You have failed to establish that this is the most relevant question to ask in regards to determining media bias. But, at any rate, the answer is yes.
Here is a critique of that study by the folks at FAIR in their latest annual review of which think-tanks got the most citations:
Was I supposed to?
Did I miss the voting info? I do see where " most journalists identify themselves as being centrists on both social and economic issues. " Identify themselves! On social issues 30% adnmitted they were left, 57% see themselves as centrist and 9% are right (5%other). That sounds like a good cross section of the country to me!
The section concludes with “Most journalists, therefore, would certainly not recognize themselves in the “liberal media” picture painted by conservative critics.” I don’t doubt that !!
The report didn’t just look at how they identified themselves. It also compared their actual views on various issues to the views held by the public. And, the results included the fact that “On select issues from corporate power and trade to Social Security and Medicare to health care and taxes, journalists are actually more conservative than the general public.” Yes, it may be true that they are to the left of the public on most social issues…which in fact happens to be remarkably in line with my statement earlier in this thread:
The conclusions section of that FAIR study makes several important points and is worth quoting in full:
I used to think FAIR was populated by intelligent, honest people. Now I am not so sure. This is a really bad critique.
From your quoted passage alone:
No, it doesn’t. The researchers don’t “explain away” cases where the politician cites the think tank to criticize it or attaches a self-conscious ideological label to the think tank. These cases do not even enter the data set.
From the article:
My emphasis.
There are no so-called anomalies in their analysis. With respect to Congress, the NRA seems like it should be a soft right-leaning organization. It is fundamentally single-issue, and given the acknowledged overrepresentation of smaller (read: rural) states in the Senate, this result should not be anomalous to anyone with a pulse. I suspect that most of the country does not consider the NRA to be hard right. As for the Rand Corporation, they are decidedly libertarian and they take positions on both sides of the left-right axis.
Consequently, the authors suggest:
The final paragraph of the critique is particularly dumb:
First, this revision falls victim to the original critique. Just as congressmen might quote politically effective think tanks, newspapers quote congressmen to sell papers. Second, the data must be corrected to reflect the fact that there are simply more republicans in elected office now than democrats. It should thus not be surprising that the dominant party would be quoted more often.
Finally, the assumption that Groseclose & Milyo chose their data to produce their desired results is vile. Did the authors from FAIR skip over the formal model and statistical analysis? Did G&M choose a spatial model with a Weibull distribution to achieve their expected results?
Finally, if there really is liberal bias in the media, both sets of data should reflect it. After all, the data generating process for citations in newspapers is the same.
This is a poor show from FAIR.
There is one issue on which the media displays a large and obvious leftish bias - gun control.
Her location is New Orleans and her last post was Aug 26. Maybe that’s why she’s been AWOL?
I think you misunderstood what FAIR said…or at least you interpretted it differently than I did. What they said was:
You interpretted to mean that they might be citing the think-tank because they disagree with it. However, I interpretted the “politically effective” part as suggesting that they might choose to cite a think-tank that their opponents can’t as effectively rebut. For example, it would be interesting to compute the average ADA score for the IPCC and the National Academy of Sciences as computed from Congressional citations. My guess is that both of these would come out with quite high. However, I think it is silly to conclude that these organizations are liberal.
Now, one might expect that this tendency to cite more mainstream think-tanks would occur equally on both sides of the spectrum and tend to cancel out, but I am skeptical that this is the case. In a climate where “liberal” is a dirty word in a way that “conservative” is not, I think that the Democrats will have a stronger tendency to skew their cites to the center than the Republicans will.
At any rate, their methodology would be skewed in favor of finding the media biased in favor of those who use less biased sources to support their arguments. That to me is an unfortunate flaw.
Well, since I think you misunderstood the original critique, your logic does not apply.
That would be a pretty small correction in light of the numbers that they quote. I do agree with you though that the FAIR methodology also has its problems.
