![](_page_0_Picture_0.jpeg) # **British Columbia Election Information Ecosystem Project** August 2025 ![](_page_1_Picture_3.jpeg) # **Executive Summary** The British Columbia Election Information Ecosystem Project set out to evaluate the dynamics of mis- and disinformation, media platform influence, foreign interference concerns, and public trust during the 2024 British Columbia provincial election. Through a combination of social media monitoring, digital trace analysis, and pre- and post-election surveys, the project offers key insights into the evolving information landscape and its impact on electoral integrity. We find: - 1. **Elections BC remains a highly trusted entity among British Columbians**, with no measurable population-level erosion of confidence detected. Pre- and post-election surveys reveal strong public trust in the fairness and integrity of Elections BC (slightly higher than trust in Elections Canada in fact), underscoring its role as a reliable source of election information. - 2. **The online political discourse surrounding the election was notably polarized**, with distinct partisan divides apparent across social media platforms. Platform segmentation was pronounced, with X (formerly Twitter) emerging as the central hub for political discussion and mass engagement, significantly overshadowing other platforms like Instagram and Facebook. The prevalence of partisan and polarized conversation on X highlights the need for ongoing assessment of each platform's unique risk profile, particularly considering factors such as high bot activity, skewed algorithms, and limited transparency. - 3. Although the official campaign period experienced minimal misinformation activity, a marked increase in online discourse questioning election integrity emerged the day of the election and in the immediate post-election phase. Notably, despite the high visibility of such discourse online, there was minimal impact of mis- and disinformation on public attitudes. These findings dem- # **Sommaire exécutif** Le projet sur l'écosystème de l'information électorale en Colombie-Britannique visait à évaluer les dynamiques de désinformation et de mésinformation, l'influence des plateformes médiatiques, les préoccupations liées à l'ingérence étrangère et la confiance de la population lors des élections provinciales de 2024 en Colombie-Britannique. Le projet offre des informations clés sur l'évolution du paysage informationnel et sur l'influence de ce dernier sur l'intégrité électorale. Pour obtenir ces données, plusieurs méthodologies ont été mises en œuvre : veille de médias sociaux, analyse des traces numériques et conduite de sondages avant et après les élections. Nous avons constaté : - 1. **Elections BC demeure une entité très fiable aux yeux des citoyens et citoyennes de la Colombie-Britannique.** - Aucune baisse mesurable de confiance n'a été observée auprès de la population. Les sondages menés avant et après les élections révèlent une forte confiance du public dans l'équité et l'intégrité de l'organisme, laquelle dépasse même légèrement la confiance à l'égard d'Élections Canada. Ces résultats consolident le statut d'Elections BC comme source d'information électorale fiable. - 2. **Le discours politique en ligne entourant l'élection était particulièrement polarisé** et des divisions partisanes nettes apparaissent sur toutes les plateformes de médias sociaux. La segmentation des plateformes était prononcée, X (anciennement Twitter) s'étant imposé comme centre névralgique des discussions politiques et du dialogue de masse par rapport à d'autres plateformes comme Instagram et Facebook. La prévalence d'échanges partisans et polarisés sur X fait ressortir la nécessité d'évaluer régulièrement le profil de risque de chaque plateforme en demeurant à l'affût, tout particulièrement, d'éléments tels qu'une forte activité des robots logiciels (*bots*), les algorithmes biaisés et la transparence limitée. onstrate that **extensive online activity does not necessarily reflect broader offline voter sentiment or engagement**, underscoring the importance of contextualizing digital narratives within real-world public attitudes. 4. Our investigation found **no evidence of foreign information manipulation or interference during the election**. However, misinformation and speculation about foreign interference appear to have become normalized in election-related discourse. The project also identified vulnerabilities within BC's election information ecosystem that merit attention. - 5. **X's disproportionate political influence:** X played a dominant role in both legitimate political discourse and misinformation spread, highlighting the need for targeted monitoring and analysis of and transparency from the platform. - 6. **Persistent misinformation narratives, even in the absence of evidence:** False and misleading claims have become normalized in online spaces, alongside a broader culture of skepticism and conspiracy. Algorithms exacerbate this dynamic by amplifying emotionally resonant content, regardless of its accuracy, thereby increasing the visibility and reach of false information. - 7. **Declining ability to identify and trust reliable information:** With nearly half of respondents unable to identify a trusted source, British Columbians (and Canadians overall) are vulnerable to confusion, distrust, and manipulation. The 2024 BC election demonstrated that while democratic institutions can be effective and continue to earn citizen's trust, they operate in an increasingly volatile information landscape. Safeguarding electoral integrity now demands anticipatory communication, ongoing monitoring and reporting, and a renewed focus on public engagement. - 3. Si la mésinformation est restée marginale durant la campagne officielle, on a observé, le jour de l'élection et durant la phase postélectorale immédiate, un pic des propos tenus en ligne remettant en question l'intégrité électorale. Cela dit, malgré la forte visibilité de ces propos, la mésinformation et la désinformation n'ont eu qu'une incidence limitée sur les attitudes de la population. Ces résultats démontrent qu'une **activité numérique intense n'a pas nécessairement d'effet sur l'opinion ou la participation des électeurs en dehors des plateformes de médias sociaux**, d'où l'importance de replacer les récits numériques dans le contexte des attitudes réelles de la population. - 4. Nous n'avons trouvé **aucune preuve de manipulation d'information ou d'ingérence étrangères pendant l'élection**. Cependant, la mésinformation et les présomptions sur l'ingérence étrangère semblent faire aujourd'hui partie intégrante du discours électoral. Par ailleurs, le projet a permis de mettre en évidence des vulnérabilités au sein de l'écosystème de l'information électorale de la Colombie-Britannique qui méritent qu'on s'y attarde. - 5. **L'influence politique disproportionnée de X :** X a joué un rôle prépondérant tant dans le discours politique légitime que dans la propagation de la mésinformation, d'où le besoin d'une surveillance et d'une analyse ciblées ainsi que d'une plus grande transparence de la part de la plateforme. - 6. **Persistance de la mésinformation, même en l'absence de preuves :** Les agoras en ligne ont normalisé le scepticisme et les théories du complot, tandis que les algorithmes et les contenus à forte résonance émotionnelle amplifient ces risques. - 7. **Affaiblissement de la capacité à reconnaître l'information fiable et à s'y fier :** Près de la moitié des répondants se révèlent incapables de reconnaître une source digne de foi. Les Britanno-Colombiens (et les Canadiens dans l'ensemble) demeurent vulnérables à la confusion, à la méfiance et à la manipulation. Les élections de 2024 en Colombie-Britannique ont démontré que, même si les institutions démocratiques pouvaient être efficaces et continuer à gagner la confiance des citoyens et citoyennes, elles évoluaient dans un paysage informationnel de plus en plus instable. La protection de l'intégrité électorale exige désormais une communication proactive et une veille constante de l'écosystème de l'information ainsi que de la diffusion des informations d'intérêt; en outre, on doit s'employer plus que jamais à amener la population à s'intéresser à ces enjeux. ### **Contents** | Executive Summary ii | | Section 4 | |----------------------------------|------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------| | Section 1 | | Election integrity
14 | | Introduction 1 | | 1
Foreign interference and influence
14 | | 1. | Objectives and scope of the report
2
2. Overview of the information
ecosystem approach
2 | 1-1
Self-reported concern among British
Columbians
14 | | | | 1-2
Social media interference
15 | | | | 2 Bot activity and astroturfing
16 | | Section 2 | | 3
Claims of election fraud and
wrong-doing
17 | | Election Information Ecosystem 3 | | | | 1. | Where British Columbians followed
the news during the election
4 | 3-1
Source of information questioning
election integrity
18 | | | 2 Role and influence of social media
platforms
4 | 3-2 Impact of these claims on attitudes
19 | | | 2-1
Social media platforms used by British
Columbians to get political information | Section 5
Lessons learned and recommendations
22 | | | during the campaign
4 | Recommendations
22
Section 6
Methodology .
24 | | | 2-2 Platform-specific use during the campaign 4 | | | | 2-3 Engagement online with BC election
related content
5 | | | 3 | Structure of the British Columbia | 1
Survey methodology
24 | | | information ecosystem
6 | 1-1
Pre-election survey
24 | | | 3-1
Ideological and conceptual division | 1-2
During-election survey
24 | | | amongst influential voices
6 | 1-3
Post-election survey
24 | | | 3-2 Polarization across platforms
7 | 2
Avatar methodology
24 | | | 3-3 Polarization among voters:
ideological divides
8 | 3
Digital trace and social
media methodology
25 | | 4 | Information about voting
8 | Contributors 27 | | | Section 3
Issues and priorities during the campaign .
9 | Notes .
