Terrible, terrible example, because your älso accurates are not at all accurate for many environmentalists.
Daniel
Exactly. That’s my point. If you are an environmentalist you have a particular ideology that repels at the alternate descriptions I wrote. However, an environmentalist sees no problem with the “preserve the refuge” statement. Why is that? Because the “preserve the refuge” statement reinforces the bias of the environmentalist.
Someone like me, however, doesn’t view this as an issue of “preserving the refuge.” I don’t think the refuge will be harmed by energy exploration. Therefore I can see the bias of the reporters comments. An environmentalist would not because he agrees with that bias.
Regardless of how you see it, it is correct to say that the environmentalist is trying to preserve the refuge. If the environmentalist believes that the refuge will be best preserved by opening it to unfettered development by Wal-Mart and McDonald’s, it would still be accurate to say that the environmentalist is trying to preserve the refuge.
However, it would be inaccurate to say that the environmentalist is opposed to energy exploration unless the environmentalist had made a statement to that effect.
Maybe your example’s not so bad after all, inasmuch as it demonstrates how a media source’s failure to use ideologically-loaded expressions ends up being interpreted as biased.
Daniel
True, it is accurate to say that. However, it also implies that those who support drilling in ANWR are opposed to preserving it. I do not think that is the case.
No, by opposing drilling in ANWR he is opposing energy exploration. He is also opposing economic development.
No, the reporter used an ideologically-loaded expression. If the reporter wanted to be neutral, he would have said given the guy’s name and then said, “who is opposing ANWR drilling” or something to that effect. It’s a straightforward description that has no ideological connotation. Instead, the reporter chose to use a description that portrays him as someone who is “preserving” the refuge, which is a very noble thing. If he were biased in the opposite way, he could have easily said that this person was “opposing economic development,” which is a very negative thing.
Is it all right with you if I knock your house down to look for oil under it? If not, is it misleading for me to characterize you as opposed to energy exploration and economic development?
Of course not: all I may fairly do is to characterize you as opposed to a specific instance of energy exploration and economic development (and for very good reasons).
Similarly, it would be misleading to characterize Joseph Lieberman as opposed to energy exploration and economic development: his opposition in one specific instance does not translate into a general opposition.
Again, you can read bias in the entrails all you want to. That’s why I think that we need to have a very high standard for finding media bias instead of just saying, “This particular story could have been reported with these slightly different words, and then I would have personally found it less biased, so therefore the media are biased.”
The study in the OP at least sets up a rigorous standard for detecting media bias. I’m not sure at all that I agree with the standard they suply, but it’s certainly better than an anecdotal approach that reveals more about individual bias than about the media’s bias.
Daniel
Too many negatives: the last line should, of course, say, “Of course it’s misleading:” Sorry about the mixup.
Daniel
Yeah but in order to avoid this sort of “bias” a news organization would need to explain how energy exploration works and how damage is mitigated or entirely avoided in each reference to the fight over such exploration in the refuge. That is an unrealistic scenario.
And yes the media is biased (Fox News?) and so is everyone. An interesting question is why the corporate owners of “the media” allow their employees to be biased against their interests in many cases. And why a group such as news people are biased in a certain direction considering their general educational level and breadth of experience. I think news people have an educational level higher than the average and it’s their business to observe closely the events of the world. If the conservative view is better and more useful than others I would think that they would be aware of this and their reporting would reflect it.
It’s certainly a possibility; the study attempted to calculate the political position of the average U.S. voter, but not that of the average U.S. media consumer (not sure how that would shift things).
More interestingly, you could adjust for disposable income or some such value that represents the viewers the media are trying to attract to pull in advertisers. It seems like I’ve read there’s a left-right divide there, but the only link I found in a quick search was this. And there was this maybe related link from a GD thread last year that seems to show blue states ranking high in the “have” categories. Not that this would indicate a lack of bias, just perhaps show it’s not ideological.
Precisely. People who want to drill in ANWR are “pro-drilling” or “pro-business” or “pro-energy-exploration” or “pro-my-stock-in-oil-companies-making-me-rich”, or something like that. That is what motivates them. Because that motivates them, they are willing to sacrifice some natural beauty and environment, which doesn’t mean that they don’t care about it, just that they value it less.