I don’t know why it is particularly vile. FAIR has its point-of-view. But that these academics also have their own point-of-view seems fairly obvious to me from looking at the web sites. And, anyone who would quote a study authored by John Lott of the AEI so approvingly as if he were some sort of serious scholar is suspect in my book. The man has made cherry-picking data a high art.
Yup. You were asked to, and we can’t answer the question in the thread title if we don’t even know how “liberalism” is to be determined.
If by “leftish” you mean “sane,” why then yes, I agree with you.
This is just one possibility. I mentioned it because the G&M explicitly remove those from the data set.
I do not think this is relevant. I’ll stipulate that journalists and politicians choose their cites strategically. Suppose they cite not to clarify their ideology but to maximize the impact of their rhetoric. I grant that both sides are able to solve this problem with reasonably equal skill.
We would expect to see citations that converge on relatively few think tanks that are unanimously well-respected and “difficult to rebut”. This does not necessarily mean that they are centrist so much as, well, difficult to rebut.
In practice, we do not see this at all. This leads me to believe that neither politicians nor journalists solve the strategic problem in this way. Rather, citations seem to diverge quite strongly.
My personal liberal bias says that the NAS should score highly because the progressive left is fact-based rather than faith-based, but that is just me. 
For expressly apolitical organizations, I do not think this measure is particularly helpful.
If this were the case, why then does the study find that there is in fact a slight liberal media bias? If democratic politicians and journalists are skewing to the right to compensate for the lack of enthusiasm for liberalism, why do the data show the reverse?
I am not entirely sure I understand this critique. What do you mean when you say “in favor of”? I don’t mean to be obtuse.
Stepping back, let me give you my opinion. I think that for the most part, the methodology is pretty solid. We have to remind ourselves continually that the authors were not trying to measure media bias in any absolute sense, merely relative to congress. What this study tells me is not that the media is somehow biased leftwards but that the right is heavily overrepresented in the legislature, and that the tastes of the public are not yet in lockstep with the tastes of their representatives.
Well, since I think you misunderstood the original critique, your logic does not apply.
That would be a pretty small correction in light of the numbers thatthey quote. I do agree with you though that the FAIR methodology also has its problems.
I don’t know why it is particularly vile. FAIR has its point-of-view. But that these academics also have their own point-of-view seems fairly obvious to me from looking at the web sites. And, anyone who would quote a study authored by John Lott of the AEI so approvingly as if he were some sort of serious scholar is suspect in my book. The man has made cherry-picking data a high art.
[/QUOTE]
How liberal is the national media? Depends where one is on the political spectrum I’d say. If you are pretty far to the left (relative to some theoretical ‘center’ in the US…and we could open a can of worms on THAT as well) then the national media is going to seem either ‘normal’ or perhaps right leaning. If one is pretty far right then something like Fox news is going to seem about ‘normal’ and the regular national media is going to seem left leaning. If you are right at the center then…flip a coin. Sometimes its going to seem left leaning, sometimes right.
I guess what I’m trying to say is that as will the myriad other threads on this same subject there is no answer thats going to satisfy really anyone. Leftists are going to trot out cites showing conclusively that there is no media bias…oh, and really the national media is right leaning and gives Bush a pass. Righties are going to insist that the national media is left leaning and has always been that way, and they are going to trot out a whole different set of cites showing conclusively that the media itself has polled to the left…oh, and really the national media had it in for Bush from the get go and has hounded him no end especially during the election.
This will drag on for several pages and then die out with both sides convinced THEY were right.
-XT
Gah, submitted too soon. Nothing like enforced brevity.
It is particularly vile because it is tautological and lazy. Criticizing the authors for a point of view which FAIR induces from the study itself is pretty low and adds no information to the discussion.
G&M discuss Lott’s paper in detail, paying particular attention to its useful game theoretical innovations. FAIR is silent on this point, just as it is silent on G&M’s own formal model. I wonder if they skipped that part. Math is hard.
