NYT gets one wrong – again

Back in January, the NYT ran a piece on women and marriage in the 21st century titled “51% of Women Are Now Living Without Spouse” – an article that the rest of the MSM breathlessly ran with, in what I would characterize as barely contained glee about the idea that a majority of women in this country were shunning marriage.

Just one problem, though. The stats didn’t add up.

A couple of days after, Michael Medved wrote a piece at Townahall.com disputing the numbers in what he called “journalistic malpractice” on the part of the NYT for running an article based on faulty data. Medved wrote:

So how could reporter Roberts read the same Census figures that any American can view (“according to a New York Times analysis”) and come up with such bizarre conclusions?

It’s all based on a fundamentally dishonest decision that Roberts never acknowledges in the entire course of his lengthy article. It turns out that in his analysis he chose to count some 10,154,000 girls between the ages of 15 and 19 as “women.” It should come as no surprise that this vast group of teenagers (yes, teenagers, most of whom live at home) are officially classified as “single.” In fact, 97% of the 15 to 19 year olds identify themselves as “never married.” The Census Bureau, by the way, doesn’t call these youngsters “women” Γ’β‚¬β€œ it labels them “females” (a far more appropriate designation).

Yet even the ridiculous inclusion of his ten million unmarried teenagers couldn’t give Sam Roberts the story he wanted to report Γ’β‚¬β€œ that most American “women” are now unmarried. As a matter of fact, the Census Bureau shows that among all females above 15 the majority (51%!) are still classified as “married.”

So the New York Times required yet another sneaky distortion to shave off that last 2% from the married majority, though this bit of statistical sleight-of-hand Sam Roberts had the decency to acknowledge. “In a relatively small number of cases, the living arrangement is temporary, because the husbands are working out of town, are in the military, or are institutionalized” he writes. In other words, in his brave new majority of “women” without spouses, he includes all those thousands upon thousands of wives and mothers who are waiting and praying at home for the return of their husbands from Iraq or Afghanistan. By arbitrarily removing this 2% of all females (2,400,000 individuals) who are classified as “married/spouse absent” from the ranks of the married, and then designating as “unmarried” his millions of middle school and high school girls who are living with their parents, together with some 9 million elderly widows who have devoted much of their lives to marriage and husbands (42% of all women over 65 are widows), Roberts can finally arrive at his desired but meaningless conclusion that “most women” now “are living without a husbands.” Eureka!

Medved was on target, and the NYT’s public editor Byron Calame went on record in a Sunday op/ed acknowledging that the numbers were wrong:

Several readers, including some who perceived the article as an attack on family values, challenged the inclusion of 15-year-olds, in e-mails to me and in comments posted on the Web version of The Times. “The article is a little deceiving because it is based on the percentage of women 15 and older who are not married” wrote one reader, noting that “it’s not even legal to marry at 15” in many states. I couldn’t agree more.

The failure to prominently and clearly explain the methodology of the survey used was one of several journalistic lapses that I found in the handling of this story. The single passing reference to the range of ages included in the overall data from the Census Bureau’s 2005 American Community Survey, or A.C.S., came below the midpoint of the article. Given the teenage “women” issue, editors should have made sure at least that the age range of the survey was more fully explained before the continuation of the article on an inside page.

But editors may have made the problem worse. I saw the top portion of a draft of the article prepared by Mr. Roberts in which the age range was first mentioned in the 10th paragraph. The first reference in the published story was in the 21st paragraph.

When readers did get to the mention of what ages were included, it was incorrect. It indicated that the numbers reflected A.C.S. data on “more than 117 million women over the age of 15.” Similarly, the footnote to the graphic accompanying the article said the data there were “for people over age 15.” Both mentions of the age cutoff were so minimal that some readers missed the supposed exclusion of 15-year-olds.

Mr. Roberts has now said that 15-year-olds were included in all the data in both the article and the graphic. But there hasn’t been any correction of the two misleading explanations indicating that the data were for females 16 and older.

When I began to look into reader concerns about the article shortly after it appeared, it became clear that there was confusion over the issue of 15-year-olds. Mr. Roberts initially told me, and wrote in an e-mail, that 15-year-olds had been excluded from the “raw numbers” cited in the article, mainly because he had discovered some states’ restrictions on marriage at that age. So the statements in the article and graphic that 15-year-olds were not counted seemed at first to be consistent with what Mr. Roberts had told me and the office of the standards editor last month.

My subsequent questions, however, led to Mr. Roberts’s eventual acknowledgment that 15-year-olds had been fully included in all the data. Seeking to explain that shift, he wrote in a Jan. 30 e-mail to me: “When I realized that nothing would change by eliminating 15-year-olds, I left the numbers as is, again for consistency.”

Actually, leaving out 15-year-olds would have cast statistical doubt on the new majority. A calculation done for me by Times consultants at the Queens College department of sociology in New York shows that the number of females 16 years old and older not living with a spouse in 2005 exceeded the total living with one, but by a small number that was well within the margin of error.

Read the rest as Calame reports how Roberts has tried to spin his methodology as accurate, spin that Calame – rightly – isn’t buying:

It was discouraging to find yet another article with an unusual angle that didn’t seem to encounter many skeptical editors as it made its way to the front page. “At the Page One meeting there was agreement that the story was especially newsworthy because of the for-the-first-time-more-living-alone-than-with-a-spouse angle” Jill Abramson, the managing editor for news, wrote to me in an e-mail. “No questions about the methodology or age categories were discussed.”

In the wake of this controversy, Bill Keller, the executive editor, has decided to meet with staffers with expertise in statistics and demographics to create a “vetting network to help with the editing of articles dealing with those subjects” Craig R. Whitney, an assistant managing editor and the standards editor, said Thursday.

After dealing with three weeks of questions from readers and from me, Mr. Roberts on Monday expressed a little less certainty about the new majority trumpeted in the first paragraph of his article. He wrote to me: “I think the essence of the article remains accurate: that, depending on how one adjusts the census’s definition, about half β€” maybe a little bit more, maybe a little bit less, depending on the age group β€” of American women are living without a spouse at any given time.”

Readers deserved this kind of more tempered perspective back on Jan. 16 β€” and a more tempered story, displayed on an inside page.

But then, that wouldn’t have fit in with the NYT’s anti-family values agenda, would it?

Comments are closed.