Saturday, June 23, 2007
Oh. That liberal media.
John Wixted takes a hard look at the mainstream media's claim that it reports the news objectively notwithstanding the revealed political inclinations of the average reporter or editor. Among other delights, he examines a study out of UCLA that constructs a shadow "Americans for Democratic Action" score for major media outlets. The results are very interesting, even if not surprising to the readers of this blog and others of its ilk.
Read the whole thing, and if your reading vision is going be sure to click on the graph for the big version.
5 Comments:
By 
, at
Sat Jun 23, 09:02:00 AM:
I'm only surprised by the relative positioning of the Wall Street Journal.
By 
Christopher Chambers, at
Sat Jun 23, 05:53:00 PM:
I'm curious--do right wing tools just sit around day and grumble and whine and bitch like Mr. Magoo...attacking but never accepting or finding common ground, prospecting the baser instincts, pre-setting the buttons and then daring anyone "different" to push them? Jesus Christ...there is no LIBERAL MEDIA!!! There's outfits like Pacifica, which is liberal as hell but a mere blip. Then there's the true MEDIA: cogs in huge communications and info-tainment conglomerates which care only about ratings and ad revenue/subscriber income, and then there's outfits like Roger Ailes and Rupert Murdoch's and of course smaller ones like Sinclair, Clear Channel to an extent, and Pat Robertson's "news" network. In short, you have a tiny, tiny true liberal media, a 20,000 Leagues under the Sea giant octopus of a "regular" media dumbing everthing down for bucks and "shareholder value," and a real Pirates of the Caribbean's
Krakken of a creature that's just plain right wing, no debate. So please come off this bullcrap. It's going to get old...
By Noumenon, at
Sun Jun 24, 02:40:00 AM:
The Wall Street Journal's response is here.
Brendan Nyhan did a pretty exhaustive roundup of criticisms of the study here, includes authors' responses.
The basic methodology is still pretty neat whether the study holds up or not.
I don't agree with liberals who recognize no liberal bias in the media at all. I side with those who think the tendency to line up with the Democrats is swamped by the more important tendency to support the government in general. A status-quo bias, if you will.
By SeekerBlog.com, at
Sun Jun 24, 03:07:00 PM:
There are some methodological problems with Groseclose/Milyo -- such as the large variance in measurement periods between say WSJ and CBS News that Karl noted. A repeat of the study with more funding and refined methodology would be a good thing [it's a very expensive technique compared to the typical polling study].
What I think is most important about Groseclose/Milyo is that the study attempts to measure media outputs -- rather than opinions or party affiliations of the journalists. The use of the think tank citings as a way to map ADA scores to journalist product is clever. I've not thought of a superior technique for measuring outputs, though I first read Groseclose/Milyo as a working paper in 2003, and first posted on the study in 2005.
For fun I tried a very crude Google News "output evaluator" in 2005, looking at whether the writer chose to qualify think tank citations as "liberal" or "conservative". A better-designed approach might actually be useful -- e.g., one objection to my little test was treating Brookings as a liberal proxy. Brookings has a proxy ADA score of 53 which indicates lots of centrists would not feel it necessary to give it the liberal tag. A more clearly liberal think tank would be "Center on Budget and Policy Priorities", but unfortunately there are few citations [2 out of 41 unqualified citations]. Of course my test missed all the language variations such as "left-leaning Center..." etc.
In sum, direct measurements are challenging, while polls are cheap and easy. But polls fail on the objection that "journalists are trained to ignore their personal bias".
By Assistant Village Idiot, at
Sun Jun 24, 10:58:00 PM:
seeker, agreed. Outputs are far more powerful evidence, though they are admittedly harder to discover.
Using think-tanks quoted as a measure is indeed quite clever, though some fish will slip through the net. A news aggregator like Drudge will look more centrist because he is taking the issues from the popular conversation. Also, those sources which quote their opposition in order to refute them might be pulled to the center as well - though I suppose if they are forthright enough to actually cite rather than misrepresent their opponents' views that tends centrist right there.