27 | | 1 | According to the voters
9 | | | | 2 According to influential voices
10 | | | 3 | Key issues
11 | | | | 3-1
Environment and climate change
11 | | | | 3-2 Health and drug policy
11 | | | | 3-3 Crime and Safety
12 | | ![](_page_4_Picture_3.jpeg) #### **Section 1** # **Introduction** The 2024 British Columbia general election, held on October 19, occurred at a pivotal moment both politically and technologically. Marked by a rapidly evolving digital information ecosystem, the election unfolded amid growing concerns about misinformation, electoral interference, and shifting voter behavior. Voter turnout was 58.45%, with just over 2.1 million ballots cast.1 Election day saw severe weather disruptions and there were several ridings with extremely close margins that led to two mandatory judicial recounts and one partial judicial recount. A defining feature of the 2024 election was the reconfiguration of BC's political right. The BC United Party, formerly the BC Liberals, suspended its campaign just weeks before the writ was issued, urging supporters to consolidate around the BC Conservatives. This dramatically reshaped the electoral map. The BC Conservative Party, led by John Rustad, surged from obscurity to win 44 seats and 43.3% of the popular vote, becoming the official opposition. Meanwhile, the incumbent BC NDP, under Premier David Eby, retained power but lost eight seats with 44.9% of the vote, finishing with 47. The BC Greens held steady with two seats, though their vote share declined. Major themes of the campaign included affordability, healthcare, housing, and public safety. Although misinformation and Indigenous rights were not dominant issues, both figured in the wider discussion of the campaign. While the information environment remained polarized, with spikes in online distrust following preliminary results, post-election surveys indicated that public confidence in electoral accuracy remained high, with 80% of voters expressing trust in the results. The 2024 BC election also coincided with the United States presidential election, a deeply polarized and contentious moment in American democracy. With significant implications for Canada ranging from economic policy and border management to climate action and international alliances, this parallel electoral cycle was closely followed by Canadians. The intense media coverage and societal divisions evident in the U.S. campaign environment amplified broader concerns about political polarization and democratic resilience, providing a backdrop to the provincial vote. On the federal stage, drama swirled around Prime Minister Justin Trudeau, whose popularity sank to historic lows. Polls throughout late 2024 showed steep declines in Liberal support (some sinking to single digits), with the Conservative Party under Pierre Poilievre surging to leads of 20+ points, favored by Canadians on issues like the cost of living, crime, and economic management. Elections BC delivered an election widely regarded as secure and transparent. The outcome reaffirmed the NDP's hold on government but also signaled a rising right-wing opposition poised to redefine the province's political future. # **1. Objectives and scope of the report** This report aims to provide an assessment of the information ecosystem surrounding the 2024 British Columbia provincial election and present the findings and recommendations for electoral management bodies, policymakers, media organizations, and civil society, to enhance resilience and proactively address risks that might impact election integrity in the future. To achieve this, we: - 1. Analyze the role and impact of social media platforms in disseminating political information, including identifying trends, risks, and the degree of polarization among political actors and the electorate. - 2. Investigate the presence and impact of misinformation, disinformation, and foreign and/or domestic interference (or influence) within the election-related discourse. - 3. Evaluate British Columbians' trust and confidence in the electoral process and in Elections BC as a trusted institution. The scope of the report includes analysis conducted before, during, and after the election, utilizing survey data, social media monitoring, and digital trace analysis. It considers key political actors and influencers across major social media platforms (X (formerly Twitter), Facebook, Instagram, YouTube, and TikTok) to capture a rich picture of the information ecosystem during the election. This report provides analysis across three key areas: - E **Section 2: The Election Information Ecosystem** explores how various information sources, mainly mainstream media and social platforms, shaped voter perceptions and political discourse during the 2024 BC provincial election. - E **Section 3: Issues and Priorities** focuses on the dominant themes and voter concerns that emerged in the public conversation, highlighting which topics mobilized attention and engagement. - E **Section 4: Election Integrity** assesses public trust in the electoral system, examines the prevalence of misinformation and disinformation, and investigates potential instances of foreign or domestic interference. The analysis demonstrates the need for vigilant yet balanced election-monitoring practices, acknowledging online-offline sentiment gaps, preparing strategic communication responses, and recognizing social media's evolving influence on electoral integrity. ### **2. Overview of the information ecosystem approach** The concept of an information ecosystem describes the complex interplay between media, technology platforms, political actors, and the public, emphasizing how information is produced, disseminated, consumed, and influenced. In the context of the 2024 British Columbia election, we employed a multi-method approach (see Methodology) to map and understand this ecosystem: - E **Survey Research:** Conducting comprehensive, repeated surveys of British Columbians across the pre-election, campaign-period, and post-election phases to capture shifts or stability in public trust, issue priorities, perceptions of misinformation, and electoral integrity. - E **Digital trace analysis:** Collecting and analyzing social media data from influential British Columbian entities (politicians, media outlets, influencers, civil society groups) to track online discourse, engagement patterns, and information dissemination across multiple platforms. - E **Avatar-based social media monitoring:** Using controlled, avatar-based observations across social media platforms to capture real-time dynamics, identify emergent misinformation narratives, and detect coordinated or inorganic activity. Through this approach, the report aims to provide nuanced insights into the strengths, vulnerabilities, and overall health of the electoral information ecosystem in British Columbia, informing strategies to safeguard future electoral processes. ![](_page_6_Picture_2.jpeg) ![](_page_6_Picture_3.jpeg) #### **Section 2** # **Election Information Ecosystem** We begin the description of the information ecosystem during the election by providing an overview of the behaviours and attitudes of British Columbian citizens, politicians, and influencers before, during, and after the campaign period. ## **1. Where British Columbians followed the news during the election** There has been considerable discussion of the growing role social media plays in providing Canadians with general and political information.2 We evaluate the extent to which British Columbians get their news from social media versus other mediums in Figure 1. The Figure displays the extent to which British Columbians reported following various news sources either *"fairly"* or *"very"* closely during the campaign period. Television (45.7%) and news websites (43.0%) were the most closely followed sources, with nearly half of respondents each. This suggests that traditional broadcast and digital formats remain dominant in the province's media consumption habits. Social media was the third most followed source at 30.7% where its reach still lags behind more established platforms. Radio (23.3%) and print media (22.2%) were followed closely by roughly one in five respondents, indicating a more modest but still meaningful audience for these legacy formats. Podcasts, by contrast, were the least closely followed source, with only 13.0% of respondents. While still trailing behind television and news **Figure 1: Primary sources of news during the 2024 British Columbia Election** ![](_page_6_Figure_11.jpeg) websites in terms of close followership, podcasts have shown steady growth in audience engagement and are increasingly recognized as influential platforms for in-depth discussion and niche topics. As their popularity continues to rise in other regions and among younger demographics, podcasts may become a more significant pillar in British Columbia's evolving information ecosystem. Overall, British Columbians operate in a hybrid information environment in which traditional broadcast sources interplay with newer digital platforms. ## **2. Role and influence of social media platforms** As the information environment in British Columbia continues to evolve, understanding where voters turn for news online, specifically which social media platforms, and how digital platforms shape political discourse, is essential. The following subsections describe how British Columbians access political information and the news on social media, and how the influence and distinct characteristics of these major social media platforms shape political engagement and conversation. ### **2-1. Social media platforms used by British Columbians to get political information during the campaign** Figure 2 illustrates both general and political weekly use of five major social media platforms during the campaign period. Facebook remains the most widely used platform overall, with 61.0% of respondents reporting general weekly use. Importantly, nearly one in four (24.5%) of all respondents use Facebook specifically for political purposes, highlighting its continued relevance as both a social and political information conduit, despite Canadian and international news being blocked on the platform.3 YouTube follows closely, with 57.9% general usage and similar political use (24.7%). It remains an important source of political information, particularly given its dual function as both **Figure 2: General and political use of social media platforms in BC** ![](_page_7_Figure_9.jpeg) an entertainment and information hub. Instagram is another platform with a large user base (48.4%) and significant political use (19.5%). For these three platforms, approximately 40% of each user base uses the platforms for political purposes. The bottom two platforms: X and TikTok, have smaller user bases, but unlike the big three, more users use the platform for political purposes, i.e. while only 17.5% of respondents reported using X for political purposes, that is more than two thirds of all X users. Similarly, 66% of all TikTok users use the platform for political purposes. This rate is substantially more than Youtube and Meta platforms, which only see about 40–45% of their self-identified users also report political use. Notably, TikTok usage is much