In their book Manufacturing Consent, Edward Herman and Noam Chomsky argue that the mainstream media functions largely as a propaganda arm for the state. When the war drum beats, the corporate media tows the government’s line, censoring facts that don’t fit the official narrative.
Outside of war, media bias is typically less overt. But to the careful observer, it can still be discerned. In this case, our careful observer is Canadian oil critic Regan Boychuk.
Boychuk lives in Calgary — a prairie city that is famous for two things. Calgary hosts the world’s largest rodeo. And it is the corporate heart of the Canadian oil business. Calgary … home to cowboys and crude-oil CEOs.
As you might guess, our story of media censorship is not about cowboys. Calgary’s main newspaper, the Herald, is staunchly pro-oil. And that means its editorial pages are filled with oilpatch jingoism. However, the rest of the paper is an archetype of neutral reporting.
Just kidding.
Unsurprisingly, the Herald’s pro-oil stance shapes the content that appears in the paper. This post takes a quantitative look at the editorial ‘curation’.
Most of the heavy lifting has been done by Boychuk, who had the brilliant idea to track the reporting of environmental journalist Mike De Souza. Between November 2010 and July 2013, De Souza wrote a series of articles documenting scandals related to the Canadian oilpatch, and its staunch defender, the Harper government.
At the time, De Souza was working for Postmedia, a news conglomerate that operated a wire service for its many subsidiaries. So when De Souza’s pieces were published, they were delivered to local papers like the Ottawa Citizen, the Edmonton Journal, and the Calgary Herald.
Here’s the catch. Although owned by the same conglomerate, these local papers had leeway to edit (or shelve) their wire-service articles. The result, Boychuk realized, was a controlled setting to analyze media censorship. Earlier this year, Boychuk published his findings in a piece called ‘Proximity to Power: The oilpatch & Alberta’s major dailies’.
My contribution here is mostly visual. I’ve taken Boychuk’s investigation and translated it into charts. The results largely speak for themselves. As De Souza’s articles approached the center of Canadian oil-and-gas power in Calgary, they were increasingly gutted, and their message changed. It’s a fascinating case study of how business interests shape the news.
The geography (and geology) of Canadian oil-industry power
This post is mostly about the myopia of human bickering. But since that’s a depressing topic, let’s start with something more majestic. Let’s frame the ideological landscape of the Canadian oilpatch by looking at the big picture of fossil fuels.
Long before humans existed, life on Earth was doing chemical magic, taking energy from the sun and converting it into biomass. In the ancient seas, dead organisms fell to the seabed where they slowly accumulated. Over the eons, the Earth ground and compressed this seafloor biomass until its chemistry changed into the hydrocarbons we call ‘oil’.
In many regions, this oil remained locked under the sea. But in other places, moving continents and changing climates pushed the seabed above water, where it awaited discovery by a scrappy group of primates.
The North American plains are one of these above-ground-waiting-to-be-discovered places.
One hundred million years ago, though, the great plains were not ‘plains’ at all. They were a sea that ran the length of today’s North America, connecting the Arctic Ocean to the Gulf of Mexico. On the seabed, biology and geology did their thing, slowly forming an immense deposit of oil. When the poles later froze and sea levels sank, the whole area rose above land, awaiting primate infestation.
It’s here that we return to human myopia. Roughly ten thousand years ago, the great plains were discovered by indigenous peoples. Much later, Europeans conquered the territory, setting up a nation-state called ‘Canada’. In the 20th century, Canadians began to exploit the fossil fuels underfoot, drilling hundreds of thousands of oil wells, building teeming oil-driven cities, and, of course, squabbling about the business of exploiting oil.
Before we focus on this fossil-fuel bickering, though, let’s connect the big picture of geology and geography to the Canadian ideological landscape. Due to quirks of geology, oil formed in the west of North America. But due to quirks of human history, Europeans conquered the continent from the east. As a consequence, eastern centers of power — cities like Toronto and Ottawa — were established long before western oil was discovered. And so these cities never became dominated by the oil business.
The same was not true for the western cities of Edmonton and Calgary. At the turn of the 20th century, both places were rural backwaters — settlements with a few thousand people. But the discovery of oil changed that. By the late 20th century, Edmonton had grown into a sprawling industrial hub for the Canadian oilpatch. And Calgary became the oil business’ corporate epicenter, with a downtown brimming with oil-company headquarters. In short, the politics of both cities became dominated by the oil business.
Pulling together this interplay of geology and geography, Figure 1 shows the big picture of Canadian oilpatch myopia. In Canada’s east sit the cities of Toronto and Ottawa — centers of corporate and government power with no oil to their name. In the west sit Edmonton and Calgary — cities that are geographically in the thick of the Canadian oilpatch, and ideologically under its spell.