I don’t disagree with you…and this is in fact my whole point. In general, if one side (say the liberals) in Congress tends to use more apolitical organizations more than the other, then the authors’ method of ranking those apolitical organizations will skew them into the liberal end of the spectrum. When one then looks at how journalists use these various organizations, you will arrive at the conclusion that the media is skewed liberal. This is a problem with their methodology. (And, their analysis did include some quite prominent organizations that are fairly apolitical or at least centrist…like Rand and Brookings…that they not surprisingly found in their methodology to lean left.)
To look at it in another (although not completely independent) way, if the general quality of the research done by a decidedly liberal organization like Center for Budget and Policy Priorities is more accurate and less deceptive than the general quality of the research done by the decidedly conservative Heritage Foundation, then to the extent that journalists know this and let it influence what they quote, this will show up as a liberal bias. I.e., if CBPP says the sky is blue and Heritage says it is purple with pink polka dots, any attempt by journalists to say anything more than that there is strong controversy over sky color, giving equal balance to both statements, will be interpretted as a liberal bias.
As I noted, this becomes even more of a problem when the organizations that the liberals tend to quote, such as NAS or the IPCC, do not have a decidely liberal view.
No…You misunderstand for two reasons. First of all, their method is only a relative one…It looks at journalists relative to Congress. So, you are incorrect in your conclusion that this effect, if it occurred equally with the politicians and the journalists, would show the reverse of what is shown. Instead the effect would tend to just cancel out. Second of all, I think you are mixing things here. You seem to be assigning the strategy to use more centrist think-tanks to strengthen their political argument equally to both members of Congress and journalists. My point is that it would apply to members of Congress but would not tend to apply to journalists because journalists are trying to present a balanced view…not to argue for one side or the other. (Even most of conservative critics, except perhaps the most extreme ones, who argue that there is a liberal media bias tend to argue that the bias is not generally purposeful but rather unconscious.)
I’m not sure I understand your confusion but perhaps it gets to the idea that the bias in the study is purposeful which I didn’t necessarily mean to imply so let me just rephrase my statement: “At any rate, their methodology would be skewed toward finding the media biased in the direction of those who use less biased sources (in some unspecified absolute sense) to support their arguments. That to me is an unfortunate flaw.”
Well, okay, that is one possible explanation. But, then I think that is a pretty strong critique of the study. Sure, you are not questioning their methodology (and I think you are being too kind here for the reasons mentioned above) but you are questioning the conclusions that they drew from it. I.e., they were trying to say that the media was biased liberal (relative to the American people) and you are saying that it is really Congress that is biased conservative (relative to the American people). So, in some sense, you are saying that their narrow conclusions about the media relative to Congress are valid but their broader conclusions about this being evidence for a media bias are invalid.
However, I see G&M (and you to some degree) as being so starry-eyed about the math that it blinds one to possible serious problems. Yes, what they did was clever in some ways, but that doesn’t mean it is necessarily very accurate at getting at the correct answer to the question (at least in a broad sense) that they are trying to address. There is definitely value in doing sanity checks on their results with different methods or with looking at whether all the implications of their method are believable.
And, speaking of those implications, one of the surprising ones that they discuss but don’t really lay to rest in my view is their result that the Wall Street Journal leans way left. They give it a score of 85, which puts it significantly further left than any other media outlet that they consider and squarely on par with the average Democrat in Congress (and about halfway in between Tom Daschle and Ted Kennedy). While I agree with their arguments that there seems to be a strong wall between the rabidly conservative editorial page and the news section of the WSJ, I find it hard to swallow the degree of liberal bias that they attribute to this news section. If the news section of the WSJ is so more rabidly left … even in comparison to other news outlets often claimed to be liberal like NPR, the NY Times, and the Washington Post … then why isn’t the Right so up-in-arms about the WSJ’s reporting? Is it that the opinion section placates them? It just seems a bit strange to me.
I read the local papers and see little indication of them being liberal. The English versions are more liberal than the Arabic ones, but I would call both quite conservative.