higher among Canadian youth under the age of 25, however, with 46% using the platform and 30% of them using it for political purposes (these numbers drop to just 6.5 and 2%, respectively, for Canadians aged 55+). While usage statistics reveal overall reach, the character of political discourse varied dramatically from one platform to the next. ### **2-2. Platform-specific use during the campaign** While usage statistics reveal overall reach, the character of political discourse varied dramatically from one platform to the next. **Facebook** was heavily used by political candidates, but BC-based influencers tended not to use the platform and news organizations were not present at all due to being blocked from sharing content. While many British Columbians still use Facebook to get news, we observed a low volume of news content and of general content about the BC-election on the platform, with the notable exception of campaign organizing on the platform by candidates. **YouTube** content regarding the election was dominated by influencer analyses of political statements and formal news coverage, including interviews and election updates. It served more as a space for in-depth analysis than real-time discussion. Engagement was often driven by longer-form content rather than immediate public reaction. **Instagram** saw active participation from political candidates, particularly the BC-NDP, as well as influencers. Like Facebook, news organizations were blocked on the platform, and we observed emerging anonymous accounts presenting themselves as news sources. These accounts have proliferated in Canada under the Meta news ban. Unlike X, interactions here were more curated, with limited use of evidence in comments. Everyday users were less visible, making it harder to gauge authentic public sentiment. **X** was central to political discourse, with candidates and party accounts driving much of the conversation, especially from the BC-Conservatives, often framing their opponents as "extreme." Influencers, journalists, and everyday users amplified and debated these points, resulting in long chains of replies. The platform hosted and amplified more polarizing discussions compared to others. **TikTok** featured a mix of content from news publishers, influencers, and other users, with the Green Party notably active in engaging the public through comments. Its algorithm quickly adapts to user preferences but generally offered a more spontaneous and grassroots view of public opinion. ### **2-3. Engagement online with BC electionrelated content** Figure 3 shows patterns of online engagement with posts related to British Columbia across five major social media platforms over the campaign period. The data captures interactions such as likes, shares, comments, and views on content generated by influential accounts including politicians, media organizations, influencers, and civil society groups. The campaign period startand end-dates are noted on the plot. Engagement trends varied substantially across platforms. The highest engagement platform by far was X during the election (56%). This platform consistently showed the highest total volume of interaction, and activity was considerably higher during and in the immediate aftermath of the election. TikTok, while somewhat more volatile, showed multiple spikes in engagement, particularly in late August and again post-election, suggesting episodic virality rather than sustained political discourse. This behavior reflects the platform's algorithmic dynamics and younger user base, where a single post or creator can drive disproportionate reach. Instagram and Facebook displayed more stable patterns, with noticeable peaks around key campaign events in early and mid-October, like the debates. These platforms appear to serve as consistent channels for political engagement, with a mix of official communications and community-based commentary. YouTube showed upticks in engagement in the middle phase of the campaign, indicating increased attention during the heart of election season. Overall engagement was extremely low, however. Overall, the data highlights how different platforms served distinct roles in shaping the campaign conversation in BC, with X dominating volume, Facebook and Instagram anchoring steady engagement, and TikTok and YouTube generating more episodic attention. **Figure 3: Engagement with BC-related posts by platform** ![](_page_8_Figure_12.jpeg) Media Ecosystem Observatory [meo.ca](https://meo.ca) **Figure 4: Information clusters in the BC political ecosystem based on semantic similarity** ## **3. Structure of the British Columbia information ecosystem** The political information environment in British Columbia during the campaign period was stratified and polarized. Online discourse was shaped not only by party messaging but also by the platforms on which it circulated and the ideological leanings of both influencers and voters. Political communication was largely confined within distinct ideological clusters, with limited cross-group interaction. While mainstream media offered some connective capacity, they did not fully bridge partisan divides. This fragmentation, combined with uneven platform dynamics and rising affective polarization, suggests a digital landscape that may hinder deliberative dialogue and amplify narrative silos. #### **3-1. Ideological and conceptual division amongst influential voices** Figure 4 visualizes the structure of British Columbia's online political information environment through a network projection. Nodes (circles) represent media organizations, political figures (NDP, Liberal and Conservative), and influencers, coloured by party affiliation and scaled by total engagement (size of the circles). Proximity between nodes reflects the degree of similarity in the content they post (based on hashtags, URLs, shared topics, and cross-platform post similarity). Three distinct ideological and communicative clusters are evident. In the middle bottom, a dense grouping of Conservative-aligned accounts, including Pierre Poilievre, Ezra Levant, and independent influencers like Kat Kanada, form a tight, relatively self-contained information sphere. These actors engage with each other frequently and share semantically similar messaging, suggesting a high degree of narrative coordination or alignment. In the top-right, NDP figures such as Jagmeet Singh, David Eby, and Peter Julian occupy a similarly clustered zone. This cluster features fewer non-partisan influencers but maintains cohesion in messaging, distinct from both media institutions and the right-aligned cluster. Roughly in the center, mainstream media organizations (e.g., CBC, CTV, Global News) and prominent journalists (e.g., Richard Zussman, Andrew Coyne) anchor a relatively neutral zone. While these accounts do not form a tight semantic bloc, they serve as connective tissue between ideological groups, indicating their role in circulating broadly accessible or less partisan content. Media Ecosystem Observatory [meo.ca](https://meo.ca) Liberal figures such as Justin Trudeau appear between the NDP and media clusters, suggesting a more moderate or mixed engagement pattern. Overall, the spatial and semantic separation of clusters reflects a polarized information environment where political discourse is largely insular, with limited crossover between ideological groups. While media actors offer a degree of bridging, partisan clusters (particularly on the right) show high internal cohesion and low integration with the broader ecosystem. This fragmentation has important implications for the reach of mis- and disinformation, the resilience of democratic debate, and the potential for narrative echo chambers. Information is likely to travel very quickly within likeminded communities, but sharing may not stretch across the entire ecosystem. #### **3-2. Polarization across platforms** While polarization appears across the entire information ecosystem, we also observed some important platform-level stratification. Figure 5 presents weekly engagement with political content posted by political candidates, official party accounts, and non-candidate members of the legislative assembly from the Conservative Party, Green Party, and NDP over the course of the campaign period, broken out by total engagement (top plot) and by platform (bottom plots). The patterns reveal both the timeline of online attention and clear partisan asymmetries in reach and visibility. Across all platforms combined (top plot), engagement peaked in early to mid-October, corresponding with the most intense phase of the campaign. Conservative politicians consistently dominated in volume of engagement, particularly between late September and mid-October, suggesting a strong mobilization strategy or greater resonance among their online follower base. NDP politicians generated steadier, moderate engagement, with notable peaks both early in the campaign and post-election. Green politicians had comparatively minimal engagement throughout, aside from several brief spikes. Platform-specific trends further contextualize these dynamics. On Instagram (bottom left), NDP politicians received substantial attention, with almost no Conservative content or engagement. This suggests that Instagram may serve as a relatively stronger channel for progressive communications. Green Party spikes are also most visible here, likely reflecting brief viral surges from visual or activist-oriented content. On X (bottom right), Conservatives dominated both the consistency and scale of engagement, with multiple peaks coinciding with national debates and campaign controversies. The platform appears to have served as a key vector for the Conservative digital campaign, particularly in amplifying candidate messaging and reactive commentary. These patterns illustrate the stratified nature of political engagement across platforms. While X favoured Conservative reach, Instagram did the same for progressive voices. The divergence **Figure 5: Weekly social media engagement with party-affiliated candidates across platforms** ![](_page_10_Figure_11.jpeg) ![](_page_10_Figure_12.jpeg) underscores the role of platform choice and audience alignment in shaping the digital effectiveness of party campaigns. ### **3-3. Polarization among voters: ideological divides** We have shown ecosystem and platform polarization, but polarization is also an attitude. Figure 6 highlights stark partisan affect and deepening polarization in the Canadian political information ecosystem by illustrating how Conservative and NDP voters evaluate each other's parties (specifically, how much they like or dislike the opposing party). The data shows a pronounced affective divide, with intense dislike of the opposing party common among both voter groups of the major parties. Both NDP and Conservative voters show strikingly negative views of the opposing party, with a large share in each group assigning the lowest possible rating in our scale (0-10). Among NDP voters, 54% rated the Conservative Party in this lowest bin, an especially sharp rejection, with very few offering higher scores. Among Conservative voters, only 41% rated the NDP between 0 and 10, a large number but less than their NDP counterparts. In contrast, in-group