Bryan Caplan  

Boaz on Media Bias

The Government Mortgage Subsid... I Lived to See the Future...
You don't have to convince me that liberal media bias is real and large.  But Cato's David Boaz points out another amusing example:

[M]ainstream (liberal) media regularly put an ideological label on conservative and libertarian organizations and interviewees, but not on liberal and leftist groups.  In a report about states accepting stimulus funds, reporter Kathy Lohr quoted "Jon Shure of the Washington D.C.-based Center on Budget and Policy Priorities," "Maurice Emsellem with the National Employment Law Project," and "Tad DeHaven, a budget analyst with the fiscally conservative Cato Institute in Washington, D.C." (Thanks! And I'd say the label is correct, even if I might prefer libertarian.)

Those are all legitimate sources for the story. But only one of them gets an ideological label -- even though the other two groups are clearly on the left...

Back on March 23, I noted but did not blog about references on "Morning Edition" to "the libertarian Cato Institute," the "conservative American Enterprise Institute," and "the Brookings Institution." No label needed for Brookings, of course. Just folks there...


It's all too typical of the mainstream-liberal media: They put ideological warning labels on libertarians and conservatives, lest readers and listeners be unaware of the potential for bias, but very rarely label liberals and leftists...

Boaz quixotically concludes:

Journalists should be more even-handed: label all your sources ideologically, or none of them. It's stacking the deck to label those on the right but not those on the left.

But what's in it for them?

Comments and Sharing

COMMENTS (27 to date)
Daniel Kuehn writes:

Although to be fair, he's talking about NPR here. That's going to be considerably more liberal than what I think the average American would call "mainstream media".

I don't think you have to be careful calling it "bias" when Cato self-identifies as libertarian and a paper calls them libertarian, and when Brookings doesn't self-identify as anything and they don't get called anything.

Daniel Kuehn writes:

* I do think you have to be careful.

Hyena writes:

Cato probably needs an ideological label more than the other two.

I think people who listen to NPR generally know the ideological positions of the major think-tanks or it becomes apparent once their commentator starts.

Cato, though, needs to differentiated from the conservative talker.

Gabriel Weil writes:

Cato and AEI and self-conciously libertarian and conservative organizations, respectively. Brookings is to the left of center in American politics, but it was not created explicitly to advance liberal ideas or causes. It is thus a more contestable point whether Brookings (for example) is accurately characterized as a liberal think tank in a way that it is not for Cato as a libertarian one. If the Center for American Progress were not given an ideological label when CATO or AEI or Heritage was, that would be a clearer sign of some sort of bias.

Dan Weber writes:

Someone pointed out how Dan Rather would do this on election nights. A conservative politician would win in Iowa, and then some guy would win in Nebraska.

Great googly-moggly!

William Barghest writes:

News isn't about current events.

agnostic writes:

In the NYT, there's a roughly 2:1 ratio of articles with "conservatives" to "liberals."

Doesn't look like it varies by election year or not. In 2008, 393 with "liberals" and 822 with "conservatives." In 2009, 474 with "liberals" and 876 with "conservatives."

Conservatives are a swarming horde, but liberals -- yes, OK, they exist, but they're just one of those sideshow fringe groups. Nothing to see there folks, just move along.

Daniel Kuehn writes:

agnostic -
But what "liberal" groups are there with the prominence of "conservative" groups? What do you expect a search count to really tell you?

First, as I said earlier and others have mentioned - if you self-identify as "conservative", as virtually all the major conservative think tanks do, then of course a paper is going to call you "conservative". If you don't self-identify as anything they're much less likely to.

But even then, the article lines NELP and CBPP up against Cato and AEI. I could believe that NELP is as liberal as Cato and AEI are libertarian and conservative respectively. CBPP though? Maybe, but that's getting a little murkier. Brookings? No way. No where near as liberal as AEI is conservative.

What are you looking for? A fifty-fifty split? What on Earth would lead you to expect that a fifty-fifty split is "fair"?

This is the same kind of logic that gets people riled up when it comes to climate change - if its not fifty-fifty then its not "fair".

Who says that fifty-fifty makes sense as what's "fair" in the first place?

Andy Hallman writes:

I think Daniel Kuehn is right, and I think agnostic's statistics may simply reveal that more people self-identify as conservative than liberal.

I work as a journalist in a small town in Iowa, and I count myself a member of the media. When I write a story about a candidate for public office, I only use words like "liberal" and "conservative" if the interviewee wants to be identified that way. So far, I've never met a "liberal" but I've met plenty of "conservatives."

It really amuses me that people think this liberal/conservative divide is the result of the media's liberal bias. The media have biases, but this is not one of them.

Ted writes:

I don't really care about liberal media bias or whatever. I actually think the complaints are grossly exaggerated and mostly a self-serving complaint that some article doesn't say what you want it to say, and thus you are going to call it "biased" because it doesn't match your own bias. But even given the existence of media bias, which of course exists, the liberal bias of the NYT is balanced out by the conservative bias of the WSJ. And the liberal bias of MSNBC is balanced out by the conservative bias of Fox News (well actually outweighed since Fox is a joke at this point). If you don't read other sources besides ideologically comforting ones, than you probably like the bias anyway. All around everyone wins. Bias is washed out for those that care. And those that want the bias are comforted. Sounds like Pareto optimality to me!