Looking at this map of Canada, keep the geography in mind as we turn to the ideologically charged business of publishing the news. We’ll watch Mike De Souza’s environmental reporting get sliced and diced as it leaves Postmedia (headquartered in Toronto) and journeys to papers in Ottawa, Edmonton and Calgary.
Of course, it’s conceivable that our three papers — the Ottawa Citizen, the Edmonton Journal, and the Calgary Herald — could treat De Souza’s work with an equal hand. But as you’ll see, that’s not what happened.
Environmental reporting during the Harper years
To set the stage for De Souza’s journalism, let’s wind the clock back to 2006. In Canada, it was a year of jarring political change — the moment when neocon politician Stephen Harper became Prime Minister.1 It’s not an exaggeration to say that Harper — a former employee of Imperial Oil and resident of Calgary — ran the country for the benefit of oil companies. As you’d expect, a string of controversies ensued. And De Souza was there to document the political flames, publishing a string of exposés.
In 2014, De Souza left Postmedia, prompting environmental magazine The Narwhal to write a retrospective ranking his twenty most important pieces. Seeing the opportunity to study censorship, muckraking investigator Regan Boychuk took these twenty pieces and tracked down how (or if) they’d been published by the Ottawa Citizen, the Edmonton Journal, and the Calgary Herald.
In ‘Proximity to Power’, Boychuk conducted an article-by-article analysis of the censorship. In this post, we’ll leave the specifics of De Souza’s articles behind. Instead, we’ll take a bird’s-eye view of how his reporting was censored as it approached the heart of the Canadian oil business.
Saving space? … Or saving face?
Every day, newspaper editors face a dilemma: far more stories are written than can possibly be published. The result is a kind of constant ‘editorial churn’ in which numerous worthy stories are axed.
Turning to Mike De Souza’s environmental reporting, editorial churn means that for banal, space-saving reasons, some of his stories will inevitably get cut as they journey from paper to paper. But amidst this churn, there could also be politically motivated censorship. So how can we separate this censorship from the non-political churn?
Regan Boychuk has devised a simple solution: we see if the ‘churn’ has a geographic pattern. We watch a sample of De Souza’s reporting leave Postmedia, and then observe how the articles get cut as they travel across the country. When we carry out this exercise, we find that the editorial ‘churn’ has a conspicuous direction.
Figure 2 shows the pattern. As we approach the heart of oil-and-gas power in Calgary, newspapers mysteriously run out of space for De Souza’s reporting. Of the 20 De Souza articles that left Postmedia, the Ottawa Citizen was able to publish 14, the Edmonton Journal had room for 12, while the Calgary Herald had space for only 11. Funny. When the Herald ‘saves space’, it looks a lot like saving the oil business’ face.
Diving shallow
Before we lay charges of ‘censorship’, we should consider some non-incendiary explanations for the pattern in Figure 2.
Here’s a possibility: maybe the Herald opted for depth over breadth. In other words, the Herald might have published fewer of De Souza’s articles because it was saving space for long-form content. If so, then counting words (instead of articles) should remove our geographic trend.
Except that it doesn’t.
Figure 3 shows the pattern. When we count the number of De Souza words published, we again find that the editorial churn has a geographic direction. As we journey to the corporate epicenter of the Canadian oilpatch, newspapers spill less and less ink publishing De Souza’s work.
Back-page news
Although the case for ‘objectivity’ looks bleak, let’s continue to hand the Herald olive branches. When it comes to De Souza’s work, maybe the Herald is substituting prominence for quantity — making a splash by publishing a few pieces at the front of the paper.
Unfortunately, the opposite seems to be true. Not only did the Herald print fewer De Souza articles, it tended to bury these pieces at the back of the paper. Figure 4 runs the numbers.
Looking at the evidence, the contrast between the Citizen and the Herald is stark. In the Citizen, most of De Souza’s pieces were published in the first four pages of the paper. But in the Herald, none of De Souza’s pieces graced the first four pages, and several were placed deep within the paper. In short, if the Herald editors aren’t biased against De Souza’s environmental reporting, they have a funny way of showing it.
Headlines hacked
Forging ahead, perhaps there’s a more creative way to show that the Herald really cares about De Souza’s journalism. Maybe the proof lurks within its headlines.
On that front, let’s turn to Figure 5. Here, each word cloud shows the top dozen (or so) words that appear in the published headlines of De Souza’s articles. If the Herald is trying to preserve De Souza’s message, it has an odd way of showing it. The Herald’s headline vocabulary seems palpably different than the source material from Postmedia.
Let’s put some numbers to this game of headline hackery. We’ll start with the ten most frequent words used in De Souza’s Postmedia headlines. Then we’ll track the frequency of these words as our headlines journey across the country.
Figure 6 shows the results, which remind me of a bad game of telephone. As De Souza’s headlines move from the Citizen to the Journal to the Herald, the original message gets progressively lost.