ratings are more muted: 25% of Conservatives gave their own party a 91-100, and 21% of NDP voters rated their own party 91 or higher. This bimodal distribution of party ratings reveals a deep affective divide and limited political cross-sympathy. The mutual dislike signals a broader erosion of shared legitimacy across partisan lines. #### **4. Information about voting** A critical component of a free and fair election is clear information for electors, often provided by an impartial election **Figure 6: Affective polarization between Conservative and NDP voters** ![](_page_11_Figure_11.jpeg) **Figure 7: Public Trust in Election Information Sources and Institutions in British Columbia** ![](_page_11_Figure_13.jpeg) administration body. Figure 7 highlights the central role of Elections BC in the provincial information ecosystem, particularly when it comes to public trust in election-related information. Despite widespread concern about misinformation and a global population-level decline in institutional confidence, Elections BC stands out as a comparatively trusted entity. On the left, just over one-quarter of respondents (27%) explicitly named Elections BC as their most trusted source for election information. By contrast, only 13% identified news media, 6% pointed to political sources, and 5% mentioned other sources. Notably, nearly half (49%) were unable or unwilling to name a trusted source at all. Trust in institutions (right plot) further reinforces this pattern. Most respondents expressed either "a lot" or "a fair amount" of trust in Elections BC (32% and 38%, respectively), placing it well ahead of all other entities assessed. By comparison, journalists and the broader news media received moderate levels of trust, but also significant skepticism, with "a little trust" or "no trust at all," accounting for over 55% of respondents in both cases. Trust in elected officials was lower still, with nearly 70% of respondents expressing only minimal or moderate trust. Tech companies ranked lowest overall, with more than a third of British Columbians indicating they had no trust at all in these platforms to act in the public interest. As media and political figures struggle to maintain credibility, nonpartisan electoral institutions like Elections BC hold a unique and important position. ![](_page_12_Picture_3.jpeg) #### **Section 3** # **Issues and priorities during the campaign** We now turn from the structure of the information environment to the substantive issues that shaped political discourse and voter engagement during the campaign period. This section examines which topics voters prioritized most, how those priorities were reflected in online engagement across party lines, and the ways in which specific policies, especially around healthcare, affordability, and climate change, became important subjects for the campaign and sites of misinformation. # **1. According to the voters** We begin this section by evaluating the priorities of British Columbian voters. Figure 8 captures the relative salience of major campaign issues we identified as ranked by voters during the 2021 federal election period, with the green dot showing the average ranking (number 1 is most important and number 8 is least important) and the grey background showing the distribution for all responses of British Columbians when asked which issue is a top priority. Among the issues presented, healthcare emerged as the most consistently prioritized concern, with a median ranking of 3 and a tight clustering, indicating broad agreement across the electorate about its importance. The economy and cost of living followed closely behind, reflecting widespread voter concern over financial security and economic recovery amid ongoing post-pandemic instability. These two issues, while closely ranked on average, show slightly more variance, suggesting pockets of diverging emphasis. Education and climate change occupied the middle tier of voter priorities. Climate had a broader range of rankings, reflecting its high salience among some voters (particularly younger and left leaning) but lower urgency among others. Housing was also mid-ranked overall, though its wider distribution suggests variation in perceived importance. Notably, misinformation and Indigenous issues were ranked lowest in terms of perceived importance during the campaign. While both topics were present in public discourse, they did not rank as top-tier issues for most voters. That said, the relatively low mean ranking does not necessarily reflect a lack of concern but may instead point to their perceived distance from immediate electoral consequences or media coverage. These findings underscore the continued dominance of healthcare, affordability, and economic management even amid heightened attention to democratic resilience and reconciliation. **Figure 8: Relative importance of key issues among potential voters during the campaign** ![](_page_13_Figure_5.jpeg) ## **2. According to politically influential voices** Political parties, while generally responsive to the overall concerns of voters, may also choose to prioritize certain issues. Figure 8 illustrates the proportion of social media content posted by candidates and influencers affiliated with the Conservative Party, Green Party, and New Democratic Party (NDP), alongside other influential voices. Healthcare and drug policy consistently appears as a top issue across all groups, with NDP-affiliated accounts posting about this topic most frequently, followed closely by Conservatives, Greens, and other influential voices. This suggests broad-based concern and a shared emphasis on healthcare across ideological lines. Housing, homelessness, and affordability ranks highly for both NDP and Conservative accounts and is also a leading issue for other influential voices and Green accounts. This indicates a widespread focus on affordability challenges, especially salient in urban regions. Public Safety and Crime feature more prominently in Conservative-affiliated posts, with Conservatives leading all groups in content volume on crime. Other influential voices appear to have focused quite strongly on this topic as well. In contrast, these topics are less emphasized by the NDP and Greens. Education and Childcare was discussed most prominently by NDP and Green accounts, while Conservative and other influencers focused on the topic less. Climate change and the environment dominates the Green Party's social media activity, representing a much larger share of their total posts than for any other group. In contrast, Conservatives, NDP, and other voices post about environmental issues far less often, underscoring the Greens' distinct thematic concentration. Immigration, race, diversity, and cultural conflict appears most frequently amongst other influential voices, and NDP accounts ranks the highest amongst partisan groups. Overall, the posting share remains modest. Other groups, Conservatives and Greens, show relatively lower activity on this issue. Reconciliation and Indigenous issues, along with gender affirmation and LGBTQ+ rights, are posted about more often by Green and NDP accounts. These issues are minimally represented in Conservative content but are moderately present among other influential voices. Overall, the distribution of posting activity reveals clear thematic distinctions across political lines. While certain topics like healthcare span partisan divides, others (particularly environmental and identity-based issues) remain highly differentiated in terms of posting focus. **Figure 9: Engagement with candidate and influencer posts on key issues, by party** # **3. Key issues** While there were a wide range of issues, we note that three issues got a lot of attention due to misinformation and/or perceptions of information manipulation: healthcare and drug policy, public safety and crime, and climate change and the environment. This is not an exhaustive list but is illustrative of the online information environment during the election. #### **3-1. Healthcare and drug policy** Throughout the campaign, the BC Conservatives faced sustained backlash over claims that they planned to cut \$4.1 billion from healthcare spending (a figure that originated from a misrepresentation of their platform). The claim was first amplified by NDP candidate Ravi Parmar, who cited a reference in the Conservatives' platform to a Deloitte report suggesting potential efficiencies in healthcare spending. Despite Conservative Leader John Rustad's public denials and clarification that no such cut was proposed, the claim spread rapidly across social media, becoming a viral talking point repeated by NDP candidates and even echoed in posts by Premier David Eby. Critics framed the Conservatives' healthcare policy as a dangerous move toward privatization, raising fears that patient care would suffer under their governance. Adding to the controversy, John Rustad's past statements on vaccination, particularly his remark regretting his decision to receive the COVID-19 vaccine, were used by the NDP to question his credibility on health policy. Eby attempted to position Rustad as inconsistent and politically opportunistic, accusing him of tailoring messages depending on the audience, specifically with regards to vaccines. Simultaneously, Rustad's strong opposition to drug decriminalization and his personal anecdotes about the opioid crisis, some of which were later contradicted or clarified, fueled further scrutiny.4 While not always outright misinformation, some of Rustad's remarks were perceived as misleading or exaggerated. In turn, the Conservatives criticized the NDP for failing to effectively handle the opioid crisis. #### **3-2. Public Safety and Crime** Crime and public safety emerged as one of the most polarizing issues of the 2024 BC election, with debates often overshadowed by competing narratives around the causes of crime and the role of government in preventing it. The BC Conservative Party campaigned on a public-safety message focused on NDP policies on issues like drug decriminalization, gun control, and bail reform. They framed the NDP as responsible for causing a wave of repeat offenders and open drug use into public spaces. This **Samples 1: Several posts and comments regarding factual accuracy of healthcare claims during the election** ![](_page_15_Figure_3.jpeg) ![](_page_15_Figure_4.jpeg) ![](_page_15_Figure_5.jpeg) gained traction online, where screenshots and videos, such as a clip shared by the Conservative Party of BC featuring a Joe Rogan guest condemning David Eby's decriminalization policy, circulated widely, reinforcing the party's claims of a failed public safety strategy. The polarized tone was heightened on platforms like X, where pro-Conservative voices accused the NDP not only of policy failure but of deliberately undermining the safety of law-abiding citizens. NDP MLA Baltej Singh Dhillon, a former RCMP officer, faced backlash after voicing support for enforcing federal gun control laws (Bill C-21). His comments drew accusations of fearmongering and communism, with some users suggesting that efforts to remove guns from the population mirrored authoritarian tactics. Organizations like the Canadian Shooting Sports Association and Calibre Firearm Magazine