The New York Times and CBS jointly published an interesting political poll this morning, showing (among other things) that Republicans hold a narrow 42%-39% advantage in the upcoming midterm elections, that President Barack Obama's approval rating (41%) is the second-lowest that this particular poll has detected over the past two years, that half or more of Republican-leaners under the age of 45 support legalizing marijuana and same-sex marriage, and that (in the words of Mediaite's Noah Rothman, anyway), "partisan Republican voters are more willing to compromise a range of issues than their Democratic counterparts."
But what caught my eye wasn't the numbers, it was the loaded and unintentionally telling adjectives the paper used to present them. Here's how the related article begins; bolding will be mine, for emphasis:
Republicans are in a stronger position than Democrats for this year's midterm elections, benefiting from the support of self-described independents, even though the party itself is deeply divided and most Americans agree more with Democratic policy positions.
Though "self-described" is technically accurate here, it is also a gratuitous modifier. Why remind readers that the "independents"—in contrast to the "Democrats" and "Republicans" and "Tea Party supporters" in the same article—arrive at their categorization through a conscious act of self-branding?
A search of the phrase on the paper's website provides a possible clue. "Self-described" is often deployed to indicate that the person in question is delusional, comically egotistical, proud of something dubious, or all three. "Jason Itzler, Self-Described 'King Pimp' Drops Names in Court," comes the top search result when filtered by relevance (a follow-up article on the King Pimp is number two). "The Artist as Bully and Self-Described Sex Machine," is the fourth item, followed by "self-described snob" at fifth and sixth. While the phrase is often used neutrally (as in the Dalai Lama being "a self-described Marxist"), even there it's in the service of providing attribution to what would otherwise be a potentially contentious claim.
You can plausibly read the NYT's lede as hinting that the main reason these divided and otherwise unpopular Republicans are eking out a lead over Democrats is that they are attracting the support of people who are either fooling us or themselves. Such a parsing exercise looks a lot less paranoid after considering the first sentence of the second paragraph:
The independents in the poll — a majority of whom were white or male or under age 45 — continued to sour on President Obama's job performance.
A-HA!!! So these self-describers are actually just a bunch of white males who don't like the black president. Much like the dangerous nutbags in the Tea Party that the Times keeps warning us about.
But those of you who graduated from 3rd grade math have probably already discovered the flaw in the paper's emphasis. A large majority of EVERYBODY in the United States–including the subsection within the New York Times newsroom–is "white or male or under age 45." According to the Census, 47.7% of U.S. residents are male, 60.5% are under 45, and 72.4% are "white." By my cocktail-napkin calculations, that means as many as 90% of Americans belong to at least one of these three categories (please correct me in the comments). With about the same amount of relevance, the Times could have re-written that sentence as: "The independents in the poll — a majority of whom believe in God — continued to sour on President Obama's job performance."
There are plenty of other odd wording-choices in the article (such as this gross oversimplification: "Republicans hold their edge despite the fissures in their party over whether it is too conservative or not conservative enough"), which all serve as a reminder that even the hardest of numbers are subject to the most elastic of interpretations and prejudice.