The Receipt
Every major Canadian news outlet is rated equally accurate by outside fact-checkers [5][–13]. Canadians who actually watch CBC, CTV, and Global trust them at near-identical, very high rates [1]. Yet trust in Canadian media is at a record low, split by a 54-point partisan gap [1]. The distrust comes overwhelmingly from people who don't watch.
The trust problem isn't in the articles. It's in which articles exist. Every editorial standard, ombudsman, and regulator in Canada governs how stories are covered. Nothing governs which stories are selected for coverage in the first place. The accountability architecture has a hole in it exactly where the trust problem sits.
Impartiality claims can't fix this because they don't reach the catalogue. The only response that can reach the problem is telling people what you choose to cover, what you don't, and why.
You've probably had the argument. Someone says CBC is biased. Someone else says it's the most trusted outlet in the country. Both cite real numbers. Both are right.
Pollara's 2026 Trust in Media study found CBC at +51 net trust [1]. The same study found a 69-point gap between Liberal voters who trust CBC (+78) and Conservative voters who barely do (+9) [1]. Across all media, the partisan trust gap is 54 points [1]. In 1992, the equivalent gap between Liberal and Reform voters was 29 points [3]. It has nearly doubled.
Reuters puts overall Canadian trust at 37%, down 18 points since it began tracking in 2016 [2]. Among English-speaking Canadians, 35% [2]. Statistics Canada's 2023 social survey found only 37% reporting high confidence in Canadian media [4]. The generational pattern has flipped: 65% of seniors trust media versus 43% of 18-to-34-year-olds [1]. In the 1990s, younger Canadians were more trusting.
The question is what "bias" is actually measuring.
It's Not Accuracy
If the trust problem were about getting facts wrong, outlets that outside reviewers rate as more accurate would earn more trust. They don't.
Media Bias/Fact Check, the most widely used independent media reviewer, rates every major Canadian mainstream outlet "High" for factual reporting [5][–11]. Here is what that looks like alongside the trust data:
| Outlet | MBFC Accuracy | Pollara Net Trust | Liberal–Conservative Gap |
|---|---|---|---|
| CBC | High [5] | +51 | 69 pts |
| CTV News | High [6] | +50 | 41 pts |
| Global News | High [7] | +47 | 40 pts |
| Globe and Mail | High [8] | +38 | 34 pts |
| National Post | High [9] | +29 | 23 pts |
| Toronto Star | High [10] | +23 | 35 pts |
| Canadian Press | High [11] | +39 (2025) | 44 pts (2025) |
| Rebel News | Mixed [12] | 0 | −16 pts |
| True North | Mixed [13] | n/a | n/a |
Trust data: Pollara Strategic Insights, 2026 [1] and 2025 [3]. Accuracy ratings: Media Bias/Fact Check [5][–13].
The accuracy column is a constant. The trust column is a variable. Whatever is driving a 69-point partisan split on CBC, it is not factual error. CBC carries the same accuracy rating as every other mainstream outlet on the list.
Accuracy does differentiate the mainstream from the fringe. Rebel News is rated "Mixed" [12]. True North is rated "Mixed" [13]. But within the mainstream set of outlets Canadians actually argue about, accuracy explains nothing. When Gallup and the Knight Foundation asked Americans in open-ended format why they distrust news, inaccuracy and bias came in nearly tied (45% and 42%), and people routinely mix the two together [16]. People say "inaccurate" when they mean something else.
The professional bias raters can't even agree on where outlets sit. MBFC rates the Globe and Mail "Right-Center" [8]. Ad Fontes Media rates it "Middle" [15]. MBFC rates CBC "Left-Center" [5]. Ad Fontes labels it "Middle" [14]. If trained reviewers using formal rating systems can't agree, what audiences call "bias" is something the professional instruments aren't capturing either.
It's Not the Content
If the trust problem were detectable bias inside published stories, the people most exposed to the content would be the most distrustful. They are the most trusting.
Pollara tested trust among actual viewers. Among Canadians who regularly watch CBC, net trust is +87. CTV viewers: +86. Global viewers: +87. Weather Network viewers: +87 [1]. Near-identical rates, regardless of outlet.
