David R. Henderson  

The Problem with Wikipedia

Steven Chu vs. Bernie Madoff... Summers Joins FIRE...

When I first heard about Wikipedia, I thought, "this can't work." My reason: there was no assurance that letting huge numbers of people fill in entries and update things would lead to correct information. That said, it works much better than I had expected.

But in my only two cases where I have paid close attention to how information gets screened, Wikipedia has worked badly. The first has to do with the entry on me. See it here. I can't remember the details because I gave up, but I do remember trying to get my birth date and year entered but not being able to do so. Wikipedia did not trust me to know my own date of birth. That's a minor problem, but still an annoying one.

The second has to do with something that happened in the last 36 hours. I posted on this blog a major mistake on Wikipedia having to do with why anti-capitalist author Thomas Carlyle called economics "the dismal science." It was because the free-market economists at the time, who dominated economics, strongly opposed slavery.

I posted it because I thought, it turns out correctly, that some reader of the post would go on to the Wikipedia site and make the change. That happened quickly. That's the good news. The bad news? The mistake is back. I'm guessing that if someone went on again and corrected it, the mistake would be back within hours. That's the problem. I don't know, offhand, a good fix. I do know that San Jose State University economics professor Jeff Hummel tells his students that if they are not already seasoned researchers (and, really, how many of them are?), they should not rely on Wikipedia.

Comments and Sharing

CATEGORIES: Economic Education

COMMENTS (31 to date)
BZ writes:

Give it time. You wouldn't believe how much arguing it took to get even a mention of the Father of Classical Liberalism, John Locke, on that article on classical liberalism. Some apparently wanted to limit the term ONLY to the group of economic "survival of the fittest" types at the end of the 19th century in order to discredit the idea. This is a common problem -- the left lies. A lot.

Anyway, the mistake you point out, while important, is smaller in comparison. You've provided us with an all important reference for citation; I had to hunt down my own to defend the notion of an 18th century "classical liberalism", but like that fight -- it will just take time and perserverance.

Jason writes:

That's a problem with Wikipedia. The true problem with Wikipedia is something completely different.

BZ writes:

Also, scroll to the bottom of this page to read the discussion of this question, which is far more substantive than the Locke debate I mentioned:


David R. Henderson writes:

Thanks. That discussion is interesting. It reminds me of how I almost pulled the plug on The Fortune Encyclopedia of Economics (now The Concise Encyclopedia of Economics) in spring 1991 when my point of contact at Fortune told me she wanted the Encyclopedia to represent what the average person thinks. I’ll tell that story in a post soon.

adam writes:

I don't think n=2 observations is quite sufficient.

More systematic studies of the quality of certain parts of Wikipedia (such as science or math where the quality is easier to measure objectively) have been conducted, usually finding that its quality is similar to, if slightly below, the average quality of an average Britannica article.

I think the real bottom line here is that you should never rely on any one source.

Jon Murphy writes:

I've always liked Wikipedia. It's an easy way to check something quick.

Wikipedia can be used in an academic setting, with caution. Wikipedia uses citations, which you can click on to find where the data is coming from. This is how I've always used it. I don't trust what is written on the site, but rather follow the citations to somewhere else.

Austinecon writes:

@BZ/David R. Henderson,

Aww shucks you guys!

Nick F writes:

Regarding the first point, about you changing your birthdate, this is a necessary heuristic restriction: one cannot edit their own page. It prevents many more distortions than it creates.

Regarding the second point: one of the wonderful things about Wikipedia is that you can look under the hood to edit history, citations, and discussion of topics. The site has rules that can create some biases (like nPOV and self-editing restrictions), but as a person reading Wikipedia, if you understand these biases, you can learn a lot, if not more still, from Wikipedia than from a traditional source. This is especially true since contentious points play themselves out in the talk pages.

For this reason, students should probably, on some level, be instructed on how to use Wikipedia by looking under the hood. I would assume this is Hummel's concern, but on a certain level, students should take any source (digital and established print) not directly provided by a professor with a strong grain of salt, for the same reasons.

Additionally, Wikipedia may be well-served, when there is a debate on interpretation of a point of fact, to link within the article to the underlying discussion, since there is nearly always a critical mass of people debating these points. Indeed, on the Classic Liberalism page, now, there is a lengthy discussion about use/interpretation of the term "dismal science" that gets at the core of the debate. All one has to do is click "Talk" at the top of the page. Since this term has been commonly misinterpreted, it should come as little surprise that there is debate here. I don't think Wikipedia is any less accurate than traditional sources on this topic because of that. This is likely the reason why

On a personal level, I often use Wikipedia, but mostly as a jumping-off point, since it's a good compedium of major citations on whatever topic I'm researching.

Ted Levy writes:

Well, David, as an economist once told me, "Nothing's perfect."