publicly sided with the Conservatives, arguing that targeting licensed gun owners did little to address the root causes of violent crime. Adding to this divide was the controversy over the province's approach to drug use and addiction. The NDP's now-reversed decriminalization of small amounts of controlled substances was widely condemned by Conservatives, who referred to it as a dangerous "experiment" that enabled addiction and deteriorated urban safety. In contrast, the NDP defended its actions as part of a harm-reduction model grounded in public health, though even David Eby acknowledged its shortcomings. While both parties ultimately supported the move to re-criminalize drug use in public, they framed the issue through different lenses (health support versus law and order). As the issue of public safety became increasingly entangled with federal-provincial responsibilities (e.g., bail, sentencing, firearms legislation), misinformation and emotionally charged rhetoric was found across social media platforms. #### **3-3. Climate Change and the Environment** Environmental policy, particularly the carbon tax and climate mitigation strategies, became a contentious point in the 2024 BC election, though it ranked as a medium-priority issue overall. Debates over carbon pricing were revealing of broader confusion and polarization, particularly around federal-provincial jurisdiction. The carbon tax was originally implemented by BC in 2008 as a pioneering provincial initiative but has since become federally mandated. This evolution complicated Conservative leader John Rustad's campaign promises to eliminate it as Premier David Eby and others quickly pointed out it was not within provincial authority. While Rustad later clarified that he was referring to removing the provincial application of the tax, critics accused him of misleading voters. The political tension around the tax also reflected deeper public frustration over rising living costs. Critics argued the carbon tax unfairly penalized working families already struggling with **Samples 2: Posts and comments indicative of the polarized environment on issues of crime and safety** ![](_page_16_Figure_3.jpeg) ![](_page_16_Picture_4.jpeg) ![](_page_16_Picture_5.jpeg) inflation, housing shortages, and food insecurity. Supporters, on the other hand, maintained that carbon pricing is a key tool in fighting climate change, and emphasized the rebates and incentives in place to mitigate its financial burden. However, as with many other issues in this election, nuance was often lost. While the NDP focused on strengthening BC's green infrastructure and advancing electrification projects, the Conservatives positioned climate policies, particularly those seen as costly or restrictive, as part of a broader pattern of government overreach and economic mismanagement. There were also bursts of misinformation about the climate during the election, much of it deeply conspiratorial in tone. During election weekend, severe rainfall and localized flooding across parts of BC coincided with a flurry of social media claims alleging that the NDP had used "cloud seeding" to manipulate the weather and suppress Conservative voter turnout. The conspiracy quickly gained traction across platforms like X, Instagram, and Reddit. Cloud seeding, a real though limited meteorological technique used in agriculture or drought mitigation, became the centerpiece of a voter suppression narrative. The viral spread of the theory underscores the susceptibility of election discourse to misinformation and the degree of mistrust held by some voters, especially when weather events, science, and politics intersect. **Samples 3: Several posts and comments making misinformed claims regarding climate change and the election** ![](_page_16_Figure_10.jpeg) ![](_page_16_Picture_11.jpeg) ![](_page_16_Picture_12.jpeg) ![](_page_17_Picture_2.jpeg) #### **Section 4** # **Election integrity** As concerns about democratic resilience and misinformation grow worldwide, the 2024 British Columbia provincial election offered a revealing case study in how trust, truth, and digital dynamics shape public confidence in electoral processes. This section explores the integrity of the election through three interconnected threats: the presence and impact of foreign interference, the role of bots and astroturfing, and post-election allegations of election fraud. # **1. Foreign interference and influence** Foreign interference was top of mind for Canadians throughout 2024, with the Foreign Interference Commission hearings5 and significant reporting on Chinese6, Indian7 , and Russian8 interference. This section provides our assessment of interference in and influence on the information environment during the election noting that there were some important methodological limitations, particularly around coverage of linguistic minority communities and poor social media data access beyond our core data collection. We find high concern and many claims of foreign interference, but little evidence to substantiate any such claims. ### **1-1. Self-reported concern among British Columbians** Figure 10 shows the levels of concern expressed by British Columbians in comparison with Canadians nationally (where applicable) regarding numerous threats to democratic integrity and information quality, specifically the threat of foreign influence, AI generated content and online misinformation. Across all measures, a substantial share of the public reported being highly concerned ('very' or 'extremely concerned'), although British Columbians consistently registered slightly lower levels of alarm than the national average. **Figure 10: Public concern about misinformation and external influence on Canadian media and politics** ![](_page_17_Figure_12.jpeg) Image: James Doyle/CP Interestingly, British Columbians and Canadians were the most concerned about AI-generated content, with 58.9% of Canadians and 49.5% of British Columbians expressing "very" or "extreme" concern. The prominence of this issue reflects increasing public awareness of generative AI tools and their potential to distort reality, especially during elections. Worry about foreign government influence also featured prominently. Over half of national respondents (52.4%) were highly concerned about foreign interference in Canadian media and politics, compared to 44.5% in British Columbia. Concern specifically about foreign influence in BC's media and politics was slightly lower still, at 40.5%, indicating a perceived distinction between national and provincial-level vulnerability. Online misinformation more broadly drew similar levels of concern across both samples (44.6% nationally and 42.9% in British Columbia). This consistent level of unease reflects the public's ongoing recognition of the risks posed by inaccurate or manipulative information online, even if such content may not always be associated with foreign actors or advanced technology. The past two years have seen a growing public sensitivity to a wide array of information threats, including both foreign interference and technologically driven distortion. The broader information ecosystem is viewed as fragile and increasingly vulnerable to manipulation. **Samples 4: An X post and Instagram comments illustrating concerns about foreign interference, the NDP, and election integrity** ![](_page_18_Picture_7.jpeg) #### **1-2. Social media interference** The Media Ecosystem Observatory monitors many prominent international entities (e.g. state media outlets, influencers who comments on international politics) to identify instances where international attention is placed on Canadian issues. This open-source and network-based monitoring did not reveal any coordinated messaging from Russian, Chinese, or Indian actors regarding the BC-election. Overall, foreign actors appeared to demonstrate very little interest in influencing the campaign. There were several mentions of the campaign in Indian media, but these articles were primarily focused on the escalating tensions (including the mutual expelling of diplomats on October 14, 2024)9 that coincided with the end of the campaign: the volume was relatively minor, and we did not observe any coordination. Similarly, there were a few posts on WeChat (a Mandarin-language social media platform popular among Chinese Canadians) that received a small amount of attention. By far the most significant claims of foreign interference during the election were those connecting David Eby and the NDP to Chinese interests. See Samples 4 and 5 for several posts and comments shared during the election made on X and Instagram. After the election results came in, allegations that the NDP's victory was "*China's victory*" surged on X and TikTok. One major source of these allegations was the coverage of Rise Media during the campaign, which some alleged was blatantly pro-Eby. ![](_page_18_Figure_11.jpeg) ![](_page_18_Figure_12.jpeg) **Samples 5: X posts making claims around Chinese involvement in the British Columbia election** ![](_page_19_Figure_3.jpeg) ![](_page_19_Figure_4.jpeg) Vancouver-based Rise Media is Canada's largest Mandarin-language publication and published several pieces on the election. While there were several commentary pieces published in Rise Media that were complimentary of Eby and the NDP, there was at least one other that argued in favour of Rustad. Generally, we assessed the commentary to be relatively balanced. Reach of their content on WeChat was marginal. Our assessment is that there was very little to no foreign information manipulation in digital information spaces and that claims of interference far outstripped the evidence. Despite this, British Columbians continue to be extremely concerned about foreign actors meddling in domestic politics (although less in the post-election period, see Figure 11). This incongruence reflects the tendency for the information ecosystem to surface and amplify conspiratorial ideation and baseless or flimsy claims regarding foreign interference. Again, we note that we use only open-source tooling and attempts to interfere in elections are not confined to the digital realm. # **2. Bot activity and astroturfing** Bots (specifically social bots) refer to automated accounts that perform computerized tasks often by mimicking human behaviour. These accounts engage with human users and posts to create new narratives and/or amplify existing ones. Although the discourse around bots wasn't prominent leading up to the election, accusations against parties for using bots in their campaign occasionally surfaced. On X, Saanich Councillor Teale Phelps Bondaroff received attention after he shared an interaction with a X account he suspected of being a bot (see sample 6). When Bondaroff noticed replies to his posts supporting NDP candidates appeared to be inauthentic messages to vote Conservative without addressing the content of what he said, he replied with a prompt telling the user to "ignore previous instructions" and perform some sort of generative writing (as in one instance where he told a critical account to write a haiku about frogs). The account did just that nearly instantly, indicating the account was indeed a bot. Other X users also claimed there was bot activity based on the