Not even remotely.
Not even remotely.
Most of said apolitical organizations have already been removed from the data set. If more need to be purged to tweak the results, so be it. This is not a major stumbling block.
First, this critique is dealt with explicitly by the authors:
They proceed to test this. You can read their results yourself if you like.
I do no such thing. However, I erred in suggesting that politicians and journalists face the same strategic decision. They don’t.
In response to your first objection, recall that the ADA score of the politician is exogenous and not determined by what think tanks the politician cites. Think tanks acquire ideology by their citation by politicians. The media are subject to more frequent feedback than politicians, and as such are more constrained to give their consumers what they want. I contend that if we grant that liberalism is a dirty word, we would be more likely to see the media shy away from it with greater intensity than politicians. I can give feedback to the NYT by not buying their paper. I can only give feedback to Charlie Rangel every two years.
Same response as above. Journalists would use centrist or right-leaning think tanks not to strengthen their political argument but to appeal to the consumers of their products. There are some interesting models for this strategic choice. G&Y tantalizingly cite a David Baron (an outstanding political economist) unpublished manuscript that gives a rigorous model for this choice. I would hold my peace about this until I read his article.
:Shrug:
I do not loose any sleep knowing that an unspecified absolute standard may exist that may invalidate the results of this study. Until we find such a standard, let alone prove that it exists in the first place, this critique is less than compelling.
The authors do not make the explicit connection that you attribute to them. From their very first paragraph:
They did not attempt to measure the bias of the American people nor did they assume that the median member of congress represents the median voter. I would reject such an interpretation vigorously. I believe that their conclusions about media bias relative to Congress are in fact pretty good. I interpret these results to mean that the right is overrepresented. This interpretation is not at odds with anything the authors explicitly say.
Starry-eyed? Really.
The model and the specification and interpretation of the statistics are the meat of the article. This is where I would focus my critique. Neither you nor FAIR have done so. Instead, these “possible serious problems” amount to little more than “what if liberal think tanks are just better” and “what if we discovered an absolute objective standard of bias.” Perhaps it is a matter of taste, but I consider neither of these critiques serious, let alone actual problems.
Is this study satisfying in that it answers definitively how the media drifts with respect to the American people?
No. And it is not supposed to. When we can measure the bias of the American people, then we can construct such a study.
Sure enough, the authors propose alternate specifications in the Appendix. This sort of thing is an extremely valuable exercise.
My answer? The Right doesn’t read the news and just skips to the editorials. 
The fact that the WSJ is contrary to expectations does not necessarily invalidate the study. It almost makes me want to read it more often.
If the American media was a politician, it would be a moderate Republican. It was for the war and accepted the White House’s case without criticism or comment. It’s afraid of Janet Jackson’s booby. It panders towards a religious view of the world, especially Christianity, and completely avoids consideration of non-theistic or skeptical worldviews or critiques of religion. It’s tacitly pro-life. It doesn’t come out and say so, but it’s loathe to ever depict abortion rights in a positive light and goes out of its way to show the choice to keep the baby as noble and heroic. It’s becoming more tolerant on gay rights but is still squeamish on SSM and adoption and still tends to prefer gay folks in a harmless, comic relief role rather than to take them seriously as three dimensional people. It thinks that ID is a real theory and that there is some sort of real debate going on with regards to evolution. It’s indifferent to environmental issues. It’s largely ethnocentric and its interest in other countries other than the US range from complete indifference or obliviousness, to paternalistic, to mocking (France) to hostile. It pays some lip service to tolerance of other cultures and religious traditions but doesn’t really mean it. It tends to ignore or gloss over the legal and ethical failings of Republican presidents but is merciless with Democrats, even when the allegations are as petty as a blow job or as bogus as the Swift Boat accusations. It’s pro-capital punishment and idiotically supportive of drug prohibition.
Above all else, it is insanely, slavishly, religiously pro-corporate.