Also, I think what is going on is that conservative and libertarian institutions tend to identify and define themselves by their ideology - which is why they get the ideological label. Brookings Institution is obviously moderate center-left, but it doesn't define themselves by that ideological label whereas Cato does define themselves by the libertarian label. I would assume the same goes for this "National Employment Law Project." They probably don't define themselves by an ideological movement, though they probably are ideological.

Also, do you really need to even identify what political spectrum an organization named "National Employment Law Project" identifies with?

ziel writes:

AEI does NOT self-identify as conservative, and Cato self-identifies as libertarian, not conservative (and for good reason - it's not conservative). So we're back to liberal bias.

AEI notes that "the Institute's community of scholars is committed to expanding liberty, increasing individual opportunity, and strengthening free enterprise."

Brookings lists its goals as "Strengthen American democracy; Foster the economic and social welfare, security and opportunity of all Americans; and
Secure a more open, safe, prosperous and cooperative international system.

The focus of each is rather clear from these missions listed on their websites: AEI is conservative, Brookings liberal. Both should be identified as such when being quoted, or neither should be.

Daniel Kuehn writes:

ziel -
"Strengthen American democracy; Foster the economic and social welfare, security and opportunity of all Americans; and
Secure a more open, safe, prosperous and cooperative international system"

Is a liberal mission statement?!?!?!?

AEI gets more explicit in many of its events, speakers, and publications.

Nevertheless, I was very careful not to say that AEI self-identified as conservative and Gabriel only said they were "self-consciously" conservative, which I would agree with.

If you think Brookings is as liberal as AEI is conservative, I would suggest that you may not be very familiar with these two organizations.

Tom Cairns writes:

Everyone seems to think the media is biased against them, so I suspect some kind of confirmation bias is happening. What kind of statistical test would you recommend to determine actual media bias, independently of reader preconceptions?

ziel writes:

Daniel - "expanding liberty, increasing individual opportunity, and strengthening free enterprise"

is a conservative mission statement?!?!?!?

Are you really incapable of deciphering what these mean? I'll help:

"Strengthen American Democracy" - American democracy needs strengthening? Obviously this is meant to increase the relative power of the so-called "disenfranchised" - a typical liberal trope.

The second point uses the very term "social welfare", a term that only has meaning for liberals - it's meaningless to anyone who does not think that individuals are responsible for their own "social welfare."

The third item - "a more...cooperative international system." That's called Internationalism - and that falls on the liberal side of the ledger.

I'm not saying they're a leftist organization, but enough on the liberal side that to pretend otherwise is misleading. The other orgs mentioned in the article are clearly left wing - there's no excuse there.

I'd say Brookings is probably a more reliable institution than AEI - I think their "scholars" are generally more scholarly (or rather Brookings has less out-and-out hacks than AEI has).

Daniel Kuehn writes:

social welfare only has meaning for liberals?

cooperative international system is for liberals?

I'm really trying to treat you even-handedly and take you at face value, but you're making it tough.

agnostic writes:

"Liberals" and "conservatives" does not refer to organizations, so you're off in talking about which ones self-identify and which don't. In plain English, "liberals" refers to a group of people, and so does "conservatives."

The NYT makes over mention of the conservative group of people much more than they do the liberal group of people, and by a ratio of 2:1. I didn't discover that the ratio was within some small range around 1:1 -- 2:1 obviously shows a bias.

I'd expect the opposite bias for Fox News, and so would anyone else. Taking all major media outlets into account, what do you think the distribution of this ratio looks like? (Rhetorical question.)

agnostic writes:

* overt mention

FredB writes:

Liberals don't see liberal bias in the way fish don't see water.

I thought they were giving libertarians/conservatives credit...

matt writes:

I think FredB is absolutely correct. I take this post as an observation and not a critique. Of course it is only acceptable if our (Bryan's) side is more correct methodologically. If an irrational voter/public demands a media, it makes sense that the media that they demand is irrational.

If however, we are the ones that are irrational, the tables would seemingly be turned. A liberal might retort FredB by saying, Libertarians don't fear bias the way that lemers don't fear gravity.

Of course, the great thing about being libertarian is that the logical problem of, "no one can be unbiased, therefore I cannot be unbiased," leads to only one philisophical conclusion...

Scrutineer writes:

Ted - ...the liberal bias of the NYT is balanced out by the conservative bias of the WSJ.

Maybe not (unless you're referring strictly to the op-ed pages).