Thinking in terms of alternatives and comparative institutions, however, it would seem you're being overly harsh regarding Wikipedia. Ignore its libertarian founder, Jimmy Wales, and ignore Wales' Hayekian insight in developing the encyclopedia. Just look at the results:

Wikipedia is an online encyclopedia. Let's compare the problem you cite with old-fashioned paper encyclopedias, like Britannia. If you found an error there, it would possibly be corrected in the next edition--that is, in a few years. And then the correction would be available to all those who bought, often at considerable expense, the new edition, that is, a small fraction of the readership.

Here, the revisions are immediate, and free, and available to all.

Yet the paper version's essays were usually written by experts in their fields. Whereas on Wiki we don't know the bone fides of those who write, partly due to the anonymity of the forum, partly because there is in fact no ONE writer.

Yet even experts make errors. I understand you've caught a few economists guilty of such things yourself. :-) And it is my understanding that one reason for the long delays between editions and updates in books like Britannia is that it was hard to find experts willing to work for what was paid and turn out copy in a timely fashion.

Then there's the pluralism argument, brought up in Bryan's concurrent discussion of extremer extremists. No one likes errors, but even those of us sure of our facts appreciate we are sometimes in error...no doubt there are some out among the Internet tubes highly frustrated that you are "incorrectly" attributing Carlyle's assessment of economics dismalism to slavery. Errors are bad, but easily correctable errors are better than difficult-and-timeconsuming-to-correct errors.

I'm pretty sure I read a study a few years back that compared Wikipedia to traditional encyclopedias. Experts in their respective fields read Wiki articles and Britannia articles and compared them. Wiki, if I recall, came out ahead, at least on a number of parameters. I could Google that for you, though I understand many academics frown on such actions... :-)

Norman Pfyster writes:

I'm pretty sure from reading Carlyle's essay that he didn't call social science the "dismal science" because it was opposed to slavery. Even in the paragraph cited, he describes social science as one in "which finds the secret of this universe in 'supply and demand,' and reduces the duty of human governors to that of letting men alone..." In context he is describing a policy which leads to starvation, and instead favors mercantilism and a managed economy (for some reason, he thinks sugar cane is more noble than pumpkins and the notion of comparative advantage seems lost on him). His economics might be stupid, but that's another issue entirely. To say that he called economics the dismal science because it supported emancipation strikes me as being fairly tendentious and not at all a justification for re-writing the Wikipedia entry.

David R. Henderson writes:

I don't think n=2 observations is quite sufficient.
True. My point is that it’s the only two where I’ve paid close attention and it didn’t work. That said, I agree with your bottom line: never rely on any one source.
I see, by following the link that BZ cited, that you’re doing yeoman’s work on this. Thanks.

Mperry writes:

Looks like they've just removed that section entirely.

Thomas Hobbes writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Adam writes:

Good point about the durability of errors in Wiki. Funny thing though, I'm sure I tried to get that same "error" corrected several years ago. I thought it was pretty cool to go in and add my correction to the site. I never checked back, however, to see if my entry stayed corrected. Hence, you're post yesterday was pretty surprising to me.

Finch writes:

> Wikipedia uses citations

I wish Wikipedia kept timestamped PDF copies of all citations rather than relying on links that degrade with time. It's only sometimes possible to follow the links and get the source. And sometimes even when the link works, the source material has changed, too.

Seth writes:

If you enter "dismal science" into Google, the first result is a Wikipedia entry on the phrase itself, which provides the correct context for the phrase and even includes a link to the original Carlyle essay.


Carl writes:

You should make a blog post with the correct date, and bring it to the attention of editors who can cite it.

M Stallman writes:

reminds me of another xkcd cartoon, where made-up wikipedia entries become "fact" if sourced for a magazine before it's corrected


David C writes:

Here is the study everybody keeps referring to:

3.86 errors per article for Wikipedia vs. 2.92 errors for Britannica
Both had 4 major errors in all of the articles examined.

That was in 2005. I'm guessing the major Wikipedia articles have since become more accurate.

I also thought it was interesting comparing the Wikipedia article on Britannica with the Britannica article on Wikipedia:
"It is regarded as one of the most scholarly of English language encyclopaedias."

"Although some highly publicized problems have called attention to Wikipedia’s editorial process, they have done little to dampen public use of the resource, which is one of the most-visited sites on the Internet."

Jealous much?

I've just had reason to compare the Wiki article on Glass-Steagall with a treatment of the same act at the Columbia Journalism Review. The CJR suffers badly in that comparison.

Wiki on GS, though repetitive and overly long, is actually quite informative. Its several footnotes are useful too.

Mark Little writes:

My user experience with Wikipedia has been very positive. For controversial, political, ideologically tinged questions, it is surely not reliable. But for scientific, mathematical, or generally non-controversial factual matters it is often excellent. (Not comprehensive and not always right, but what do you want from an encyclopedia?)

For most of the queries I've used it for, Wikipedia is sometimes useful and sometimes less so, but it is almost always more valuable than the other links that come up on a Google or Bing search. For a quick factual summary it is hard to beat.

I don't even fault Wikipedia much for the bias that does appear. They seem to try hard, we all have our biases, and I'm not sure I could do any better writing on areas I know, if judged by someone reading with a different perspective than mine.