Similarly, people who don’t want drilling in ANWR are “pro-environment” or “pro-preservation” or something like that, because that is what motivates them. They are willing to sacrifice some potential US economic development, which doesn’t mean that they don’t care about it, just that they value it less.
Someone who was truly anti-energy or anti-economic-development would be some kind of weird luddite or anarchist or Marxist or something.
As for the study as a whole, I have to say that it strikes me as just silly. What on earth does sheer-number-of-citations-from-think-tanks necesarily have to do with bias? There’s a false assumption of equivalency built into that claim, which is that there are similar numbers of think tanks on both sides, and they’re equivalently reputable and quotable.
Suppose that there’s a very big and reputable one-side-leaning think tank which, while having somewhat of a bias in what it chooses to investigate, is also known to produce very rigorous and fair statistics and positions. Then presumably it would be used as a citation for lots of stories, without necessarily implying that the particular stories had any bias. Alternatively, citations from think-tanks might mean that the story is LESS meaningful, if a paper writes paragraphs of anecdotes and personal heartfelt stories about some social issues, and then throws in some quotes from a think tank on the other side to give an appearance of balance, etc.
Also, what does “fair” mean, anyhow? What would a completely fair and objective newspaper account written in 1855 have said about slavery? Take a contentious issue like “Did Bush lead the country into war on false pretenses?”. Should a fair and unbiased story about that issue give 50% weight to both sides of the story, or should it tell the truth, whatever that happens to be? (Whatever the truth is, it’s unlikely to coincidentally be precisely the average of the two partisan viewpoints…)
There’s also a HUGE HUGE HUGE point that I think needs to be addressed, which is the difference between bias on social issues and political bias. You might (might) be able to convince me that most reporters subconsciously assume that their positions on things like abortion are the “norm”, and tend to present stories on those issues with that bias built in. But that in no way implies that these very same reporters are (for instance) going to treat Kerry more favorably than Bush.
If the article is accurate, it sounds like the study’s conclusions rest on some pretty shaky assumptions. For example:
From the linked article:
(Bolding mine.) They seem to be assuming that the voting of members of Congress perfectly represents the wishes of their constituents (on average). This ignores the reality that Congressmen very often vote according to the wishes of the powerful interests and corporations that pay for their election campaigns (or, as in the case of Duke Cunningham, simply bribe them). Sometimes industry lobbyists literally write pieces of legislation. It’s got nothing to do with the average political character of the constituents. (Ideally, every member of Congress would vote based on his own conscience and best judgement, after careful consideration.)
I think it’s a very interesting study, and is certainly measuring something. I’m unconvinced at this point that what it is measuring is what people normally call “bias in the media”.
For the record: I’ve no doubt at all that bias exists in various media. However, I’m skeptical that the media in general share a significant bias in a particular direction (other than the bias to keep or attract readers / viewers / listeners). I think the bias exists mostly in subtle word choice and turns of phrase and is probably unquantifiable in any meaningful way.
Most of the comments I wanted to make have been made by others. There’s one that hasn’t been mentioned yet. The linked article states:
One of the many interesting things about the 2000 presidential election was the candidacy of Ralph Nader. (As The Daily Show’s “America: The Book” puts it: “Of the three candidates in the 2000 election, Ralph Nader was second only to Al Gore in costing Al Gore the presidency.”) If we assume that, in general, the supporters of Nader were more left-wingish than the supporters of Gore, then by excluding Nader supporters the studies authors have likely caused their research assistants to skew conservative.
Now, let me be clear - I am by no means stating that the study should be dismissed on these grounds. It’s certainly possible that even if the assistants were skewed ever-so-slightly conservative that they wouldn’t let their biases affect their collection of data. The article linked to in the OP says only that they were “directed” to “scour US media coverage of the past 10 years.” It doesn’t say how free they were to pick what media to scour. Perhaps their biases would be irrelevant. I don’t know.
I only raise this issue because I think it’s a perfect example of how subtle bias can be, and how fiendishly difficult it is to measure.
