« Assorted links | Main | How will the financial crisis affect the economics profession? »
Which countries does the NYT cover?
Here is further discussion, noting that I am a little confused about how the U.S. is counted and why it doesn't appear more often.
Posted by Tyler Cowen on November 3, 2008 at 06:36 AM in Data Source | Permalink
Comments
France being 3rd really shows you how sycophantic the NYT Style section is...
Posted by: Andrew at Nov 3, 2008 7:56:27 AM
Hey Andy, pass me some freedom fries to go with the zealotry.
Could you be a little more predictable next time?
Posted by: James at Nov 3, 2008 8:16:38 AM
On the undercount of the "U.S."
The "U.S." isn't counted. According to the note on the post, "United States: searched "America," and subtracted "Latin America," "South America," and "Central America"."
Posted by: John B. Chilton at Nov 3, 2008 8:23:33 AM
James, could you be a little more predictable next time?
Posted by: at Nov 3, 2008 9:47:57 AM
Actually, the inquiry of economic concern would be if the NYT is so cosmpolitan in its outlook as
to be of interest the world over, if that has anything to do with its present finances.
Posted by: Superheater at Nov 3, 2008 10:03:30 AM
May be heresy to say so here, but the only things I (rarely) read in the NYT are Bruni's blog (mostly to mock it and him) and the Science section.
Online is Google news, this blog and a few econ blogs, a few libertarian blogs, print is the WSJ, Economist, Reason, City Journal, and Policy Review. Just no time for what I perceive to be the very low ROI of reading the NYT.
Posted by: at Nov 3, 2008 10:26:41 AM
I think John B. Chilton's got it. It looks like the data on number of articles per country were pulled from a simple "Advanced Search" on the NYT website, set to fetch articles from Jan. 1, 2000 onward. Then they were standardized by dividing by the total number of articles published since 2000, and multiplied by 10⁵. A quick check of Germany and France gives 3181 and 4020, which seem close enough since both have doubtless been mentioned a few times since the chart was cooked up. So, shouldn't this be standardized using the total number of articles mentioning a country? This would bias the U.S. back up, at least in terms of the "total articles" figure.
Posted by: ecmendenhall at Nov 3, 2008 11:23:59 AM
ecmendenhall : Probably yes, but the relative proportions are still right.
Posted by: Neuroskeptic at Nov 3, 2008 11:38:38 AM
The US is probably under-represnted because not all articles about the US have the tags "America," "United States, etc in them. For example, an article about a hurricane in Flordia might not actually say "America" in the text. Unless, of course, we are only trying to measure relative coverage of foreign policy issues.
Posted by: at Nov 3, 2008 11:48:25 PM
