There’s another point to keep in mind, Johnston said, which is that there’s inherent imprecision when dealing with large numbers, or when making these conversions. Convert $5,700 in 1977 to 2007 dollars with one of these Web-based inflation calculators, and you’ll get $19,294.35 and 19,502.47, respectively. Drum’s formulation—“about $20,000” —is appropriate, Johnston said. An alternative, and possibly superior, approach would have been to say that costs in 2007 were more than 2.5 times as much as in 1977.
But while those cautions are worth keeping in mind, the general rule of thumb remains: When the data spans more than a few years, always adjust. That might cut against the instinct of some journalists—especially those who didn’t get into the business to work with numbers—to avoid “manipulating” figures, but it’s the right thing to do.
That being said, a couple words on Brown’s behalf are warranted. Some comments left at both Drum’s site and the Post’s have suggested that his motivation in not adjusting the figures was to make the increase appear larger, just as the movie industry’s habit of comparing box office revenues without adjusting for inflation makes it appear that nearly all of the top-grossing movies of all time were made in the last fifteen years, when adjusted grosses show a quite different picture. But there’s no reason to believe that’s so.
It’s also worth asking how much harm was incurred by this error. Drum, via e-mail, called it “a small misstep in an otherwise useful piece.” Elixhauser, who provided the data, was even more supportive, saying she “didn’t think it affects the storyline at all.” “I was really impressed with the quality of work that he did,” she said.
Still, a mistake is a mistake, and this one could have been prevented. Asked whether the Post has a policy on adjusting dollar figures, Sellers said that she was not aware of one, adding, “Our goal is transparency, as always, and David clearly said that these numbers were not adjusted for inflation.”
If the Post does not in fact have such a rule, it should add one—and make sure its dwindling ranks of editors are aware of it—in order to avoid repeating the mistake. “Transparency” is important, but so is meaningful information—and historical data, without appropriate context, doesn’t qualify.
A coda on this story: When Brown’s article is reprinted in the Aug. 3 edition of The Washington Post National Weekly, he said, the parenthetical in question will be changed to read: “The 1977 charge is the equivalent of $19,500 in 2007.” That’s still not ideal, but it’s a step in the right direction.