In every instance I examined, Kohn compared what a specific NYT article said with what he, Bob Kohn, believed the article should have said. This is a horrible, totally unreliable way of assessing bias. Far better would have been to to do things like:
---compare what the NYT said about democrat presidents vs what they said about republican presidents. E.g., When unemployment was 6.0% how did the NYT report about the 6.0% with each category of president? This test could have revealed a bias favoring one of these two US political parties.
Note: pages 99-102 of the book Tilt? The Search for Media Bias, by David Niven describes a study that did exactly this kind of test for US newspapers as a whole. It found that "...for those who see partisan bias at the heart of media behavior, the party variable was not significant in any of the five models. In short, coverage of the unemployment situation was driven by the total size of the unemployment situation and not the party of the White House occupant." (p. 101).
---compare what the NYT articles said (about various newsworthy events) with what other news sources said (about the same events). This way of testing NYT articles could have found falsehoods in the NYT. It could also have revealed ways that NYT articles differ from news from other news organizations.
Kohn's manner of assessing the quality of NYT articles gives readers little reliable information at all about the NYT. In every instance that I examined, Kohn's negative opinion of the NYT article could have been explained by biases and other flaws not in the article at all, but rather in Bob Kohn himself.
For example Kohn cited articles that he claimed treated 43rd president George W. Bush unfairly with regard to Bush's controversial sale of $848,000 worth of Harken stock on June 22 1990 (stock that lost 75% of its value on Aug. 20 1990, suggesting that Bush, as an employee of Harken, might have used inside information to unfairly dump soon-to-be-devalued stock onto another investor lacking inside information about the stock). However, Kohn didn't show any evidence to suggest that the NYT treated Clinton or any other democratic president any better. In this instance what Kohn regarded as negative treatment of Bush by the NYT about this stock sale could have been fully explained by an anti-president bias or a bias that emphasizes negative things about any famous person. News organizations of all kinds publish sharp (and often blatantly unfair) criticisms of all kinds of famous persons all the time.
I also found what seem to me to be lies in Kohn's book. I used the Lexis-nexis database of newspaper articles to check the accuracy of Kohn's quotings of fifteen (thereabouts...I inexcusably failed to record many details about my investigation of this book) different NYT articles. I found that all but two were quoted completely accurately. However, the two exceptions were very damning. In each case I found that Kohn had made around eight significant changes to the article. It appeared obvious to me that all of the changes had an effect of making G.W. Bush appear better. [27-Feb-2005 NOTE: I plan to soon have a full documentation of Kohn's alterations of these two articles on this web page.]
Kohn had many complaints about how the NYT treated Bush's sale of Harken stock. Kohn claims in the book that the NYT egregiously failed to mention Bush having filed an SEC Form 144 (which is used to tell the SEC that one plans to sell some stock).
Kohn bizarrely (without supporting justification) asserts in the book that form 144 is far more important to mention than is form 4 (used to tell the SEC about a completed sale) when assessing the ethics of someone's sale of a stock. (This claim by Kohn seems absurd to me, because one can always change one's plans. The really important thing is to tell the SEC whether or not [and exactly when] the sale in question occurred, which neither the SEC nor the stock-owner can know for certain until the sale has been done.)
However, when I searched Lexis I found that although a specific NYT article Kohn cited actually did mention Form 4 without mentioning Form 144, that same NYT issue did mention Form 144 in another article! As if this weren't damning enough, I found that when I compared how the NYT mentioned this form with how all other papers did, the NYT exhibited more pro-Bush coverage (by Kohn's measure) than did any other US newspaper!
I found that the NYT mentioned Form 144 in two different articles -- more than any other newspaper in the database! (see the counts below)
I did two different types of searches (one for each of the two forms) for all 277 newspapers in the Lexis-nexis category called "U.S. News". These were the counts I got:
Search 1: U.S. News, for Bush AND Harken AND Form 144
(anywhere in article)
All mentions: 11
NYT mentions: 2
Search 2: U.S. News, for Bush AND Harken AND Form 4
(anywhere in article)
All mentions: 23
NYT mentions: 4
Please note that the ratio of total mentions of each type of form implies that US newspapers in general agreed with my assessment of the relative importance of mentioning form 4 vs form 144. The total for all papers was 23 mentions for form 4 and 11 mentions for form 144, a ratio of roughly 2 to 1. Note that when the NYT mentions are removed from these counts, the ratio is still heavily slanted in favor of mentioning form 4: 19 to 9.
It's also interesting to see all the details about all the mentions of the form Kohn said was far more important to mention than the other. Keep in mind that Kohn was using the number of times the form was mentioned as a measure of how pro-Bush or anti-Bush the NYT's coverage of Bush was. Sorted (by number of mentions) this is the list:
FIRST PLACE for PRO-BUSH COVERAGE (two mentions):
The New York Times: July 4, 2002 and July 9, 2002
Ten Second-place (one mention):
The Advocate (Baton Rouge, Louisiana): Dec. 3, 2002
The Austin American-Statesman (Texas): July 4, 2002
Copley News Service: July 8, 2002
Deseret News (Salt Lake City, Utah): July 4, 2002
The San Francisco Chronicle (Calif.): July 4, 2002
The Seattle Post-Intelligencer: July 4, 2002
The St. Petersburg Times (Florida): July 21, 2002
The Sunday Oregonian (Oregon): July 14, 2002
The Washington Post: July 9, 2002
The Washington Times: July 11, 2002
266 also-rans (no mention): all other 266 US newspapers in the Lexis-nexis U.S. News database