is surprising to note that very little has been written on the comparative representation of provincial legislators over time. This dataset not only fills that gap but is also designed for easy updates to incorporate more recent biographical features, including 2SLGBTQI+ people, visible minority status, and First Nation, Inuit and Métis origins, as this information becomes increasingly available publicly.
This dataset should offer researchers the opportunity to study various dimensions of representation, including gender, language, religion, elections, parliamentary careers and parties in parliament. It follows the publication of recent datasets that focus on federal politicians (see Taylor et al., Reference Taylor, Lucas, Kirby and Hewitt2023; Sevi, Reference Sevi2021) by looking instead at the provincial level (see Sevi et al., Reference Sevi, Blais, Arel-Bundock, de Greus, Tolley, Goodyear-Grant and Loewen2021). This new dataset should therefore enhance our understanding of the multi-level nature of Canadian politics and the extent to which provincial trends in representation align with national ones. Such a comprehensive data collection is a necessary first step in understanding over-time representational patterns in Canada's four oldest provinces.
Data Collection and Variables
Data were collected for the legislatures of Ontario, Quebec, New Brunswick and Nova Scotia. These provinces were chosen not just because of their historical importance, but because they already maintained established political institutions such as responsible government, political parties and defined patronage networks that existed well before Confederation in 1867.
Data collection begins in Ontario and Quebec in 1792, 1785 in New Brunswick and 1758 in Nova Scotia. Each elected legislator represents our unit of analysis and is identified in the dataset by their name and surname in addition to a unique identification number (ID). Legislators have an identified year of birth and year of death. This biographical information was parsed from various sources (see Table 1). First, we consulted each legislatures’ website, notably official biographies about current and past parliamentarians.Footnote 2 We also utilized the Dictionary of Canadian Biography (2023) to complete missing information, especially for elected officials who occupied public office before Confederation.Footnote 3 In cases where the biographical information was still lacking, we looked online using a search engine to fill the holes, notably on Wikipedia pages (when references were included). Research assistants began by collecting this information through these various sources. They first coded biographical information for 50 legislators in each province. These results were then verified by the authors to ensure that there were no disagreements between the authors and the assistants.
Data Sources by Province
Year of birth is readily available for most legislators (92.3%) but is mostly lacking for those legislators elected very early on and for those presently in office who might not have yet made their biographic information fully public. Year of death suffers from the same problem given the, sometimes, spotty records from the pre-Confederation period and that not all legislators have yet to pass.
Further biographical information includes the legislator's gender and place of birth, listed by province (if Canadian) or country (if foreign-born). We recognize the significant limitations of coding gender identity using archival data. Though research has shown that individuals who identify with genders other than men and women have now been elected to Parliament (see Everitt and Tremblay, Reference Everitt and Tremblay2023; Malloy, Reference Malloy2023), we lack access to this information for most of the period under study. We thus decide to report a dichotomous classification of gender based on the historical sources consulted. This is a limit of our data gathering effort that represents an opportunity for future researchers, as gender identity and its representation in political contexts has significantly changed over time.Footnote 4
Regarding place of birth, we make use of the era-appropriate location such as Upper Canada and Canada West for Ontario and Lower Canada and Canada East for Quebec. We also include an indicator of religion. However, we have religious information for just over half all legislators, and most of this this information is concentrated within Quebec where more exhaustive biographical records were available. Our final biographical indicator records the legislator's occupation for which we have information on 93.5% of legislators. These occupations fall into many professions, from lawyers to farmers, and they were reported without alteration or categorization.
In parsing legislators’ biographies, our research assistants recorded family links between the elected legislator and their previous and future kinships. These links are generally included in the legislator's biography. In this regard, each legislator has a dichotomous variable indicating whether they had a family member elected prior to the legislator's entrance or if they had a future family member elected. We further include in the dataset a series of variables identifying the name of the relative as well as the familial link. The easiest way to think about kinship is in terms of nodes or steps in a genealogical line, where an individual node represents a legislator in a family tree. The connections between nodes correspond to the relationships between family members, such as parent and child, sibling or marital status. Legislators can have multiple horizontal or vertical connections. Horizontal nodes are family links between individuals who are from the same generation (for example, siblings, cousins, spouse), while vertical nodes refer to generational movements between ancestors and descendants (for example, child, grandchild, grandparent). The maximum number of vertical nodes in our dataset is five (someone's great-great-great-grandfather), while there are horizontal family links with two nods (someone's half-sibling), and a maximum of three nodes in hybrid family links (someone's grandparent-in-law). Family links are coded both ways so that Daniel Johnson Sr. is the father of Daniel Johnson Jr. and of Pierre-Marc Johnson, while Daniel Johnson Jr. is the brother of Pierre-Marc Johnson and the son of Daniel Johnson Sr.
We likewise include electoral data aggregated and collated from the Canadian Elections Database (Sayers, Reference Sayers2017). These data include the election type (general election or by-election) for the winning candidate, their raw vote, their party, their total number of valid votes cast in the district, their vote share and their margin of victory. Electoral data for by-elections was acquired from each province's respective electoral body.Footnote 5 Margins were calculated in two ways. First, it simply takes the form of the difference in vote share in percentage points between the candidate and their main challenger. Second, when the districts are multi-member districts (MMD), as was the case in New Brunswick and Nova Scotia until the mid-1970s, then margin takes the form of the candidate's vote share less the first loser's vote share. We further include a variable indicating whether the legislator was appointed to cabinet after the election, whether the legislator was elected by acclamation and whether the riding was an MMD.
Each legislator has a variable that identifies the year, or years, in which they left office (exit year). Electoral data begins in the post-Confederation era: 1867 in Ontario, Quebec, and Nova Scotia, and 1870 in New Brunswick. Legislators have as many rows as parliamentary mandates: 4,108 pre-Confederation mandates and 12,909 post-Confederation mandates.
Applications
The dataset can be used for a variety of analyses to observe biographical, electoral and representational trends. Figure 1, for example, tracks the share of women represented in each of the provinces’ legislatures since 1940.
Female Share of the Legislature across the Four Provinces
Figure 1 confirms that the share of women in provincial parliaments has been growing into the contemporary era (post-1940) (Elections Quebec, 2024; Sevi et al., Reference Sevi, Blais, Arel-Bundock, de Greus, Tolley, Goodyear-Grant and Loewen2021). This is obviously only one part of the story since others (Janovicek and Thomas, Reference Janovicek, Thomas, Francheschet, Krook and Tan2019) have shown that the barriers to participation occur in previous stages of a political career. However, this analysis remains interesting because it demonstrates long-term regional trends, which highlights how cultural or structural factors influence the legislative representation of women across different provinces.
Figure 2 plots the share of legislators who were born outside of the province in which they were elected. In the first elections held in New Brunswick and Ontario, fully 100 per cent of all members elected were born outside their respective province. However, we note that this proportion quickly falls by the mid-nineteenth century.
Share of the Legislature, by Province, Born Outside Their Province of Representation
That legislators born outside of their province dominated early elections except in Quebec is not altogether surprising, particularly when one considers colonial legacies, notably with the settlement of Upper Canada following the American Revolution (Lamonde, Reference Lamonde2022) and the British colonization of New Brunswick and Nova Scotia. Quebec is evidently the exception given that there was already a settler population descended from French colonizers dating back to the seventeenth century. As a result, members of this community were elected to the assembly at the time of the province's first legislature.
Contribution
This dataset is designed to help researchers study representation both cross-sectionally and over time. It incorporates a large number of variables, including a unique identifier for each legislator, geographic details like province and constituency, personal information such as name, gender, birth/death years, birthplace, career length, prior occupation and religion. It also contains electoral data and legislative information, including election type, multi-member district, election year, valid votes, vote total, vote share, margin of victory, the year of departure from office, party affiliation and ministerial status. Finally, the dataset reports familial connections by indicating if a legislator had a relative elected before and/or after them, their name and the type of relationship.
This consolidated dataset of legislators spans nearly 250 years of history across the four original Canadian provinces. Given its extensive coverage and potential for comparative analysis, we believe that future researchers could use these data to answer several important research questions related to the evolution of political representation. One potential application would be to see how the composition of legislatures may have been influenced by structural changes, such as the modification of parliamentary procedures or the introduction of new electoral rules. And because provinces experienced many of these changes at different times, one could also conduct a comparative analysis to determine if any of these changes had an impact on career length, electoral turnover, the diversity of legislative assemblies or the nature of political competition. Finally, our dataset could potentially be combined with other existing archival data sources—such as the voting or speech records of parliamentarians, Canadian census data or other civil servant registries—to assess the influence of political dynasties on legislators and examine the impact of patronage and family wealth on representation.
Open access
