Can Wikipedia Stay Neutral in the 2024 Presidential Election?

Are Wikipedia editors politically biased toward the left? And does Wikipedia's very mission statement necessarily imply a partisan worldview?

An often overlooked indicator of Wikipedia’s influence is the extent to which complaints exist about its bias. Its prominence on the web and integration with social media, data platforms, and now generative AI has conferred upon it a rare degree of influence about what people think, believe, and even know exists.

These complaints might be easily shrugged off if not for Wikipedia’s own claim to neutrality. The second of Wikipedia’s five pillars is: “Wikipedia is written from a neutral point of view”. Like The New York Times (“All the news that’s fit to print”) or CNN (“The most trusted name in news”) it asks to be held to a higher standard. Therefore, it should be no surprise when someone tries to do just that.

While no one has recognized or criticized Wikipedia’s potential for bias as much as Wikipedians themselves, the loudest complaints tend to originate on the political right. With the U.S. presidential campaign now at full tilt, a trio of otherwise unrelated developments show how this plays out in real life: a debate about Kamala Harris’ record as Vice President, a study finding evidence of a left-liberal bias on Wikipedia, and a new manifesto about Wikipedia’s purpose.

Czar Wars

Less than a week after President Biden tapped out of the 2024 presidential election and Kamala Harris rotated in, a single Wikipedia editor made a handful of edits to a little-viewed and seldom-edited article, “List of U.S. executive branch czars”. The fallout from this addition is still unfolding:

The archaic Russian title of “czar” has long been used for executive branch appointees managing policy areas not overseen by existing federal agencies. Starting with “energy czar” during the U.S. oil crisis of the 1970s and reemerging as “drug czar” under Reagan, titles such as “AIDS czar”, “cybersecurity czar”, and even “democracy czar” (figure that one out) have proliferated, typically for hot-button issues where the president must be seen taking some kind of action.

In the spring of 2021, amid a surge of migrants at the U.S.–Mexico border, Biden tasked Harris with working with Central American countries to “address the root causes of migration”. While the White House never used the phrase “border czar”, a few prominent outlets like Axios did. Harris never seemed to give the issue her full attention, and worse, gave a bad interview about it, so once she became the new nominee Republicans made it a key line of attack against her candidacy.

Within a few hours of Harris’ addition to the “List of czars” page, her name was removed and added back several times by editors with little to no editing history, then conservative writers and influencers complained on X / Twitter, and a big debate kicked off on the talk page. Meanwhile, the page was frozen in the status quo ante, angering some who saw its omission as “partisan hackery”—especially as page views temporarily exploded by 6,700%.

So… was Harris the border czar? What is a border czar, anyway? There has never been any U.S. government office with the title of “czar”. It’s insider jargon, an invention of the media, never intended to be interrogated like this. Does it matter that the phrase was not widely used at the time? Were media sources wrong to use it then? Are recent corrections just the liberal media doing Harris a solid?

It turns out that either a) the answers are not so simple, or b) hairs can be endlessly split to the point of absurdity. As difficult as consensus can be to obtain when political advantage is at stake, Wikipedia’s neutrality policy, and especially its prohibition against “POV forks”, forces editors to decide if these terms are plain facts or political spin, a ridiculous argument or serious business.

The pro-mention editors have a point that some reliable sources used the phrase “border czar”, which fits the article’s criteria, however their intellectual curiosity ends there. Mention skeptics have a point that the term was never used by the administration, that media usage was inconsistent and infrequent, and her remit did not entail an active role in border security.

As of this writing, the matter has not been definitively settled, but it seems plausible a “disputed mention” will be included. The controversy over her title belongs somewhere on Wikipedia, right? As a matter of fact, it already is: it’s explained quite clearly in Wikipedia’s biographical entry about Kamala Harris.

First We Take Manhattan

Earlier this summer, a second cousin of this argument made the rounds when a conservative think tank, the Manhattan Institute, published a study by a freelance researcher answering in the affirmative the question it posed: “Is Wikipedia Politically Biased?”

By scraping Wikipedia for right-leaning and left-leaning topics and performing an AI-assisted sentiment analysis of words in close proximity to each, the author showed a “mild-to-moderate” tendency to associate right-leaning topics with negative sentiments and left-leaning topics with positive sentiments.

