Grading PolitiFact:: Karl Rove and President Obama's approval rating

The issue:

Take careful note of the way PolitiFact potrays Karl Rove's statement:

The fact checkers:

Louis Jacobson:  writer, researcher
Greg Joyce:  editor

Analysis:

Regular readers know by now that establishing the context of the claim in question stands as a chief priority:

Barack Obama has won a place in history with the worst ratings of any president at the end of his first year: 49% approve and 46% disapprove of his job performance in the latest USA Today/Gallup Poll.

Rove's statement, as noted in the story, came from an op-ed appearing in the Wall Street Journal.  The paragraph occurs at the start of his column, and sets the stage for Rove to analyze the reasons behind President Obama's historic slide.  Where PolitiFact paraphrases Rove to the effect that Obama's approval ratings are the worst, PolitiFact introduces an ambiguity later exploited by author Louis Jacobson.

We'll turn our attention now to Jacobson and his account of Rove's statement:

We'll begin by clarifying a few points.

First, we should note that Rove was guilty of rhetorical excess when he said that Obama's numbers are the worst of "any president" at this point. Gallup's historical data -- the longest-running of the major polling firms -- dates back to Harry Truman, who was the first president whose entire tenure was polled in a fashion that modern experts would consider scientific. So Rove should have argued that Obama's numbers were the worst of any post-World War II president at this point in his term.

Jacobson's first point is fair.  The distinction between a vice president elevated to the presidency by circumstances compared to elected presidents certainly affects any comparison such as the one Rove would undertake.  Charitable reading would encourage the interpreter to consider that point assumed in Rove's writing.  After all, it followed from the publication of a USA Today/Gallup poll that included that observation in specific terms:

In comparison to the approval ratings for modern elected presidents in December of their first year in office, Obama's standing is the worst, though he's close to Ronald Reagan.

Rove did not take the same care in explaining himself.  And though that should not be enough to significantly fault Rove in terms of normal communication, it is the sort of omission that PolitiFact often uses to justify a "Mostly True" rating.

Jacobson's next point:

Second, Rove misspoke when he referred to measuring Obama's ratings at the end of his first year. By definition, those numbers won't be available until late January. But in assessing his statement, we've sidestepped that problem by asking Gallup to provide us the approval ratings for December of the first year in office of the postwar presidents, so that the ratings for each can be compared more directly with the figures we have for Obama.

This objection is essentially like the first.  It is charitable to take as understood in Rove's writing what USA Today made more explicit in its presentation of the data.

Jacobson's third objection treads similar territory by noting that some presidents were not elected to office, such as Nixon's successor, President Ford.

I note with approval that PolitiFact, with some apparent encouragement from pollsters--elected not to count the above three objections against Rove.  That is in keeping with the practice of charitable interpretation, though I would have been OK with PolitiFact using its "Mostly True" rating based on the three objections.  I do fault PolitiFact slightly for using some prejudicial language, however.  Rove was "guilty of rhetorical excess" and "misspoke" with respect to the first two points.  Those terms offer a grudging version of interpretive charity.

Those observations out of the way, how does Jacobson evaluate Rove's claim?

• Eisenhower, 69-22
• Kennedy, 77-11
• Nixon, 59-23
• Carter, 57-27
• Reagan, 49-41
• George H.W. Bush, 71-20
• Clinton, 53-39
• George W. Bush, 86-11
• Obama, 49-46

Compared to these predecessors, Obama's numbers are indeed the weakest -- but he's tied with Reagan in that unflattering achievement.

Note that Jacobson finds Obama "tied" with Reagan, whereas USA Today found Obama "worst" but close to Reagan.  The difference comes from the ambiguity I mentioned earlier.

There are at least two ways to understand "approval ratings."  One puts the focus squarely on the approval number.  The other considers both approval and disapproval numbers.  Jacobson used the former method, while USA Today/Gallup and Rove used the latter.

Obama only warrants a tie with Reagan if the disapproval number is thrown out.  Otherwise it serves as a definitive tiebreaker, and would fully justify Rove's statement with other considerations set aside.

PolitiFact's final assessment of Rove ("Mostly True") is fair  for the wrong reasons.

The grades:

Louis Jacobson:  F
Greg Joyce:  F

I was so pleased with PolitiFact appearing to use a reasonable semblance of charitable interpretation that I wrestled with giving Jacobson and Joyce passing grades.  The PolitiFact effort in this instance has some very positive features, and it seems only fair to credit both men on that point.

But PolitiFact claims to be in the business of fact-checking.  And dropping the disapproval number for the sake of slipping Obama into a tie with Reagan so that Rove gets dropped to "Mostly True" cannot be accepted in fact-checking.