And what is the alternative? Would you trust, e.g., Britannica more, on any subject with ideological salience? True unbiasedness and complete accuracy is not obtainable. But the crowd sourcing model of Wikipedia seems to do better than anything previous that had anywhere near this ambition of scope.

db writes:

Wikipedia's standard for "truth" is verifiability. If you can't provide a third party citation for an edit it is likely to get rolled back. Wikipedia does not care if you have first hand knowledge that something in an article is wrong, they are adamant the cite remain a tertiary, rather than primary source.

Obviously that huristic can lead to bad information and biases the site against new scholarship, but it seems to do a pretty good job of agregating consensus, middle-of-the road, positions. Just be especially wary, as you should with any tertiary source, of information on topics undergoing a paradigm shift or more generally those that lack a solid consensus.

Jeremy, Alabama writes:

Mark Little has it right. The more objective the science, the better the Wikipedia article is likely to be. The gold standard for me is the article on the Pioneer velocity anomaly:


which discusses all the proposed explanations, and experiments that might prove or disprove each of them. This is how science should be, with a vigorously interested and non-judgmental debate.

Compare this with Wiki's various global warming articles, where climate non-catastrophists are marginalized as weird neanderthals.

Nicholas writes:

Wikipedia suffers on contested topics where the moderators lock pages, then select users who conform to their biases on the subject to write the articles. Like most conflicts, this stems from scarcity: there is only one article per topic and more than one idea about what the article should say. But it's the internet, so the scarcity is artificially imposed. The solution is to run concurrent articles on these topics, and use heuristics to server the version of the article you are most likely to be looking for.

David R. Henderson writes:

Thanks to all. I agree with Mark Little too. Just so you know, I was pointing out a pitfall of Wikipedia. I’m not one who commits the Nirvana Fallacy, to use the term of my mentor, Harold Demsetz. The world is a much better place because of Wikipedia. My goal with this post was to make it slightly, slightly better.

Steve Cronk writes:

A problem with Wikipedia is similar to a problem in the market. Individual attempts to solve it may fail, but it's constantly moving in the direction of improvement.

John David Galt writes:

Steve Cronk has part of the puzzle. Wikipedia is a bit like a whole marketplace. An individual participant in the market can develop knowledge, but the market as a whole never can.

I hope that Wikipedia's IP protection is flexible enough that someone (preferably many people, on all sides of issues) can use it as a starting point to create a good, stable, reference site that is really useful (at least to people who share that person's views).

But Wikipedia as it is now shows the ill effects of trying to be "democratic" and offend no one. I will never trust a Wikipedia article to be complete, accurate, or even neutral. At most, the site is useful as a starting point for research on topics that are new to you.

Larry writes:

Wikipedia policy is that it's own articles do not qualify as "reliable sources" for other articles.

As a result, especially for controversial topics, I use it as a good place to look for sources. The list of references are in effect a manually curated version of a search engine result. That's the real gold in wp.

I'm often amazed how bitterly some editors fight over points/ articles that are highly non- controversial to the rest of us. Article quality does seem to improve as time goes by...

Kevin Driscoll writes:

I love Wikipedia and use it all the time, but what many of you have pointed out is true; it suffers from all the same defects as a traditional tertiary source. The problem is that in certain subjects there just isn't a consensus and so a Wikipedia article will almost always be wrong. Anyone involved enough to correctly state their position is also so involved that they probably violate the nPOV standards. Wikipedia doesn't let one part of an article be biased and another part be biased the other way and have the two cancel out. The whole thing has to be nPOV and sometimes that isn't possible.

For most high-level science and math Wikipedia is a great resource. For example, the wikipedia article on perturbation theory in Quantum Mechanics carries the calculation WELL beyond what you will find in even graduate-level textbooks.

Kevin Driscoll writes:

I love Wikipedia and use it all the time, but what many of you have pointed out is true; it suffers from all the same defects as a traditional tertiary source. The problem is that in certain subjects there just isn't a consensus and so a Wikipedia article will almost always be wrong. Anyone involved enough to correctly state their position is also so involved that they probably violate the nPOV standards. Wikipedia doesn't let one part of an article be biased and another part be biased the other way and have the two cancel out. The whole thing has to be nPOV and sometimes that isn't possible.

For most high-level science and math Wikipedia is a great resource. For example, the wikipedia article on perturbation theory in Quantum Mechanics carries the calculation WELL beyond what you will find in even graduate-level textbooks.

Dave B writes:

I was editing to add Mises with Hayek and Friedman to classical liberalism over a year ago. I provided three books by credible authors in the "Discussion" page that explained how Mises was among the heavyweights and one of the very few carrying the torch of liberalism's revival in the early 20th century. He even wrote his own book on classical liberalism! They kept undoing it in Wiki and I finally gave up our of frustration.

I was pleasantly surprised to find him now included. I don't know who or how, but he is now accurately listed. That may give us a sign of hope about the long term. The problem is the short-term battle.

Comments for this entry have been closed
Return to top