Grading PolitiFact: Rand Paul and the public-private employment gap

The issue:

The fact checkers:

 Louis Jacobson:  writer, researcher
Martha Hamilton:  editor


Analysis:

Rand's statement in context:

PAUL: My -- my hope now -- my hope is to be on the Budget Committee and to go through all of these numbers and, by January, to have a balanced budget that I will introduce. I want there to be a Republican alternative -- whether it wins or not, I want the Republican message to be one of balanced budgets. If they won't do it in a year, we'll say, how about two years? If they won't do it in two years, how about three years? But someone has to believe it.

AMANPOUR: Give me one specific cut, Senator-elect.

PAUL: All across the board.

AMANPOUR: One significant one. No, but you can't just keep saying all across the board.

PAUL: Well, no, I can, because I'm going to look at every program, every program. But I would freeze federal hiring. I would maybe reduce federal employees by 10 percent. I'd probably reduce their wages by 10 percent. The average federal employee makes $120,000 a year. The average private employee makes $60,000 a year. Let's get them more in line, and let's find savings. Let's hire no new federal workers.

(yellow emphasis added)

PolitiFact:

 We investigated this question 10 months ago, when we looked at a statement by Sen. Scott Brown, R-Mass., that "federal employees are making twice as much as their private counterparts." At the time, we ruled it False. But we're taking a fresh look.

PolitiFact also investigated a similar question three months ago, finding it "Half True" that the average federal worker makes more than $100,000 while the average private worker makes less than $70,000.  A fresh look seems like a fine idea.  Perhaps PolitiFact will reconcile the apparent discrepancies in its reporting.

The story goes on to make a couple of important distinctions.

Government employment does not perfectly mirror the different types of employment from the private sector.  There are fewer government burger-flippers, for example.  So the compensation comparison is an apples-to-oranges comparison to a significant degree.  Also, overall compensation is not what people normally think of when discussing salary.

Prior to finding Rand's claim "False," PolitiFact's Louis Jacobson finds it literally accurate:

According to the Bureau of Economic Analysis, a federal statistics-gathering agency, federal worker compensation in 2009 averaged $123,049, which was double the private-sector average of $61,051. That's a gap of almost $62,000 -- and is pretty close to what Paul said on This Week.

In fact, Rand's numbers are perfectly accurate when rounded to the nearest 10,000.  Rounding to the nearest 5,000 Rand could have accurately said $125,000 and $60,000, respectively.
PolitiFact offers that comparing similar jobs makes for and apples-to-apples comparison, finding that federal compensation does trend higher than private pay compensation.  If we were interested in underlying arguments ("Let's get them more in line") then PolitiFact's apples-to-apples comparison would support Rand:

Despite Paul's exaggeration of the numbers, critics of federal compensation patterns do have some valid points.

For instance, Chris Edwards, an economist at the libertarian Cato Institute, notes that federal pay has risen faster than private-sector pay in recent years, despite the recession. "BEA data show that average federal salaries rose 58 percent between 2000 and 2009, which was much faster than the 30 percent increase in the private sector," he writes.

Despite indulging in an comparison that is arguably apples-to-oranges, then, Rand is literally accurate and has a valid underlying argument touching the disparity between federal and private compensation.  "Half True," then, like PolitiFact's ruling on Mike Keown's similar claim?

No such luck.  The rating instead matches the one given earlier to Sen. Scott Brown (R-Mass.):

But let's return to Paul's assertion. Paul said that the "average federal employee makes $120,000 a year. The average private employee makes $60,000 a year." Most people hearing that would assume he was talking about salary alone, but  he was talking about total compensation, including benefits such as retirement pay and paid holidays. Although studies show federal employees typically earn more than their private-sector counterparts, the difference is nowhere near as much as the doubling Paul says. So we rate his statement False.

Rand's supposedly "problematic" phrasing is actually in line with Keown's.  Brown's choice of words was only marginally more misleading since his use of the term "counterparts" might move listeners to infer a strict apples-to-apples comparison.  And the latter probably cannot justify the "False" rating given to Brown.

The grades:

Louis Jacobson:  F
Martha Hamilton:  F

Selection bias:  Referencing past rating of Brown while ignoring past rating of Keown, failing to credit Rand's underlying argument in the final rating.


Failure to follow standards:  Finding Rand "False" despite the literal truth of his statement when given a reasonable charitable reading (total compensation rather than gross monetary salary).

That's no way to conduct a fact check.  Nor does it constitute objective analysis.


Afters:

Though it was close, I wasn't the first to point out the basic failure of this fact check on PolitiFact's FaceBook page:

Terry Kinder Alers

But PolitiFact, suppose Rand Paul had clearly stated that these numbers INCLUDED BENEFITS? Would the result still be false?

I'm worried that your group is beginning to slant your findings. I want to "like" the results, because I myself lean left - but I don't want any of my opposers to be able to find fault in your analyses.