The distrust comes from non-viewers. Canadians who don't watch CBC form their impression of it from somewhere other than its journalism. That impression is strongly negative among Conservative voters (+9 net trust, versus +78 among Liberal voters) [1]. The people who watch it trust it. The people who don't watch it distrust it.
Researchers have a name for this pattern. The "hostile media effect," first documented in 1985 [17], shows that partisans perceive identical neutral coverage as biased against their side. A review of 34 follow-up studies confirmed the pattern holds consistently and gets stronger the more people care about the topic [18]. In 2024, a study published in PNAS Nexus found that partisan gaps in judging what's true held firm even when people were paid to get the answer right. Sixty-six to 78% of the gap persisted regardless of incentives [19]. People don't just say coverage is biased. They sincerely believe it is, regardless of what the content actually says.
The strongest case against this reading deserves full weight. The Senate Standing Committee on Transport and Communications, after a two-year study with roughly 60 witnesses, recommended in June 2026 that CBC commission annual third-party bias audits [20]. Dave Snow of the University of Guelph examined 209 CBC corrections published between 2021 and 2026 and found that of eight substantive partisanship-related corrections, seven ran against conservatives and one favoured Liberals [21]. These are real findings. But they are corrections reviews at one outlet, not side-by-side comparisons of what different outlets actually cover. They do not explain why trust varies by up to 69 partisan points across outlets that outside reviewers rate as equally accurate. No study does, because the comparison has never been conducted.
Systemic Bias
If it's not accuracy and it's not the content, what's left?
Researchers identified the answer in 2000. D'Alessio and Allen broke media bias into three kinds: how a story reads once it's published, how much attention it gets, and which stories make it to print in the first place [22]. That third kind is the one nobody measures. Groeling, writing in the Annual Review of Political Science, explains why: you can't count the stories that were never written. To measure what's missing, you'd need to know every story that could have been covered, and nobody has that list [23].
This is what the trust data is pointing to. When Reuters asked people who distrust news to explain why, 67% cited bias, spin, and agendas [24]. When Reuters surveyed 26 countries in 2026, 70% said media owners and politicians exert "undue influence" on the news [2]. These are not complaints about factual errors in published stories. They are complaints about what gets covered and why.
Call it systemic bias. Not because individual journalists are biased. The accuracy data shows they're doing their jobs. But because editorial choices about what to cover produce invisible patterns in the public record without anyone violating any standard. The parallel is structural: systemic risk in banking isn't about individual bad banks, it's about how the system is wired. Systemic bias in media isn't about individual bad journalists. It's about which stories reach the public and which don't, and the fact that nobody tracks, governs, or discloses that process.
Nobody's Watching
Every editorial standard, ombudsman, and regulator in Canada governs treatment. None governs the catalogue.
CBC's Journalistic Standards and Practices, the most detailed editorial code in Canadian media, defines five principles: accuracy, fairness, balance, impartiality, and integrity [25]. Its "balance" principle calls for "a wide range of subject matter and views," but this covers the range of opinions in stories they've already decided to run. Nothing in the document commits CBC to cover specific topics or types of events. The Globe and Mail's code requires that "news operations must be, and must be seen to be, impartial" [26]. The Toronto Star prescribes objectivity as a principle. CTV, Global, Postmedia, The Canadian Press, and La Presse follow the same pattern. Every code governs the same thing: how a story is handled once it's been selected. None addresses the selection itself.
The accountability mechanisms match. The CBC Ombudsman reviews published content against the JSP. The office's settled position is that "which stories need to be covered" is a management decision, not something the Ombudsman rules on [29]. In September 2024, a complainant raised exactly the kind of argument this article describes, saying that CBC systematically ignored perspectives unfavorable to the government in its Emergencies Act inquiry coverage. Ombudsman Jack Nagler explicitly refused to rule on that question. "I will confine my adjudication to the three stories at hand," he wrote [28]. He assessed the published content, found no violations, and stopped. The question about which stories get covered was raised and set aside.
The National NewsMedia Council, Canada's voluntary media watchdog, states in writing that it "does not deal with overarching trends in reporting" [27]. CBC, CTV, and Global are not members [27]. The CRTC regulates broadcaster licences, not news content.
