Fuzzy journalism and the number of the beast

Printer-friendly versionSend by email

In the increasingly strange world of daily newspaper journalism, where editors and publishers serve on local chambers of commerce and pen right-wing blogs while dissing their staffs for signing recall petitions, the Milwaukee Journal Sentinel is setting the standard for strangeness.

The recall petition flap didn't affect the Journal Sentinel as it did a half dozen other newspapers around the state. Moreover, the J-S has won several Pulitzer Prizes in recent years for feature and investigative reporting. However, the Milwaukee newspaper's editorial page has veered sharply rightward politically in recent years. And the news department doesn't, in some respects, seem very far behind.

We're thinking mainly of the Journal Sentinel's can't-see-the-forest-for-the-trees feature called Politifact, which rates politician statements as shades of true or false, but which too often chooses to inspect rather unimportant snippets of rhetorical navel lint. At some point soon, should this trend continue, the column will rule on whether it was accurate for some public official to call Wisconsin "America's Dairyland."

Another lesser but still significant element of questionable journalism is the Journal Sentinel's continuing disinclination to report crowd estimates at public events -- even to the extent of avoiding publication of estimates by police or other agencies. Apparently, estimates cannot be trusted because they're not exact, they're only ... estimates.

This is pertinent given the following context. Weeks ago, the newspaper's Politifact took United Wisconsin to task, giving it a "false" rating for saying petition organizers had submitted more than a million signatures to recall Gov. Scott Walker, when -- gotcha! -- the official estimate by state government was closer to (raise pinky finger to lips) 931,000. Let's see how truly inconsequential this difference is, using the Journal Sentinel's own measuring standards.

The Journal Sentinel's policy on avoiding crowd estimates has evolved to the point where it is clearly capable of being more conservative than the Walker administration itself. Take, for example, the paper's coverage of a March 10 rally at the State Capitol by progressives upset with Walker's policies.

Organizers estimated a crowd of 60,000. Now, apparently, that figure could not possibly be reported by the Journal Sentinel, because it's from a self-serving source, and the newspaper simply never reports any self-serving statements by any politician or interest group. Ever. Right?

So, moving on, better to go with the estimate of the Walker administration, which has every reason to low-ball the attendance number. Walker's Department of Adminstration estimated the day's turn-out at a paltry, pathetic 35,000. But no, that estimate is not reliable, either, by Journal Sentinel standards.

One might reasonably split the difference or report the range, but the Journal Sentinel simply wasn't taking any chances. It went off in a different direction. Here was the newspaper's headline the next day covering the rally:

Thousands rally at state Capitol

That's right: The Journal Sentinel decided that outside estimates were unreliable, so it created its own estimate, which happened to be an entire order of magnitude lower than everyone else's. Indeed, the newspaper whipped up an estimate roughly one tenth as small as the one from the state government that had every reason to provide a low-ball estimate, lest the governor be embarrassed more than necessary.

As if that weren't enough, the Journal Sentinel reporter added this observation:

Saturday's crowd wasn't as large as the throngs that showed up a year ago, but on a bright, sunny, springlike afternoon, enthusiasm was high.

Technically accurate, but not particularly illuminating, especially since those year-earlier gatherings were reported to have attracted as many as 70,000 to 100,000. In context, the above line from the paper's March 10 coverage further diminished the sheer scope of the rally, which stretched for blocks in all directions.

The practical effect of this J-S brand of fuzzy journalism is to compress the difference in strength between political activism on both ends of the spectrum, so that tiny tea party rallies of a few dozen or few hundred seem nearly as important as a rally of progressives tens and hundreds of times larger. Photographs can help readers see otherwise, but only if they show the entire crowd, not just closeups.

Remember Mitt Romney's “big” rally in a Detroit stadium a few weeks ago? Some news outlets published closeup photos suggesting that the entire place might be packed, Obama-style; most media, however, showed wide-angle shots making it clear how empty the stadium really was. The Washington Post, another prize-winning newspaper that doesn't shrink from crowd estimates, reported an attendance of 1,200. Now, how hard was that?

In another career, I once as part of my work had to estimate large crowds on the fly and on deadline. It was without question difficult, but with a bit of work it wasn't impossible. I would check with police, organizers and other public sources. I would climb to a high point (say, an upper story building window) and mentally section off the crowd into little squares to create my own estimate. Then I'd calculate from all those sources a range, then an average. But I also provided descriptions of the scope of the crowd visually, including how much area it took up. It wasn't that hard. No one expected me to be the US Census. My bosses just wanted a reasonably honest sense of proportion.

The Journal Sentinel, instead, prefers taking the easy way out, and in so doing does its readers a disservice.

So, just how many individuals did surround the Capitol this March 10? At least as many as on average attend a Milwaukee Brewers game, if you believe the governor's team. Twice that and more if you believe the organizers. But only a tenth or twentieth as many, if you take the Milwaukee Journal Sentinel's reporting at face value.

That simply isn't quality reporting. It's scaredy-cat, look-the-other-way, this-ain't-that-big-of-a-deal, good-enough-for-deadline reporting. It's overly careful, fuzzy journalism. And in the context of the bean-counting standard we are regularly treated to in Politifact and elsewhere in the paper's news columns, it's inconsistent if not shameful.

Big-city Journalists who live in glass houses need to look out at the world a bit harder than this -- just as the govenor should be doing.