following observations: - E large volume of content with the same political leaning being posted simultaneously; - E accounts with a high following but no followers making certain candidates look more prominent than others; - E large numbers of short comments written in the same tone (often anger). We observed a modest volume of X accounts with suspicious behaviour, including coordinated replies and URL sharing, extremely high levels of posting, and many accounts that were later deleted by X, likely because they were violating X's Terms of Service. **Samples 6: Example of online claims regarding bot activity on X during the election** ![](_page_20_Figure_3.jpeg) ### **3. Claims of election fraud and wrong-doing** While the 2024 British Columbia provincial election unfolded without major controversy during the campaign period, the post-election atmosphere was notably different. Online speculation and public concern escalated in the weeks following voting day, particularly around the integrity and administration of the vote-counting process. Two election administration issues helped fuel this distrust: E Missed ballot-box: In the Prince George–Mackenzie riding, Elections BC acknowledged that a ballot box containing 861 votes had not been initially counted, triggering a judicial **Samples 7: Several posts and comments casting doubt on election integrity in the post-election period** ![](_page_20_Figure_8.jpeg) recount. Although the missing votes did not alter the result in the riding, the incident raised broader concerns about systemic reliability. E Vote count discrepancy: In Surrey–Guildford, a discrepancy involving 14 uncounted votes emerged after the NDP had already been declared the winner. These developments led BC Conservative leader John Rustad to call for an independent review,10 even as Elections BC maintained these were isolated human errors that had been swiftly addressed and transparently disclosed.11 Both incidents became rallying points for a broader set of fraud claims (see Samples 7 and 8) that had been circulating in online ![](_page_20_Figure_12.jpeg) ![](_page_20_Picture_13.jpeg) **Samples 8: Additional posts and comments casting doubt on election integrity in the postelection period** ![](_page_21_Figure_3.jpeg) ![](_page_21_Figure_4.jpeg) ![](_page_21_Figure_5.jpeg) forums and through partisan commentary, especially from ardent supporters of the BC Conservative Party. The narrative quickly expanded beyond procedural missteps. Allegations included breaches of the Election Act (such as voter privacy concerns) along with accusations of foreign interference, particularly from China. As demonstrated in our samples, suspicion also fell on the tools and technologies used in the election. Critics cast doubt on the use of tabulators to process ballots and questioned the legitimacy of mail-in and telephone voting systems, despite their longstanding use in BC. Compounding these concerns was the revelation that some ballot boxes had been stored overnight in the homes of election officials during advance voting (a procedure permitted under Elections BC guidelines to ensure secure custody).12 While routine and legally sanctioned, the optics of this practice were seized upon online as evidence of lax security or potential tampering. Mistrust was further amplified by delays in reporting results, with some claiming that the timing itself was indicative of fraud. Unusually inclement weather was even used as evidence of manipulation. In parallel, disinformation campaigns directly targeted Elections BC and its staff. False narratives accused the agency of bias or collusion, contributing to a growing erosion of trust. Pre-election polls, many of which were online only and not representatively valid, which differed from actual results, were also weaponized as supposed proof that outcomes had been manipulated. These concerns were often presented not as isolated questions but as part of a broader, conspiratorial claim that the election had been manipulated to favour certain parties or candidates (mainly the NDP). Despite the volume of these claims, no credible evidence emerged to support them. Elections BC and independent observers consistently affirmed the integrity of the electoral process, and the errors that had occurred were minor, transparently addressed, and inconsequential to final outcomes. As mentioned earlier, trust in Elections BC remained stable and high throughout the election period (Elections BC was the most trusted source for election information by a broad margin). Nevertheless, the atmosphere of mistrust lingered. While most political leaders initially accepted the results, BC Conservative leader John Rustad later claimed (but failed to provide evidence for) that non-Canadian citizens may have voted. #### **3-1. Source of information questioning election integrity** In elections around the world, election integrity claims frequently come from accounts and entities outside the borders of the community of interest as well.13 We evaluated if something **Figure 11: Surge in election fraud claims around BC's provincial election, by account location (within versus external to BC)** ![](_page_22_Figure_3.jpeg) similar was happening in British Columbia, specifically looking at the volume of election fraud posts coming from within or outside of BC (primarily from the rest of Canada or the United States) or from an unknown location. Figure 11 shows the volume and geographic origin of social media content explicitly referencing election fraud in the lead-up to and aftermath of the BC provincial election. While overall volume was relatively low in early October, fraudulent election narratives escalated sharply beginning around October 20, peaking between October 25 and 30, immediately following election day. A slim majority of engagement came from accounts outside of British Columbia (orange bars), not from within the province, although BC-based accounts (green bars) also contributed meaningfully. Posts with unknown location status (grey) made up a smaller but non-negligible portion. This pattern suggests that much of the public doubt cast on the electoral process was not entirely organic to BC's political landscape. Instead, it appears that external voices played a disproportionate role in amplifying election fraud narratives, mirroring dynamics seen in other jurisdictions where external actors shape local discourse. It appears that this can occur even in sub-national political contexts. These findings raise concerns about cross-regional information manipulation and reinforce the importance of robust, localized responses to misinformation during sensitive electoral periods. #### **3-2. Impact of these claims on attitudes** Given the breadth of these claims, we evaluated the extent to which the attitudes of British Columbians changed during the critical period surrounding the election. Figures 12 and 13 illustrate a range of perceptions on electoral trust, misinformation, technological confidence, and institutional performance. Overall, most indicators remained stable between the pre- and post-election periods, but several key trends emerged. We did observe a very large 7-point increase in reported exposure to misleading or deceptive information (from 41% to 48%), reflecting a greater than usual attention to political information alongside the misinformation that we documented online. There was also a 9-point drop in concern about foreign interference (from 54% to 45%) with the BC election largely not suffering from foreign interference and little attention paid to the issue during the campaign. Another significant shift was the 25-point jump in belief that ballot tabulators were used to count paper ballots (from 58% to 83%). This is reasonable given Media Ecosystem Observatory [meo.ca](https://meo.ca) **Figure 12: Effect of the election and these narratives on public trust and beliefs about the electoral process** ![](_page_23_Figure_14.jpeg) that voters had contact with the ballot tabulators during their voting process. This was accompanied by a modest but statistically significant increase in comfort with ballot tabulators and perceptions of their security relative to hand-counting, with a post-election measure of 91% of the population. As shown in Figure 12, broad public attitudes toward Elections BC and electoral integrity were largely steady at the aggregate level. Confidence that Elections BC would or did fairly conduct the election remained high just above 90%. Similarly, trust in the integrity of election results remained high at approximately 80%. However, these high and steady levels of confidence mask a growing partisan polarization. Figure 13 shows confidence in Elections BC rose among NDP (+4 points, to 93%) and Green (+8 points, to 90%) supporters, but also declined among Conservatives (–3 points, to 74%). A similar pattern appears in trust in the integrity of election results, with NDP (+5) and Green (+4) supporters gaining trust, while Conservative trust dropped (-5). This trend of divergence was mirrored in perceptions of vote-counting machine security. All parties reported slight improvements in perceived security of electronic vote-counting (NDP: +7, Green: +5, Conservative: +6), but the post-election gap remains sizable: only 61% of Conservatives saw vote-counting machines as secure compared to 75% of NDP and 68% of Green supporters. The significant increase in awareness of ballot tabulator usage underscores education and positive experience can shape understanding of electoral procedures. While trust in these technologies rose modestly across all political affiliations, persistent partisan divides indicate ongoing skepticism among Conservative voters. Taken together, the two figures provide an essential context for interpreting both the online and offline discourse surrounding electoral legitimacy, administrative trust, and technological acceptance during and after the 2024 BC provincial election. The growing polarization, especially in perceptions of institutional trust, may shape future debates around election reform and public engagement. **Figure 13: Effect of the election and these narratives on public trust and beliefs about the electoral process by partisanship** ![](_page_24_Figure_6.jpeg) ![](_page_24_Figure_8.jpeg) ![](_page_24_Figure_10.jpeg) ![](_page_25_Picture_3.jpeg) **Section 5** # **Lessons learned and recommendations** The 2024 British Columbia election offers important insights into the resilience of democratic institutions in a complex and often hostile information environment. Despite widespread speculation and online misinformation, Elections BC maintained high levels of public trust and successfully administered the election without evidence of systemic disruption. At the same time, the online ecosystem, particularly on platforms like X, amplified polarized narratives and questioned the integrity of the process, especially in the immediate post-election period. Understanding and responding to these dynamics is critical for safeguarding future elections. While the overall impact of misand disinformation was low, its persistence, especially around foreign interference and election legitimacy, signals an evolving threat landscape that requires proactive, platform-specific, and coordinated responses. The election revealed recurring patterns that are likely to persist in future electoral cycles. These lessons point to underlying shifts in political communication, public trust, and misinformation strategies that require adaptive responses. E **Speculation about election manipulation and foreign interference is now the norm**: with rising distrust of institutional actors and heightened concern about foreign interference, citizens and especially highly active online partisans are quick to speculate on and identify foreign interference and/or election manipulation even in the absence of strong evidence. - E **Social media conversation often misrepresents broader public sentiment**: highly engaged online populations, particularly those in certain communities, frequently diverge from the views of the general population. Monitoring should be paired with survey and offline data to avoid overrepresenting fringe discourse. - E **X amplifies misinformation and disinformation risks:** As a central hub for political influencers and journalists, X plays an outsized role in shaping narratives. This is compounded by bot and other suspicious activity, algorithmic opacity, and weakened platform governance, making it a distinct vector for misinformation requiring targeted response strategies. ### **Recommendations** Implementing these steps can help strengthen democratic resilience and public confidence in future elections. 