Brian Clendinen writes:

Bias is usually displayed in what is not said verses is said. So the two examples are not proof by themselves but when you combine them with 100's of other example their is clear proof. I am talking about the same news organization. I have listened to NPR news long enough to know they are socialist for the most part and I have listened almost an hour a day for years. For example never once with the health care debate did they ever get someone from even the middle to discuss their opposition. I must of listened to countless hours of reports and commentary on the issue. Not even regular contributors from Fortune or Forbes said anything about it before it passed. The news casters would only summarize the opposition argument or give a 10 second sound bit. Only after did it pass did I here someone from the opposition explain why it was bad and that was by a ranking congressman. I mean a lot of it was in somewhat neutral terms, but there was plenty of pro sections but never a real con section. However, that was until right before the vote. Then pretty much all you heard was people explaining why this organization was wrong to oppose and betrayed them, how great this was going to be, or how people really wanted this. My tax dollars at work, not a lot of difference from communist funded “news” I have listened to and read some. I guess I should not be surprised considering the 2008 democrat convention felt exactly like a communist party political event, and for the most part was covered like one to.

twv writes:

Newspaper journalists, at least, tend to do what Boaz describes around where I live, though often it's easiest to see in frankly political articles.

I live in a heavily Democratic-Party/insider-controlled voting district, and the newspapers routinely give the incumbents the last word, even in articles about challengers. When a challenger brings up a fact that makes an incumbent look bad, the newspaper, witlessly, calls it an "allegation" or "alleged fact," even though the journalist could have fact checked the charge.

Instead, newspaper journalists rely on the established "neutrality" of using "alleged" and similar words. This brings out the patina of journalistic integrity, but better demonstrates an unwillingness to do any research whatsoever . . . at least that would show an incumbent to be what is often quite accurately alleged: a liar, a backroom deal-maker, a two-face, a what-have-you.

This practice is very like the careful stacking of the identity deck that Boaz writes about in relations to "libertarian" and "conservative." It is how those in power, or close to power, stay in power: By reinforcing the terms of the outsiders' marginality.

This being said, libertarian writers too often self-identify themselves as libertarian. Why? Because they are trying too hard to push their ideology. This is how the small fish reinforce the prejudices and biases of the big fish.

This also occurs regarding academic schools, as in "Austrian Economics." The mainstream used to dismiss an economist as "Austrian"; and the so-called "Austrians" routinely self-identify as members of the "Austrian School."

This sort of compliance with marginalization is the linguistic analog to Etienne de la Boetie's "voluntary servitude."

And if one always self-identifies, maybe one has little standing to complain about the mainstream marginalizing identification, no matter how ill-mannered it may be on a more rarefied plane.

Honesty may demand that one self-identify at some point in a discussion. It does not require that one lead with the label, or use it early on in a self-defense.

Ted writes:


It's hard to take that study seriously. Firstly, their proxy is incredibly crude. And it would force me to accept that the Wall Street Journal is "more liberal" than the NYT, and given how poorly the study is done I'm not ready to assume my eyes are lying to me.

For example, the Center for Responsive Politics is coded as being "liberal" by their proxy. There is nothing liberal or conservative about the institution, it just reports political contributions. Or the NRA is coded as just ever-so slightly conservative, with the ACLU being more conservative. Yeah, I'm sure the NRA is less conservative than the ACLU. I'm just sure of it.

Sorry, that study is a poorly done. It's not surprising you'd get such absurdity using the dumb proxy they use.

Marceli writes:

Wikipedia does the same thing.

Cato institute article: First sentence starts with, "The Cato Institute is a libertarian think tank". Large panel on the right side that says, "Part of a series on Libertarianism".
Center on Budget and Policy Priorities: No mention of political stance or ideology.
National Employment Law Project: No page on Wikipedia
Bookings institution: Mentions a possible liberal political stance towards the bottom of the article.

You guys should edit the Wikipedia pages.

Andy Hallman writes:
Instead, newspaper journalists rely on the established "neutrality" of using "alleged" and similar words. This brings out the patina of journalistic integrity, but better demonstrates an unwillingness to do any research whatsoever . . . at least that would show an incumbent to be what is often quite accurately alleged: a liar, a backroom deal-maker, a two-face, a what-have-you.

I am a newspaper reporter, and this is absolutely correct, I'm afraid. I try not to use "he said" when the person is making a serious charge or an easily verifiable charge against someone else. However, it is not possible to cross-reference every claim someone makes in a story. Unless you want the news delivered a week late.

Marceli suggests:

You guys should edit the Wikipedia pages.
Actually, I wrote most of Wikipedia's page for Econlib (with some subsequent, much-appreciated rewording help from a Wikipedia expert more experienced than I with Wikipedia's official tone preferences). I worked hard to keep any remotely politically-oriented labels, e.g., "libertarian," out of it. I try to check back on the page occasionally, but it was so hard to keep outsiders from inserting unwanted material that I decided to not create a Wikipedia page for EconLog--which would be more of a hot button than Econlib's general page. A student of Russ Roberts's created a Wikipedia page for EconTalk, and he has wonderfully and responsibly managed to keep any politically-charged labels off that page so far.

Liberty Fund does have a Wikipedia page, which I see does have a categorizing label--"libertarian"--on it. Plus, looking quickly at the page history, a lot of bot activity, but at least not removing an effort I made to actually link to Liberty Fund's own official website. I'll let Liberty Fund know in case they want to work on it. Of course, a problem with Wikipedia is that anyone can modify the text at any time. Convincing Wikipedia that you are the authoritative source or owner can be annoyingly difficult.

Comments for this entry have been closed
Return to top