In his May 30 “Best of the Web Today” column, Wall Street Journal OpinionJournal.com editor James Taranto criticized a May 29 Associated Press article titled, “Democrats Eye November Landslide,” and wrote: “Our theory about the mainstream media is that they are generally biased in favor of liberals and Democrats, but this ends up helping conservatives and Republicans by breeding complacency on the Democratic side.” Taranto offered “a way of putting our theory to a test,” and challenged readers to find an article from 1994 or 1980 “speculating about the possibility of a Republican landslide” -- presumably under the expectation that no such article would be found. Media Matters for America answered Taranto's challenge by presenting eight articles from 1994 that met his criteria. Since then, Taranto has twice addressed Media Matters' response but has yet to acknowledge that his theory of media bias did not pass his own test.

In his May 31 column, Taranto wrote simply:

The professionally indignant liberals at Media Matters, however, did uncover one Times piece anticipating a GOP takeover. Written by Adam “Major League” Clymer, the piece ran Oct.26, 1994, 13 days before the election. MM found seven other articles speculating about the possibility of big Republican gains, including a takeover of one or both houses of Congress: one each from the Associated Press and USA Today, two from the Washington Post and three from the Christian Science Monitor.

The earliest, from the Post, ran June 26, almost a month later in the cycle than the AP “landslide” article we cited yesterday, and it was considerably more modest in its prediction: “Some [Democratic] party officials believe the Democrats are in danger of losing control of the Senate and that their losses in the House could leave the Republicans holding the largest number of seats since the mid-1950s.”

In his June 1 column, Taranto again addressed Media Matters' response, and attacked the “rabid and often illiterate Angry Leftoids” who he claimed had emailed him in response to Media Matters' item. Once again, however, there was no discussion of his theory:

As we noted yesterday, the left-wing outfit Media Mutters responded to our Tuesday item about media predictions of a Democratic landslide by coming up with eight stories from 1994 speculating about big Republican gains. We must confess, we had mixed feelings about this. On the one hand, we're delighted that we were able to trick George Soros into subsidizing our research. On the other, MM, as always, included our email address in its posting, which meant we got spammed by rabid and often illiterate Angry Leftoids.

Given that the point of Taranto's challenge was to test his theory, one would expect that he would inform his readers whether he's going back to the drawing board.