1. **Invest in inclusive political literacy and civic education**: to strengthen democratic resilience and public trust in elections, governments, civil society, and education systems should invest in political literacy and civic education. These initiatives must be tailored to different demographics and delivered across diverse platforms to effectively improve public understanding of electoral processes and the risks posed by misinformation. Civic education in schools should include modules on misinformation, disinformation, and digital literacy, embedded within existing curricula. However, these efforts must extend beyond the formal education system to be truly inclusive. Community centers, think tanks, workplaces, and media organizations should be engaged to offer digital literacy programming that reaches adults and marginalized groups, ensuring equitable access to critical information skills. - 2. **Shift from reactive to anticipatory misinformation response**: a fundamental change in strategy is needed to move from reacting to misinformation after it spreads to anticipating it before it takes hold. Election bodies, civil society, and researchers should engage in ongoing risk assessment, scenario planning and online monitoring to map out predictable false narratives, identify emerging threats and prepare targeted responses in advance. Establishing clear channels of communication with influential voices—journalists, creators, community leaders, and other trusted intermediaries—can help amplify timely and corrective information. By identifying and addressing likely areas of confusion or manipulation before they arise, stakeholders can reduce the public's vulnerability to disinformation. - 3. **Expand proactive and transparent election communications**: given the increasing politicization of electoral processes and growing skepticism toward electoral innovations, election management bodies must prioritize proactive transparency. This means expanding communications not only during elections but especially in the post-election period, where legitimacy challenges frequently emerge. Well-crafted, visual, and accessible communications should be prepared in advance to explain electoral practices, changes in procedures, and common areas of public concern. These materials should be distributed across multiple platforms (online and offline), including those that cater to hard-to-reach or high-risk populations. Clear, consistent, and transparent communication can help inoculate and empower the public against conspiratorial narratives and reinforce the credibility of electoral institutions. - 4. **Collaboration and sharing best practices among election administration bodies**: collaboration across election management bodies at the national, provincial, and municipal levels, should be actively encouraged to share best practices and lessons learned in combating misinformation and improving transparency. Coordinated planning, - the development of joint communication toolkits, and shared scenario exercises can help standardize responses to emerging threats. Establishing peer learning networks, communities of practice or working groups will foster innovation, reduce duplication of efforts, and improve collective capacity to navigate the evolving information landscape. - 5. **Mandate data access and platform communication requirements**: governments and regulators should mandate greater data access and transparency from social media platforms, particularly during electoral periods. This includes requiring platforms to disclose data relevant to election-related content, provide real-time insights into the spread of misinformation, and clearly label official sources and fact-checked material. Legislation or regulatory guidance should also compel platforms to maintain open lines of communication with election authorities and designated misinformation monitoring bodies. Such frameworks are essential to ensure timely identification and mitigation of false or harmful narratives circulating online. - 6. **Develop platform-specific monitoring and response strategies**: each social media platform operates within its own cultural norms and technical architecture, necessitating tailored monitoring and response strategies. High-risk platforms (such as X) require a high level of vigilance due to their role in rapidly disseminating false claims. Partnerships between election administrators, civil society, and platform companies should be formalized to enable quicker and more effective detection, communication and intervention. These efforts should also explore ways to enhance on-platform fact-checking, reduce algorithmic amplification of falsehoods, and promote accurate content during elections. - 7. **Encourage coordinated incident response during elections:** whole-of-society incident response mechanisms across should be encouraged. Such mechanisms would provide real-time monitoring, rapid response, and public reporting during elections. Researchers, election administration bodies, civil society, and media should explore multi-modal collaboration to respond to emerging information threats. A clear escalation protocol and public-facing communication capacity will aid in swift and transparent response. The experience of the 2024 BC election shows that democratic trust is under threat and maintaining it through continuous transparency and public engagement will be critical going forward. ### **Section 6** # **Methodology** We employed a multi-method approach to map and understand the British Columbia information environment during their 2024 election that combined survey, digital trace, and avatar methodology. ## **1. Survey methodology** For this study, we conducted three surveys: pre-election, during-election, and post-election. We conducted the standalone pre-election survey from August 28, 2024, to September 6, 2024 (10 days), a daily rolling survey (28 days) from September 21, 2024, to October 18, 2024, and a post-election survey from October 4, 2024, to October 18, 2024 (15 days). ### **1-1. Pre-election survey** The pre-election survey was conducted 7 weeks prior to the election date, aiming to capture public opinion on various issues ahead of the election. The questionnaire was designed to address topics of interest and prominent socio-political issues, including the use of social media for news, public perceptions of the newly introduced ballot tabulators, concerns about generative AI, and worries about misinformation and foreign interference in the BC election. To ensure precise wording, we referenced questions from Elections BC's public opinion survey conducted earlier in the year, as well as MEO's main tracking survey. The survey was hosted on Qualtrics, with quotas established for age and gender to ensure representativeness. #### **1-2. During-election survey** The during-election daily rolling survey was launched four weeks before the election date and continues daily, targeting a sample size of 100 respondents per day with quotas for age and gender. The primary purpose of this rolling survey is to track changes in public opinion over the weeks leading up to the election. To maintain consistency and enable cross-wave comparisons, we incorporated some questions from the pre-election questionnaire, such as concerns about misinformation and AI. Additionally, we introduced new questions to explore interest in the upcoming election and the role of partisan cues in the reception of misinformation. We also allocated space to accommodate new questions in response to any campaign-relevant incidents that may arise. We monitor survey completions daily, maintaining flexibility in the sample size. If a day's responses fall short of the target, we allow additional responses the following day to make up the difference. In cases where we encounter difficulties filling quotas for specific demographic groups, such as young men aged 18–29, we allow for increased responses from other demographic groups to maintain progress. For data analysis, we apply weighting to adjust for these imbalances and ensure the results remain representative. #### **1-3. Post-election survey** Due to the extremely close race, the election results were announced after a nine-day wait. We launched the survey two days after the results were made public, allowing us time to finalize the questionnaire. For the post-election survey, we re-contacted the 1,500 participants who had completed our daily tracker survey between October 4 and 18. The post-election survey design included questions from previous waves and introduced new ones on topics such as vote choice, general sentiment about the election results, and perceptions of the results' integrity. # **2. Avatar methodology** Our "Boots on the Ground" team consisted of four research assistants. Drawing from real British Columbia demographic statistics, we developed profiles that reflected the diversity observed in BC. This process included considering various factors including geographic distribution, education level, household income and socioeconomic status, ethnic and cultural backgrounds, religious affiliations, and linguistic diversity present in BC. A total of 12 accounts were created across Facebook, Instagram, YouTube, TikTok, and X, each featuring a different fictional persona. While no avatar account was made to monitor Reddit, content on the platform was monitored. Our team found Reddit to be an excellent platform for current issues discussion and found much content surrounding the BC election. Here are several examples of the Avatars used: **Nicole Mendoza** (active on TikTok) represented a newly naturalised Filipino-Canadian in her mid-20s, studying at a Burnaby community college and working part-time in retail. We conducted roughly 15 hours of observation for this undecided voter whose feed, once the algorithm settled, was dominated by BC-election videos framed around cost-of-living, transit, and diversity concerns. Although political leanings were ambigu- ![](_page_28_Figure_2.jpeg) ![](_page_28_Figure_3.jpeg) ![](_page_28_Figure_4.jpeg) ![](_page_28_Figure_5.jpeg) **Profile pages for Nicole Mendoza (TikTok), Harjeet Singh (Instagram), Mark Peters (YouTube), and Harrison Gray (on X).