I think Evil One brings up an important point here. News itsems pick up from news releases on behalf of this or that institution, having this or that (partly subjective, it must be said) “score”.
So, a vital part of this study is what the average score of all such institutions was, and how many releases there were in total. If the scoring of the study was such that the average score of institutions was >50, we ought not be surprised at all to see the media show a score of >50. The question would become why are there more left-leaning institutions?
Or, if the average score was 50ish, but the left-labelled institutions put out more releases, then that again would be utterly unsurprising if it yielded a >50 score for the whole media: the question begged would be why institution labelled leftist by the study put out more material worthy of an actual press release.
For example, take the PIPA report of voter misapprehensions in the last election. The very numbers themselves clearly show something which leftists would want to be widely disseminated. Where is the equivalent right-leaning study, which got less attention? If it does not exist, could it be that the right simply does not carry out studies which it feels worth releasing to the same extent?
jshore’s post seems to have been mostly glossed over by many here, but I’m a big proponent of following the money in something as wildly subjective as this. I trust this study as much as I’d trust a study about media bias paid for by Soros and moveon.org, performed by a professor who is a frequent mouthpiece for liberal stances.
Groseclose and Milyo are the same people who have stated in the past that PACs aren’t really that influential in elections and actually say “a billion dollars just isn’t that much” when talking about buying influence.
Excellent points. And, as I suggested before, there’s a flip side to it: it may be that the right-leaning institutions, instead of distributing their numbers to the media, give them directly to Congresspeople. Right now, Congress is controlled by the right, and so it may be more efficient for the right-leaning people to deliver their numbers directly to Congress, leading to their being cited by Congressfolk to a degree out of proportion with their actual numbers.
BlackKnight said it well. This study shows an interesting pattern, but there’s more than one explanation for this pattern; more research can determine whether it’s indicative of media bias, access to the halls of power, funding for various nonprofits, or something else.
I’m not averse to the idea that US news media show a left-leaning bias; it certainly seems plausible to me. But I’m not convinced that it’s the case, and this study isn’t the nail in the coffin that the author seems to think it is.
Daniel
Yeah, I noticed that. And, I agree that it is important to understand the bias of the study’s authors. They make a big deal about using unbiased research assistants (i.e., having the same number of Bush and Gore voters) but we should also know the biases of the ones who actually designed the study, not just the drones who did the scut-work. As for me, I didn’t even know there were as many conservative fellowships available to an academic as Groseclose seems to have won.
You’re certainly right that their bias is noteworthy, and that it’s far more significant than the bias of the research assistants (those who are making tally marks have a lot less influence on the study than those who are determining when the tally marks will be made). It’s just that it’s impossible to determine whether their bias has any real effect, except by showing how their bias comes out. So once we acknowledge their own political leanings, I’m not sure there’s much more to be said o the subject until we’ve determined that the study either stands or falls on its own merits.
Daniel
While it is impossible to make any final conclusions based on a press release (and not the entire published report, which apparently is not yet available), this study and responses to it furnish more evidence that no such research will ever satisfactorily quantify bias in media.
Think tank mentions and comparative ADA scores are very crude gauges of how rigorously a news organization attempts a balanced presentation of a controversial issue. They say nothing about how a particular story was presented. And they are a poor substitute for a simple and rational test that each viewer/listener can apply - is there clear evidence of the reporter’s beliefs and how he/she wishes the viewer/listener to believe, rather than a marshalling of relevant information with conclusions left to the audience?
The latter has gone increasingly out of style, in tandem with the vanished broadcast editorial (how many of you remember that the major TV networks frequently used to feature editorials in the nightly news?). Advocacy journalism is the current “standard”, accompanied by interview programs in which advocates for various causes shout at each other, and which are confused with news reporting by many people (including members of this board).
Seeing that even the definition of what is news cannot be recognized here, these debates strike me as increasingly meaningless.
Debunking the “study”:
There’s something ironic about a “study” of bias that conceals the bias of its own researchers, isn’t there?
You gotta love a study that concludes the RAND Corporation is more liberal than the ACLU. :rolleyes: