Election Timing, Roles, and Language Shape Incivility in Canada's Question Period

Why Study Incivility in Question Period?

Jacob Morrier and R. Michael Alvarez investigate how uncivil behavior—insults, toxic language, and rude exchanges—emerges in one of parliament's liveliest forums: Question Period in the Canadian House of Commons. Question Period is a routine, high-visibility process where legislators interrogate ministers; it serves democratic accountability but also creates a structured setting where incivility can surface. Understanding when and why members become uncivil matters for theories of legislative behavior, party competition, and democratic norms.

Measuring Incivility: Machine Learning on Parliamentary Transcripts (April 2006–June 2021)

The authors apply open-source, state-of-the-art machine learning models to transcribed Question Period speeches from April 2006 through June 2021 to measure the incidence and evolution of uncivil language. Their approach operationalizes incivility with classifiers that identify categories such as insults and toxicity, producing a continuous record of uncivil interventions across time, parties, and individual members.

What the Analysis Tested

Using these automated measurements, Morrier and Alvarez estimate multivariate regression models to explain variation in incivility. Key covariates include the time remaining until the next general election, parties' institutional roles (for example, government versus opposition), the balance of power in the House, and the language of interventions in Canada's bilingual legislature.

Key Findings

• The analysis uncovers significant evidence of uncivil behavior during Question Periods, with insults and toxic language prominent among the detected categories.
• Incivility varies over time and across MPs and parties, and these differences are systematically correlated with institutional and contextual factors.
• Higher incidence of uncivil language is associated with proximity to a general election, party roles in the legislature, shifts in the balance of power, and the language used by speakers (English vs. French).

What This Means For Research and Practice

By combining automated text classification with regression analysis, the study provides a replicable way to track legislative incivility and links observed behavior to plausible institutional drivers. These results inform debates about parliamentary norms, the strategic use of aggressive rhetoric by parties, and the institutional conditions that either exacerbate or restrain uncivil discourse in legislatures.