What are some examples of this bias? Unfortunately, you won’t find that in this purely quantitative survey. What’s more, the freely downloadable data set (credit to the author for making it available) consists of literally hundreds of thousands of isolated passages, separated from their original context, and is simply not a useful starting point for a qualitative review. What the results actually mean—assuming they’re even correct—is thus a matter of speculation.

The steel man version of this argument is probably that left-leaning editors are placing disproportionate emphasis on controversies involving conservatives. But an equally plausible reading is that Wikipedia is describing things in proportion, and conservatives tend to frame their issues in negative terms. If one considers the campaigns of the respective presidential candidates, is that so hard to believe?

Other studies focused on Wikipedia’s political tendencies have found more nuanced results. An almost identically named paper from 2012, “Is Wikipedia Biased?” in the American Economic Review, suggested that Wikipedia had an observable Democratic lean in its earlier days, but as it has expanded, its articles “appear to be centered close to a middle point on average.”

And later research by the same authors has also demonstrated that Wikipedia editors who focus on political topics tend to become less polarized over time—something that absolutely cannot be said for that other battleground in the war of ideas, social media. Given Wikipedia’s stated goal of neutrality, its consensus requirement, and its anyone-can-edit ethos, that isn’t so hard to believe, either.

Here again, one sees the tendency of conservative critics to halt their inquiry at a level of analysis that raises questions, but doesn’t wait around for the actual answers. They dig until they hit something solid, and announce they’ve hit treasure—without checking first to see if it’s actually a rock.

Right Wrongs, Wrong Reason?

“We are not here to right great wrongs” is a phrase familiar to Wikipedians, often coming up in heated disputes among editors. In concept it derives from the “What Wikipedia is not” policy, but specifically comes from an explanatory essay about “Tendentious editing” and is referred to in shorthand as “RGW”.

Like “Play it again, Sam” and “E.T. phone home”, this phrase never appears in its most memorable form in the original source. Rather, it says:

Wikipedia is a popular site, and its articles often appear high in search engine rankings. You might think that Wikipedia is a great place to set the record straight and right great wrongs, but that is absolutely not the case.

The point is that Wikipedia exists to accurately record what “has already been sorted in the real world”, and leave the “novel connections and statements” to reliable sources.

Earlier this month, a new and seemingly contradictory essay appeared: “We are absolutely here to right great wrongs”. It argues:

True neutrality is neither passive nor indifferent. It is an active commitment to providing an inclusive portrayal of the world, which involves addressing the wrongs that have shaped our history and continue to influence our present.

It is not exactly an argument against RGW, but it does relocate the emphasis from process to outcome. It considers Wikipedia’s oft-quoted mission statement of providing “free access to the sum of all human knowledge” and asks: to what end? In this way it suggests Wikipedia’s process-minded editors and administrators may not be pursuing their own dearly-held assumptions to their logical ends.

This essay does aim to provide answers, which focus on progressive concerns and contain strongly left-coded rhetoric. It argues the free spread of knowledge is meant to support “marginalized groups”, call out “historical underrepresentation”, and critique “economic inequality”. It reads like a Port Huron Statement for Wikipedia.

This interpretation of Wikipedia’s mission is not at all implausible.

In fact, Wikipedia’s mission is stated in terms of Isaiah Berlin’s concept of positive liberty, or the “freedom to”, as in the freedom to become self-actualized. This contrasts with the concept of negative liberty, or “freedom from”, as in freedom from coercion. The former tends to be more associated with a liberal-left perspective, and the latter with a conservative-right perspective.

But the essay itself has been met with skepticism, both on the essay’s talk page and in a popular Facebook group for Wikipedia topics. It’s worth noting that none of these editors object to the political viewpoints espoused in the essay, only how Wikipedia should consider them. They point out that RGW does not call on editors to overlook injustices, merely to “wait until it’s been reported by reliable sources”, and Wikipedia would be less effective if widely viewed as an ideological or partisan instrument.

Undoubtedly, some critics already do see it this way, and could make a case that the very existence of this essay proves there is a liberal bias on Wikipedia. On the other hand, a Wikipedia supporter could say that editors’ defense of RGW suggests the site’s left-leaning tendencies are tempered by other values.

The stakes of this year’s U.S. presidential election may be higher than any other in living memory, and Wikipedia’s role in shaping perceptions may be greater than any election since its founding. Editors, readers, researchers, and commentators will continue to scrutinize it for bias, not to mention accuracy and proportionality. All that’s certain is Wikipedia will take its role very seriously.

---