Something you often see in media analysis is that campaign reporters have a bias toward the horserace—that is, because they want an audience, they have an incentive to present electoral races as more interesting and competitive than they actually are.
It’s completely untrue. The fact is, when real politics are at stake, corporate media often go out of their way to make races seem as boring as possible—to declare them over long before most citizens have had a chance to vote.
Take this item from the homepage of the New York Times today (3/2/16) summing up Bernie Sanders’ Super Tuesday performance: “Wins for Sanders in Liberal Strongholds.” Well, that’s kind of a dog-bites-man story, isn’t it? (That’s another myth about journalism—that media love a man-bites-dog story. Actually, most news outlets adore a story that reinforces their preconceptions.)
In addition to being boring, this headline has the further demerit of being wrong: Of Sanders’ four Super Tuesday victories, one was in Oklahoma, which hasn’t voted for a Democratic presidential candidate since 1964, more than half a century ago. In recent years it’s voted Republican by about a 2-to-1 margin. Sanders won it with a 10-point margin.
Another victory was in Colorado, a solidly purple state—it voted for Obama twice, but also for George W. Bush twice—and for Bill Clinton only once, in 1992. Before that, it was reliably Republican going back, again, to 1964. Sanders had a 19-point margin of victory.
And Sanders won these non-liberal strongholds not just by tapping into hidden reserves of liberalism but by appealing to a broad range of ideologies. In Oklahoma, according to exit polls, Sanders did about equally well with self-described liberals and those who identify as conservative—the latter making up about 19 percent of Oklahoma’s Democratic voters. Clinton won in Oklahoma among those who see themselves as “moderate.”
Colorado didn’t have an exit poll, but Sanders did well in Weld County, described by the New York Times (10/23/14) as a “conservative stronghold where 20,000 oil and gas wells pump day and night,” whereas Clinton won Boulder’s Logan County, across what the Times called “the border between the liberal and conservative corners of the American West.”
So why report that Sanders won in “liberal strongholds”? In part, it’s laziness: That’s what you thought was going to happen, so that’s what you report. In fact, that what happened was what was supposed to happen was a major theme of Patrick Healy and Amy Chozick‘s report: “Mrs. Clinton succeeded in containing Mr. Sanders to states he was expected to win, like Vermont and Oklahoma.” How very reassuring! Actually, though, 538 gave Clinton a very slight edge to win Oklahoma, essentially predicting a tie vote—rather than the 19-point blowout Sanders achieved.
538 didn’t predict an outcome in either Colorado or Minnesota, but the last Colorado poll listed by Real Clear Politics (Quinnipiac, 11/11-15/15) had Clinton up by 28 percentage points—47 points better than her actual performance. The latest Minnesota poll (Star Tribune, 1/25/16) had Clinton 34 points ahead—whereas actual caucus-goers there picked Sanders by 24 points. These are surprising turnarounds—in Minnesota’s case, a shift of 58 percentage points in a little over a month. But rather than stressing the surprising and dramatic nature of these outcomes—and the difficulties of predictions in this election season in particular—the New York Times instead chose to mislead its readers by telling them that that’s what was expected all along.

(NYT, 3/1/16)
The motivation for this perverse coverage is not hard to imagine: The New York Times is a paper that consciously sees itself as speaking for an establishment that Bernie Sanders is challenging, and the sooner that challenge comes to an end, the better. Thus the Times‘ retrospective on a campaign still very much underway, written by Jason Horowitz (3/1/16): “Democrats Turn to Hillary Clinton After Flirting With Bernie Sanders.”
Under a photo from a Virginia Clinton rally, a caption relates that the young voter pictured “said the odds against Bernie Sanders had him looking at Hillary Clinton.” Mission accomplished, New York Times.
Jim Naureckas is the editor of FAIR.org. Follow him on Twitter at @JNaureckas.
You can send a message to the New York Times at letters@nytimes.com, or write to (outgoing) public editor Margaret Sullivan at public@nytimes.com (Twitter:@NYTimes or @Sulliview). Please remember that respectful communication is the most effective.







“. . . the odds against Bernie Sanders had him looking at Hillary Clinton.” This makes absolutely no sense whatever.
“Boulder’s Logan County”? Logan County is in the conservative northeastern corner of Colorado, nowhere near ultra-liberal Boulder. FWIW, Boulder County went for Sanders, 64%-36%; while Logan County marginally went for Clinton, 51%-46%.
FAIR really should check its facts better.