Thursday, August 21, 2008

Is H.L. Mencken Alive and Well at the NYT?

Mencken's famous bathtub hoax was an invented history of the bathtub, designed to appeal to what readers wanted to believe about the ignorance and irrationality of people in the past. He published it as a demonstration of human credulity, reported with glee on how many people repeated it as gospel despite its obvious inconsistency with easily established historical facts, published multiple retractions pointing out how obviously false it was—and, by his account, never managed to kill the story.

My previous post pointed to a modern equivalent. The NYT (and, I think, the AP, but not Reuters, which got it right) misread a GAO report in a way that drastically altered its meaning, converting it from a plausible but boring result (a substantial majority of corporations reported no taxable income in at least one year out of an eight year period) to a wildly implausible result that nicely fitted what a lot of people wanted to believe (two-thirds of corporations reported no taxable income over that eight year period). They simultaneously made another mistake almost as bad, calculating what the corporations "should" have owed on the basis of their revenue, not their profit. The Times discovered the latter mistake and corrected it; so far as I can tell, they have not yet noticed the former mistake.

Googling around, I found an enormous numbers of online references to the story. So far, I have not found a single one, other than my piece on this blog, that spotted the mistake. I've posted the actual facts on a fair number of them, but it's like a teacup in a tempest. I have no doubt that, years from now, millions of people will still remember the scandalous, and wholly imaginary, fact from the Times article.


Anonymous said...

It has potential to become the new "47 million uninsured." The basis of that number was a report that said, in my paraphrase, "We're not really sure, but we think maybe that's the number of people who were uninsured at some point during the year."

--Steve Brecher

Mike Hammock said...

I'm still not convinced it was a mistake. Looking at second page of the report (reproduced later as Figure 1 on page 11), the table there shows that if one looks at all USCCs, a majority go without paying taxes. It is only when one looks at just the large USCCs--the light dashed line--that this is not true.

Are you suggesting that we should just be looking at the large USCCs?

Anonymous said...


There's a big difference between saying 2/3 each year and 2/3 over an 8 year period (it's a different 2/3 each year). I suppose one could argue that the Times is merely guilty of very bad writing rather than being factually wrong.

Mike Hammock said...

That's a fair point, anonymous. It would be inaccurate to say the same companies comprise those 2/3 who aren't paying taxes for all eight years. To put it another way, the original Times quote was:

"Two out of every three United States corporations paid no federal income taxes from 1998 through 2005, according to a report released Tuesday by the Government Accountability Office, the investigative arm of Congress."

This could be made clearer by rewriting this as:

"In each year between 1998 and 2005, two out of every three United States corporations paid no federal income taxes, according to a report released Tuesday by the Government Accountability Office, the investigative arm of Congress."

I don't think it's unreasonable to believe that this is a case of sloppy writing that encourages an incorrect interpretation. The writer may have had the facts clear in his head and failed to realize that the way it was written created confusion.

Anonymous said...

Could you please let me and everybody else know how we too can avoid paying taxes every year.

I would love to take a year or two off of paying taxes.

Thank you very much.

Anonymous said...

Reddit figured it out.

They also figured out that S-corps and LLC's don't pay taxes, but the owners of those corporations pay income taxes on profits.

jimbino said...

The real howler is the reported number of "alcohol related accidents." Every time I call the cops or reporters on this, they admit that the statistic includes parked cars hit with a drunk asleep inside, tipsy pedestrians killed, and the like.

DAMM, we drunks need to go after the mad mothers out there.

David Friedman said...

"Reddit figured it out."

As far as I can tell, they only saw the Reuters report, not the NYT one.

Anonymous said...

So did you write a letter to the NYT pointing out the misinterpretation? Letters to the editor are a fine old custom. If you wrote one and they ignored it, that's one thing; if no one has pointed out the problem to them, that's another thing, and can be attributed to stupidity rather than malice.

David Friedman said...

I wrote a letter to the NYT, and sent an op-ed to the WSJ. The Journal didn't want the op-ed--understandable, since by the time I had corrected the story to "they were converting two-thirds in each year into two-thirds for all years" it wasn't as striking an error as in the earlier version. I haven't heard from the NYT, and don't expect to.

Dink Singer said...

This sort of total statistical ignorance seems to be epidemic. Back on July 29, AP distributed a story with the headline "S&P: Home prices drop by record 15.8 pct. in May" and a lead reading "Home prices tumbled by the steepest rate ever in May, according to a closely watched housing index released Tuesday, as the housing slump deepened nationwide." In fact, the 15.8% decline was for the year ended in May. The monthly decline in May was .87%, the smallest monthly decline in this index since September 2007 and less than a third of the record decline of 2.7% last February. Further, the index actually rose in seven of the twenty markets tracked, so the slump certainly did not deepen nationwide.