We located a scanned version of the paper on the Internet, and it did appear to be written by Prof. Mark Kroll. His institution wasn’t listed, but after some additional searching, we found that Kroll had been teaching at the Louisiana Tech College of Business in 2005 and that he is now the business dean at the University of Texas at Brownsville. But it turns out he didn’t write the paper after all.

"Actually, I am the ‘recipient’ of the paper you refer to," Kroll told us in an e-mail. "The paper was done as a class project by three of my students in a graduate class back in 2005. The 475-to-1 ratio that you reference is listed in a table in the paper the students wrote. They do not give a specific citation for the data in the table."

The paper’s cover sheet fooled us -- as it fooled others -- because the professor’s name appears in the middle of the page, and the three students’ names appear together at the bottom in a less prominent spot. We tried to reach the co-authors -- Adam Choate, Dana Rowzee and Jerrod Tinsley, all of whom were working on their Master of Business Administration in 2005 -- but we did not hear back.

So how about the substance of the chart?

From previous fact-checks, we knew that American CEOs are generously paid, and we had confirmed that for ordinary Americans, incomes are stagnating.

But on the specific comparison of CEO pay and average-worker pay, we found two liberal groups -- the Economic Policy Institute and the Institute for Policy Studies -- that have produced long-running studies of this question.

. . .

. . .the most recent ratio from the Institute for Policy Studies is also smaller -- for 2010, it was 325 to 1. In previous years the ratio on two occasions has exceeded 475 to 1 -- to be specific, 516 to 1 in 1999 and 525 to 1 in 2000.