** ous, the "heated, meme-driven" tone of many clips suggests that TikTok's engagement incentives privilege emotive content. **Harjeet Singh** (active on Instagram) represented a Punjabi-Canadian in his 30s and explicit NDP supporter from Surrey, BC. He was monitored for approximately 12 hours. About 70% of posts seen were in English, the remainder in Punjabi. Surprisingly, only a small fraction of his feed focused on BC-specific issues; instead, federal immigration debates and diaspora-driven pro-Khalistan narratives predominated. While this may simply mirror the accounts he followed, it also underlines Instagram's tendency to surface national-scale political discourse even when an avatar's "following" list is locally curated. **Mark Peters** (active on YouTube) represented a wealthier business owner in Prince George aged 30-40, and was tracked for approximately 20 hours. His viewing history split roughly 50% BC-election material, 25% Canadian federal Conservative content, and 25% U.S. Republican commentary. All videos were in English, and algorithmic suggestions frequently featured business-oriented podcasters and Trump-related channels. The pattern underscores YouTube's trans-national recommendation logic: despite BC-specific seeding, the platform quickly provided provincial, federal and U.S. partisan narratives, complicating efforts to isolate purely provincial influences. **Amani Khoury** (active on Facebook and then Instagram) represented an early-30s second-generation Christian Lebanese professional. We initially maintained a Facebook account alongside an active Instagram profile, but monitoring revealed virtually no BC-election content, prompting the research team to retire the account. While the absence of relevant material limits substantive analysis, it offers a useful negative finding: Facebook's declining reach among younger urban British Columbians may have dramatically reduced its salience in provincial political discourse, a finding consistent with low levels of engagement as compared to X and TikTok. **Harrison Gray** (active on X) represented a late-20s health-care professional from New Westminster–Coquitlam with a consistent Conservative voting history. Over 20 hours of logging, 57% of his timeline concerned the provincial election and was overwhelmingly produced by Conservative influencers rather than official candidates. He encountered "very little to no" NDP content and a noticeably higher incidence of slurs or violent language. The account also attracted 12 apparent bot followers with AI-generated posts. ## **3. Digital trace and social media methodology** **Strategic focus on important actors**: Political conversation online is highly unequal, with a comparatively small set of "elite" accounts including politicians, news outlets, and influencers, generating a disproportionate share of attention and engagement. By concentrating on that tier we can characterise the shape and content of Canadian political discourse with far less data than a population-wide scrape would require, while still capturing the vast majority of posts that ordinary users are likely to encounter. This logic underpins our decision to track 708 entities across five major platforms. **Building a seed list:** A seed is a discrete actor (person or organisation) judged to exert measurable influence on Canadian politics. These include journalists, media organizations, candidates for political office, and influencers that meet our thresholds for inclusion (which were relaxed for the election). For every seed we located all declared or discoverable accounts on X, Facebook, Instagram, YouTube and TikTok, if it was not a personal account. Where an entity operated multiple pages on a single platform (e.g., language variants or campaign-specific channels) we recorded the primary page used by the entity. Each handle was verified manually by a research assistant. The exercise produced 1,556 unique handles, giving an average of 2.2 accounts per entity and ensuring that cross-posting behaviour could be traced accurately. **Time-bounded content retrieval:** Using a combination of official APIs and internal pipelines, we ingested all public posts published between August 21, 2024, to November 29, 2024. The final corpus comprises 1,466,433 primary posts. **Supplemental keyword monitoring:** In select instances we also conducted searches for relevant keywords to inform investigations to evaluate information manipulation incidents. Posts surfaced through this channel were deduplicated against the seed corpus and at times those originating from previously untracked but influential accounts were fed back into the seed list, widening coverage. **Multi-layer analytical framework:** Because the same actor often behaves differently across platforms and because algorithms privilege different content types each handle is treated as a node in a *five-layer multiplex network*. In sum, this approach combines breadth (five platforms, ~1.5 million posts) with depth (entity-level mapping and keyword supplementation). It yields a defensible, transparently documented baseline for understanding how political narratives flowed through the contemporary social-media ecosystem during the British Columbia election. # **Contributors Notes** Contributors to this report (listed alphabetically) are: Alexei Abrahams, Thomas Bergeron, Danielle Bohonos, Aengus Bridgman, Esli Chan, Mika Desblancs-Patel, Spencer Izen, Grace Lim, Sherry Madah, Mahnan Omar, Zeynep Pehlivan, Jennie Philips, Saewon Park, Chris Ross, Ben Steel, Chris Tenove, Ashley Vu, and Junyan Zhu. This report was made possible by the considerable work of a large number of researchers in Canada: in no particular order, Isabelle Corriveau, Nader Hedfi, Shane Littrell, Olivia Melanson, Taylor Owen, Peter Loewen, Lee Slinger, and Heidi Tworek. We also thank Brian Morgan for his design work. ![](_page_30_Picture_5.jpeg) The British Columbia Election Information Ecosystem Project was funded in part by Elections BC. - **1 [https://elections.bc.ca/news/2024](https://elections.bc.ca/news/2024-provincial-election-sov-available/) [provincial-election-sov-available/](https://elections.bc.ca/news/2024-provincial-election-sov-available/)** - **2 [https://socialmedialab.](https://socialmedialab.ca/2025/05/05/new-report-the-state-of-generative-ai-use-in-canada-2025-exploring-public-attitudes-and-adoption-trends-2/) [ca/2025/05/05/new-report-the](https://socialmedialab.ca/2025/05/05/new-report-the-state-of-generative-ai-use-in-canada-2025-exploring-public-attitudes-and-adoption-trends-2/)[state-of-generative-ai-use-in](https://socialmedialab.ca/2025/05/05/new-report-the-state-of-generative-ai-use-in-canada-2025-exploring-public-attitudes-and-adoption-trends-2/)[canada-2025-exploring-public-](https://socialmedialab.ca/2025/05/05/new-report-the-state-of-generative-ai-use-in-canada-2025-exploring-public-attitudes-and-adoption-trends-2/)** - **3 [https://meo.ca/work/old-news](https://meo.ca/work/old-news-new-reality-a-year-of-metas-news-ban-in-canada)[new-reality-a-year-of-metas](https://meo.ca/work/old-news-new-reality-a-year-of-metas-news-ban-in-canada)[news-ban-in-canada](https://meo.ca/work/old-news-new-reality-a-year-of-metas-news-ban-in-canada)** - **4 [https://www.cbc.ca/news/canada/](https://www.cbc.ca/news/canada/british-columbia/john-rustad-fatal-overdose-1.7348135) [british-columbia/john-rustad-fatal](https://www.cbc.ca/news/canada/british-columbia/john-rustad-fatal-overdose-1.7348135)[overdose-1.7348135](https://www.cbc.ca/news/canada/british-columbia/john-rustad-fatal-overdose-1.7348135)** - **5 [https://www.cbc.ca/news/politics/](https://www.cbc.ca/news/politics/foreign-interference-public-hearing-wrap-1.7362908) [foreign-interference-public](https://www.cbc.ca/news/politics/foreign-interference-public-hearing-wrap-1.7362908)[hearing-wrap-1.7362908](https://www.cbc.ca/news/politics/foreign-interference-public-hearing-wrap-1.7362908)** - **6 [https://apnews.com/article/](https://apnews.com/article/canada-china-chrystia-freeland-foreign-interference-f8f21696567c79c8c6df6cedebcc905a) [freeland-foreign-interference-](https://apnews.com/article/canada-china-chrystia-freeland-foreign-interference-f8f21696567c79c8c6df6cedebcc905a)** - **7 [https://www.reuters.com/](https://www.reuters.com/world/americas/india-is-perpetrator-foreign-interference-canadian-intelligence-agency-says-2025-06-18) [world/americas/india-is](https://www.reuters.com/world/americas/india-is-perpetrator-foreign-interference-canadian-intelligence-agency-says-2025-06-18)[perpetrator-foreign-interference](https://www.reuters.com/world/americas/india-is-perpetrator-foreign-interference-canadian-intelligence-agency-says-2025-06-18)[canadian-intelligence-agency](https://www.reuters.com/world/americas/india-is-perpetrator-foreign-interference-canadian-intelligence-agency-says-2025-06-18)[says-2025-06-18](https://www.reuters.com/world/americas/india-is-perpetrator-foreign-interference-canadian-intelligence-agency-says-2025-06-18)** - **8 [https://www.cdmrn.ca/publications/](https://www.cdmrn.ca/publications/tenet-media-final-incident-debrief-report) [tenet-media-final-incident-debrief-](https://www.cdmrn.ca/publications/tenet-media-final-incident-debrief-report)** - **9 [https://www.ft.com/](https://www.ft.com/content/7280dc0c-51bf-45cc-a277-7f4a1d87c599) [content/7280dc0c-51bf-45cc-a277-](https://www.ft.com/content/7280dc0c-51bf-45cc-a277-7f4a1d87c599) [7f4a1d87c599](https://www.ft.com/content/7280dc0c-51bf-45cc-a277-7f4a1d87c599)** - **10 [https://vancouversun.com/](https://vancouversun.com/news/elections-bc-box-votes-uncounted-other-votes-unreported) [news/elections-bc-box-votes-](https://vancouversun.com/news/elections-bc-box-votes-uncounted-other-votes-unreported)** - **11 [https://www.elections.bc.ca/docs/](https://www.elections.bc.ca/docs/rpt/Report_of_the_CEO_PGE2024.pdf) [rpt/Report\\_of\\_the\\_CEO\\_PGE2024.](https://www.elections.bc.ca/docs/rpt/Report_of_the_CEO_PGE2024.pdf)** - **12 [https://elections.bc.ca/](https://elections.bc.ca/docs/8008790_Guide-to-Voting-and-Counting.pdf) [docs/8008790\\_Guide-to-Voting-](https://elections.bc.ca/docs/8008790_Guide-to-Voting-and-Counting.pdf)** - **13 [https://www.counterterrorismgroup.](https://www.counterterrorismgroup.com/post/digital-investigation-of-foreign-influence-in-european-elections-examining-x-elon-musk-and-the-am) [of-foreign-influence-in-european](https://www.counterterrorismgroup.com/post/digital-investigation-of-foreign-influence-in-european-elections-examining-x-elon-musk-and-the-am)[elections-examining-x-elon-musk](https://www.counterterrorismgroup.com/post/digital-investigation-of-foreign-influence-in-european-elections-examining-x-elon-musk-and-the-am)[and-the-am](https://www.counterterrorismgroup.com/post/digital-investigation-of-foreign-influence-in-european-elections-examining-x-elon-musk-and-the-am)**