Grading PolitiFact: Rep. Miller versus Recovery.com

I heard a tale on the radio a few weeks ago about a competitive archer who thought he had won his event on his final shot only to find that his arrow struck the bull's eye of the target used by an adjacent competitor.  He did not win the competition.

This PolitiFact entry on Rep. Jeff Miller (R., Fla.) reminds me a bit of that story.

The issue:

Recovery.gov listed congressional districts that "do not exist."
Jeff Miller on Thursday, November 19th, 2009 in a newsletter

Nonexistent congressional districts were on Recovery.gov. They're gone now.

The "Mostly True" Truth-O-Meter rating occurs just below and to the left of the deck information quoted above. Readers cannot be blamed, I think, for wondering why the claim is not true based on the information so far.

The fact checkers:

Angie Drobnic Holan:  writer, researcher
Bill Adair:  editor


Analysis:

Though a simple solution ought to have sufficed for this case, Drobnic and PolitiFact make it much harder than it ought to be.  Drobnic gets on the wrong side of the line for objective reporting in her first paragraph:

The Obama administration has promoted its Web site Recovery.gov as a bold new step in government transparency and a convenient way for voters to see that the economic stimulus program is working. Some Republicans say the site is filled with unreliable propaganda.

"Some Republicans"?

Try numerous media outlets and some Republicans.  Drobnic partially acknowledges the former as a key finding in her story.  A pity that the omission serves to help marginalize Republican opinion in the opening paragraph.  It's easy to dismiss "some Republicans," isn't it?

Drobnic continues:

Florida Republican Rep. Jeff Miller ridiculed the site in a newsletter Nov. 19, writing about how he looked on Recovery.gov and found jobs created in Florida's 34th, 53rd, 86th, and "00th" districts.

The characterization "ridiculed" slips over the line into editorial judgment.  Though it also qualifies as a very reasonable judgment.

Drobnic quotes Miller:

"The problem is, these congressional districts do not exist," Miller wrote. "Florida only has TWENTY-FIVE congressional districts.

"We know that the Administration is pulling a 'jobs created or saved' number out of thin air despite the fact that the unemployment rate remains high," Miller wrote. "The people of Florida know. We know that although Democrats represent only 10 of Florida’s 25 districts, their districts received 60% of the stimulus funds. These numbers reek of partisanship and potential corruption."

Drobnic completed the quotation without incident.  She continues:

Miller was right that Recovery.gov did have incorrect information on it. Officials blamed it on errors entering the data and have since replaced the erroneous districts with the notation "unassigned congressional district." ABC News broke the story on Nov. 16 and the corrections were made about two days later. The error caught our attention here at PolitiFact, and we archived the Florida page that confirms Miller's observations on bad district data.

So Miller was right?  What's with the "Mostly True" rating?

After a sentence noting a separate rating of a different Miller claim, Drobnic continues:

Here, we're verifying Miller's statement that Recovery.gov listed congressional districts that do not exist. That was indeed the case until news reports brought attention to the wrong data. Officials removed the data the evening before Miller posted his newsletter, so that now the jobs and money are attributed to "unassigned congressional district." So we rate Miller's statement Mostly True.

Rep. Miller says that Recovery.gov had inaccurate information on it, but since the information he wrote of was removed by the time he published therefore Miller's claim is only "Mostly True"?

That makes no sense.

If I say "It is raining" as it pours down cats and dogs around me but by the next day it has stopped raining, it does not mean that I was wrong.  Not even to the point of "Mostly True" as the Truth-O-Meter might say.

In like manner, PolitiFact had no business grading "Mostly True" the statement they chose to evaluate.  It was simply true.

But there is a little more to the story.

Though Drobnic fails to mention it in the version published at PolitiFact, Miller did make a statement that could be graded "Mostly True" on the technicality of tense:

In fact, according to Recovery.gov, over $11 million has been spent in Florida in districts that just aren’t there.

The above statement from Miller, unlike the one chosen by PolitiFact, permits the interpretation that Miller is claiming an inaccuracy as of the time he published.

I consider it quite possible that Drobnic considered the above statement but lost it during the shuffle of the editing process.  But I can't use such conjecture in assigning grades.  This story is yet another embarrassment for PolitiFact.


The grades:

Angie Drobnic Holan:  F
Bill Adair:  F



Nov. 30, 2009:  Substituted "Miller's claim" for "the information" in order to greatly enhance the clarity of one paragraph.