The Broadcasting Act requires "varied and comprehensive" programming and "a reasonable opportunity for the public to be exposed to the expression of differing views on matters of public concern" [30]. No court has read this as a duty to cover specific topics. The Online News Act imposes no coverage obligations. It explicitly protects editorial independence without creating accountability for how that independence shapes what gets covered [31].
The international test is definitive. The BBC's Editorial Guidelines contain the strongest commitment to coverage breadth found anywhere in a comparable system: a pledge that "no significant strand of thought" be "knowingly unreflected or under-represented" [33]. In 2025, a researcher who had analyzed 14,000 hours of BBC programming filed a complaint alleging systematic under-representation of pro-Brexit perspectives over two decades. Ofcom declined to investigate. The UK High Court upheld Ofcom's refusal, ruling that impartiality rules apply to individual programs, not to what a broadcaster covers overall [32].
Even the strongest external impartiality regulator in any comparable democracy confirmed it cannot rule on the catalogue. Canada has no Ofcom equivalent for news content. The strongest regulator in a peer country ruled the catalogue out of scope. Canada hasn't built the weaker version.
The Fix Nobody Publishes
The current strategy for maintaining media trust is claiming impartiality in the stories they publish. The data shows that trust is fine at that level. Viewers trust what they watch at +86 to +87 [1]. Trust is collapsing at the catalogue level. Impartiality claims cover the stories. They don't cover which stories get chosen. The strategy is aimed at the wrong problem.
Governing the catalogue isn't the answer either. Nobody can define the full universe of stories that could have been covered [23], and the Keighley ruling confirms no regulator can rule on what should have been covered [32].
But you can make the catalogue visible. Telling people what you choose to cover, what you leave out, and why is the only fix that reaches where the trust problem actually lives. Not regulating the catalogue. Making it transparent. An outlet that says "we are impartial" is making a claim the system can't verify and audiences increasingly don't believe. An outlet that says "here is what we cover, here is what we don't, and here is why" is making a claim audiences can check for themselves.
This publication operates on that principle. Our selection criterion is disclosed: we cover entities that hold authority, whether statutory, regulatory, fiduciary, or fiscal. We do not cover opposition parties because they do not hold the mechanisms we document. That is not impartiality. It is a stated editorial logic you can see, disagree with, and test against our output. It does not solve the catalogue gap for Canadian media. It shows that the alternative to an impartiality claim is not partisanship. It is transparency about what you choose to cover.
The strongest case against this position: Reuters' 2026 data shows 45% of Canadians still prefer news that "doesn't take sides," versus 22% who prefer news reflecting their own view [2]. People still say they want impartial news. The hostile media effect persists even when audiences know an outlet's perspective [35]. Disclosure as a trust-building strategy is a hypothesis with supporting evidence, not a proven cure [34]. But the status quo is failing measurably: claiming impartiality over a dimension that isn't measured, governed, or made visible has cost Canadian media 18 points of trust in a decade [2].
What Would Change This Assessment
Four conditions would require this analysis to be revised:
- If a mainstream Canadian outlet's trust score tracked its accuracy rating regardless of the reader's politics, the case that accuracy drives trust would hold and this argument would weaken. Current data shows the opposite: accuracy is constant across the mainstream; trust varies by partisanship [1][5][–11].
- If an enforceable editorial standard existed at any news organization committing it to cover specific topics or categories, the claim that "nothing governs the catalogue" would need qualifying. None was found. The closest version, the BBC's breadth commitment, was confirmed unenforceable by the UK High Court [32].
- If any accountability mechanism had ruled on a complaint that a topic was systematically not covered and declared a breach with consequences, the accountability gap would be narrower than described. None has. The CBC Ombudsman was presented with exactly this kind of complaint and explicitly confined his review to the published content [28].
- If a rigorous, multi-outlet study showed that mainstream Canadian outlets cover different topics in ways that match the partisan trust gap, the argument that distrust tracks perception rather than actual coverage decisions would need rethinking. No such study exists. The Senate committee recommended one [20]. It has not been conducted.