This paper explores how politicians respond to the public salience of policy issues when determining which topics to publicly address. Using new data and state-of-the-art methodology, our study provides a fresh perspective on this fundamental question. We focus on a multi-party parliamentary system, specifically the Canadian House of Commons, with a specific emphasis on the issue of climate change. To assess the attention given by political parties to various policy issues, we analyze transcripts from the Question Period spanning from April 2006 to June 2021. To gauge the public's level of concern for these issues, we incorporate data obtained from Google Trends. Employing an instrumental variable estimation strategy, our study causally estimates the extent to which the public salience of climate change influences elite attention. Our findings reveal that the public salience of climate change significantly influences the attention given to this issue by parties, albeit with noticeable partisan variations. Moreover, our research highlights the effectiveness of the Question Period in compelling the government to address challenging or potentially embarrassing issues. Lastly, we uncover evidence suggesting that the Liberal Party of Canada successfully increased the public salience of climate change during its tenure in government.

Leading up to the 2022 Congressional midterm elections, all predictions pointed to a Republican wave, given factors such as the incumbent president’s low approval rate and a struggling national economy. Accordingly, the underwhelming performance of the Republican Party surprised many, resulting in an election that became known as the “asterisk election” due to its unusual and seemingly unpredictable outcome. This study delves into the specifics of the 2022 midterms, exploring factors that may have influenced the results beyond those traditionally considered by political scientists. Our analysis particularly seeks to understand whether a sudden shift in the public salience of specific issues could have influenced voters’ preferences, leading them to consider factors they might not have otherwise. To achieve this, we analyzed data from a nationally representative sample of registered voters surveyed immediately after the midterm elections. Our findings reveal that the issue of abortion played a pivotal role during this election. The prominence of abortion was not predestined, as evidenced by a comparative analysis with data from a survey conducted after the 2020 presidential election. Indeed, it seems that the decision by the Supreme Court to overturn Roe v. Wade in June 2022 significantly increased the salience of abortion. This unexpected policy shock had a significant impact on the behavior of voters in the 2022 midterm elections.

Online competitive action games are a very popular form of entertainment. While most are respectfully enjoyed by millions of players, a small group of players engages in disruptive behavior, such as cheating and hate speech. Identifying and subsequently moderating these toxic players is a challenging task. Previous research has only studied specific aspects of this problem using curated data and with limited access to real-world moderation practices. In contrast, our work offers a unique and holistic view of the universal challenges of moderating disruptive behavior in online systems. We combine an analysis of a large dataset from a popular online competitive first-person action title (Call of Duty®: Modern Warfare®II) with insights from stakeholders involved in moderation. We identify six universal challenges related to handling disruptive behaviors in such games. We discuss challenges omitted by prior work, such as handling high-volume imbalanced data or ensuring the comfort of human moderators. We also offer a discussion of possible technical, design, and policy approaches to mitigating these challenges.

In parliamentary systems, it is common for legislatures to offer a regular opportunity for their members to question government ministers. While these institutions fulfill an essential function for democratic accountability, they also represent an occasion for incivility to creep into political discourse. This article investigates the incidence of uncivil behavior in these institutions and identifies some of its covariates. Our focus is on the Canadian House of Commons. Using cutting-edge, open-source machine learning models, we measure the incidence and evolution of incivility in Question Periods from April 2006 to June 2021. We find significant evidence of uncivil behavior, especially insults and toxicity. Through a multivariate regression analysis, we show that variations in the incidence of uncivil behavior over time and across members of various parties are correlated with the time remaining until the next general election, the institutional roles of parties, the balance of power, and the language of interventions.

In this article, I study the effect of endogenous challenger entry on electoral accountability in the presence of adverse selection. To this end, I analyze a two-period electoral agency model wherein a potential challenger freely chooses whether to run for office. The effect of endogenous challenger entry on policy decisions in this model is ambiguous: depending on model parameters, it can worsen or ease policy distortions. Analogously, marginally increasing the cost of running for office can deepen or reduce these distortions. This uncertainty regarding the effect of endogenous challenger entry on policymaking leads to equally ambiguous welfare implications. Nonetheless, I identify conditions under which endogenous challenger entry improves policymaking and voter welfare. This suggests that, in some circumstances, imposing higher barriers to entry in elections can improve policymaking and voter welfare.