Content, creatively curated
So far, the Herald’s editorial churn looks a lot like censorship. But let’s not give up. Perhaps the key to the Herald’s lack of De Souza bias lies in the content of its published articles.
Speaking of content, there’s no substitute for actually reading the different versions of De Souza’s articles and observing how the text changes. If this close reading interests you, head over to ‘The De Souza Files’, where you can browse annotated versions of each published De Souza article.
Here, though, I’m going to skip the close reading and instead, count words. In what follows, I use word frequency to measure how the content of De Souza’s writing varies across papers. The patterns are … not subtle.
Communication clawbacks
Looking at our De Souza content, let’s start with words that describe communication. For some reason, the frequency of these words declines precipitously as we journey into the Canadian oilpatch. Figure 7 shows the pattern.
So why is the Herald cutting communication content? Here’s my guess. These communication words are associated with the nuts and bolts of good investigative journalism — the part where a journalist describes what people say. If you happen to gut this investigative reporting, a plausible side effect is that you gut the language of communication. In short, the pattern in Figure 7 smells of biased editing.
Glossing over government
Next, let’s look at vocabulary associated with government. Figure 8 shows how this vocabulary varies across the different versions of De Souza’s articles. After a modest surge in the Ottawa Citizen, government language collapses as we head into the oilpatch.
Given that the Citizen is published in the national capital, it’s understandable that it would have a governmental emphasis. But what’s with the Journal and the Herald? Why would they gut governmental language?
To understand the pattern, recall that our De Souza articles documented a series of scandals involving the Harper government — a regime that was staunchly pro oil. Now, if you happened to publish a pro-Harper, pro-oil newspaper, you might want to dampen news about government scandals. And if you carried out this dampening, you might end up cutting lots of government vocabulary, giving rise to a pattern much like Figure 7.
Hmm … there’s that censorship smell again.
Indigenous exclusion
Speaking of the aroma of censorship, let’s look at Figure 9. Here, I’ve plotted the frequency of words referring to indigenous peoples. For some reason, in the Herald versions of our De Souza articles, this indigenous vocabulary collapses.
To understand this pattern of indigenous extirpation, it helps to understand the relationship that indigenous peoples have with big oil.
Going back a century, indigenous peoples signed a series of treaties with the Canadian government — treaties that they believed would preserve indigenous land. But by ‘treaty’, the Canadian government mostly meant ‘piece of paper that we will honor until we discover natural resources on you land.’ When the resource boom started, the (hidden) understanding was that different pieces of paper took precedence — the legal documents that leased indigenous lands for corporate exploitation.
Fortunately, the Canadian government has since embarked on a mission of ‘truth and reconciliation’. So there’s no need to worry about the on-going exploitation of indigenous lands, or the censorship of indigenous issues.
What’s that you say? Something about the ‘oilsands’? Well that’s uncomfortable. Yes, there does seem to be a huge land rush to secure exploitation rights to the oilsands. And this massive resource trove does seem to cover a significant swath of treaty land, as Figure 10 illustrates. But I’m sure that oil companies have a harmonious relationship with indigenous groups, with no need for corporate censorship.
What’s that you say? Indigenous groups have been taking provincial governments to court, challenging oilsands lease agreements … and winning? Well, that does look bad. But I’m sure oil companies are playing fair.
True, this ‘fair play’ sometimes has the air of one-sidedness. For example, it seems that historically, when the Alberta government consulted indigenous groups about oilsands exploitation, it did so on a project-by-project basis. And it seems that this tunnel-vision had the effect of downplaying cumulative impacts, much to the chagrin of First Nations communities.
In this light, it does seem that a pro-oil, pro-corporate paper would have compelling reasons to censor indigenous issues from the reporting of an environmental journalist. So again, the Herald’s behavior has the scent of censorship.
Removing resistance
Continuing with our censorship sniff test, it seems that when the Herald published De Souza’s work, it rescinded words describing resistance. Figure 11 shows the pattern.
Now admittedly, this evidence doesn’t look good. At face value, it seems like the Herald is running a PR machine — a machine that guts language describing anti-oil resistance. Alternatively, the Herald was just chronically short on space, and these resistance words were by chance, the first to go.
Glorifying growth-speak
So far, we’ve quantified the Herald’s tendency to cut certain themes from De Souza’s writing. While the direct effect of this cutting is to de-emphasize what was gutted, the indirect effect is to emphasize what remains. And what remains, it turns out, is growth-speak.
This is my term for words like ‘income’, ‘jobs’, and ‘profit’ — vocabulary that gets wielded whenever someone wants to justify ecological harm in the name of economic growth. In De Souza’s reporting, growth-speak appears as he seeks comments from various government and industry representatives.