Toxicity is a widespread phenomenon in competitive online video games. In addition to its direct undesirable effects, there is a concern that toxicity can spread to others, amplifying the harm caused by a single player's misbehavior. In this study, we estimate whether and to what extent a player's toxic speech spreads, causing their teammates to behave similarly. To this end, we analyze proprietary data from the free-to-play first-person action game Call of Duty®: Warzone™. We formulate and implement an instrumental variable identification strategy that leverages the network of interactions among players across matches. Our analysis reveals that all else equal, all of a player's teammates engaging in toxic speech increases their probability of engaging in similar behavior by 26.1 to 30.3 times the average player's likelihood of engaging in toxic speech. These findings confirm the viral nature of toxicity, especially toxic speech, in competitive online video games.

This article proposes a new approach for assessing the quality of answers in political question-and-answer sessions. We measure the quality of an answer based on how easily and accurately it can be recognized in a random set of candidate answers given the question's text. This measure reflects the answer's relevance and depth of engagement with the question. Like semantic search, we can implement this approach by training a language model on the corpus of observed questions and answers without additional human-labeled data. We showcase and validate our methodology within the context of the Question Period in the Canadian House of Commons. Our analysis reveals that while some answers have a weak semantic connection to questions, hinting at some evasion or obfuscation, they are generally at least moderately relevant, far exceeding what we would expect from random replies. We also find a meaningful correlation between answer quality and the party affiliation of the members of Parliament asking the questions.

This paper investigates whether polarized issues weigh more heavily in voters' electoral choices. To address this question, we formulate a methodology that uses conjoint experiments to measure how responsive respondents' electoral choices are to agreeing or disagreeing with candidates' policy positions on policy issues, which we argue reflects their importance. Our approach is grounded in the potential outcomes framework and designed to minimize the burden on respondents. Immediately after the 2022 Congressional midterm elections, we implemented this approach on a nationally representative sample of 2,109 U.S. registered voters. Using the resulting estimates, we measure the correlation between issue importance and their polarization. We consider two types of political polarization: policy and partisan polarization. Our findings reveal that partisan polarization exhibits a strong and significant correlation with issue importance, whereas policy polarization does not. This offers new insights into the electoral behavior of American voters.

This article seeks to provide accurate estimates of the causal effect of exposure to toxic language on player engagement and the proliferation of toxic language. To this end, we analyze proprietary data from the first-person action video game Call of Duty®: Modern Warfare® III, published by Activision®. To overcome causal identification problems, we implement an instrumental variables estimation strategy. Our findings confirm that exposure to toxic language significantly affects player engagement and the probability that players use similar language. Accordingly, video game publishers have a vested interest in addressing toxic language. Further, we demonstrate that this effect varies significantly depending on whether toxic language originates from opponents or teammates, whether it originates from teammates in the same party or a different party, and the match's outcome. This has meaningful implications regarding how resources for addressing toxicity should be allocated.

Online platforms take proactive measures to detect and address undesirable behavior, aiming to focus these resource-intensive efforts where such behavior is most prevalent. This article considers the problem of efficient sampling for toxicity detection in competitive online video games. To make optimal monitoring decisions, video game service operators need estimates of the likelihood of toxic behavior. If no model is available for these predictions, one must be estimated in real time. To close this gap, we propose a contextual bandit algorithm that makes monitoring decisions based on a small set of variables that, according to domain expertise, are associated with toxic behavior. This algorithm balances exploration and exploitation to optimize long-term outcomes and is deliberately designed for easy deployment in production. Using data from the popular first-person action game Call of Duty®: Modern Warfare® III, we show that our algorithm consistently outperforms baseline algorithms that rely solely on players' past behavior. This finding has substantive implications for the nature of toxicity. It also illustrates how domain expertise can be harnessed to help video game service operators identify and mitigate toxicity, ultimately fostering a safer and more enjoyable gaming experience.

This article considers how ``"self-imposed term limits," voluntary pledges by candidates for elected office to step down at the end of their term if elected or reelected, can arise endogenously within a formal model of electoral accountability. To this end, I extend a standard two-period political agency model with adverse selection by appending an election campaign during which two candidates compete with the option to commit not to seek reelection. I show that term limit pledges can emerge in equilibrium without commitment power, revealing their relationship with informative cheap talk. The subsequent analysis considers scenarios where commitment power is necessary for upholding term limit pledges. In these cases, self-imposed term limits allow strong candidates to signal their private type to voters, giving them an electoral advantage over those seeking reelection. However, these candidates' appeal comes not from the informational content of their pledges but from their ability to make policy decisions free from career considerations. I characterize the equilibria of a model specification in which politicians have differing policy preferences, identifying conditions under which term limit pledges are simultaneously informative and improve voter welfare.