As we head into the oilpatch, hefty chunks of De Souza’s reporting get cut. But for some reason, the language of growth-speak avoids the knife. The result is the trend in Figure 12 — an magical rise of growth-speak on the road to Calgary. It’s as if the Herald edited De Souza’s reporting in a way that emphasized oil-business revenue. Now why would a pro-oil paper, located at the corporate heart of the oipatch, do a thing like that?
Oil-business consent
Looking at the weight of evidence, we get the clear impression that Mike De Souza’s environmental journalism didn’t get a fair shake. As his pieces made the journey into oil country, they were curated in a rather conspicuous way.
Numerous articles were cut. And the ones that did get published were pushed to the back of the paper. Headlines were changed. Text was gutted, de-emphasizing themes of communication, government, indigeneity, and resistance. Meanwhile, as if by magic, the language of growth-speak got emphasized.
In sum, the editorial evidence reeks of censorship.
I doubt that Herman and Chomsky would be surprised. In Manufacturing Consent, they note that since the corporate media is funded by ads, it is beholden to the interests of business.
That said, business capture doesn’t mean that all corporate journalism is heavily censored. Far from it. Topics that offend a niche set of businesses are fair game, so long as that niche is a marginal advertising client. But when that niche becomes the dominant player — like the oil business is in Calgary — the demands of censorship set in.
Thus we get the Calgary Herald’s blatant gutting of Mike De Souza’s environmental reporting. It’s disturbing behavior … if you think that newspapers have a duty to deliver inconvenient truths. But while this duty may be part of newspapers’ journalistic motto, it’s not part of their business model. So in a business sense, the Herald was just doing its job — making sure that journalism didn’t interfere with the more important business of selling oil.
Support this blog
Hi folks. I’m a crowdfunded scientist who shares all of his (painstaking) research for free. If you think my work has value, consider becoming a supporter.
Stay updated
Sign up to get email updates from this blog.
This work is licensed under a Creative Commons Attribution 4.0 License. You can use/share it anyway you want, provided you attribute it to me (Blair Fix) and link to Economics from the Top Down.
Sources and methods
The De Souza Files
For the interested reader, I’ve made a webpage (below) where you can browse the published versions of De Souza’s articles.
https://sciencedesk.economicsfromthetopdown.com/data/2023/desouza/home.html
Article position
Here’s how I quantify the page number of De Souza’s articles. In the simple case that the article is published in the A section, the page number is simply the section page number (i.e. A7 = page 7). This covers all but two cases — both of which are in the Herald.
The Herald saw fit to bury two De Souza articles in the D section. To quantify the position of these articles, I used ProQuest data to figure out how many pages were in the A, B and C sections of the relevant Herald issue. Then I summed this page count and added it to the section page number in the D section.
Map data
- Shape files for Canada (Figure 1) and Alberta (Figure 10) are from Statistics Canada and can be downloaded here.
- Shape files for the geological regions of Canada (Figure 1) are from Natural Resources Canada, and can be download here.
- Shape files for Alberta oil wells (Figure 1) are from the Alberta Energy Regulator, and are available here. I used depth and location data for ‘bottom holes’.
- Data for BC oil wells (Figure 1) is from the BC Energy Regulator (formerly the BC Oil and Gas Commission). A list of available data lives here. I used the following datasets:
- Drilling Data for All Wells in BC [BCOGC-41984]
- Directional Survey Data [BCOGC-2354]
- Data for Saskatchewan oil wells (Figure 1) is from the Government of Saskatchewan, and can be downloaded here.
- Data for Manitoba oil wells (Figure 1) is from the Government of Manitoba and is available here.
- For oil wells in the Northwest Territories (Figure 1), I couldn’t find data about actual wells. As a substitute, I used data for oil-and-gas rights (i.e. well leases). The data is available here.
Note: In Manitoba and the Northwest territories, I couldn’t find data for oil-well depth. So I assumed a generic depth of 1 km.
- Shape files for First Nations Treaties (Figure 10) are from Global Forest Watch, and can be download here.
- Shape files for oilsands leases are from the Alberta Energy Regulator, and can be downloaded here.
Notes
Further reading
Achbar, M., & Wintonick, P. (1992). Manufacturing consent: Noam Chomsky and the media. https://www.youtube.com/watch?v = Li2m3rvsO0I
Boychuk, R. (2023). Proximity to power: The oilpatch & Alberta’s major dailies. Capital as Power. https://capitalaspower.com/2023/01/proximity-to-power-the-oilpatch-albertas-major-dailies/
Herman, E. S., & Chomsky, N. (1988). Manufacturing consent: The political economy of the mass media. Pantheon Books.
Linnitt, C. (2014). Mike De Souza’s 20 most important articles for Postmedia. The Narwhal. https://thenarwhal.ca/mike-de-souza-s-20-most-important-articles-postmedia/
How about Andrew Nikiforuk? Does his writing enter Herald?