Two recent PolitiFact assessments of statements about Wisconsin public workers are badly flawed -- both in ways that lend support to critics of public workers and unions.

On Sunday, PolitiFact awarded a dubious “True” rating to George Will's statement that Wisconsin public employees would be better off than their private sector counterparts even after Governor Scott Walker's proposed benefit cuts. Here's PolitiFact's conclusion:

[I]t's important to consider an employee's total package of wages and benefits when comparing public and private sector compensation. But that's not the issue before us here. Will clearly stated that he was talking about just the benefits side of the equation. And with regard to the changes Walker has proposed -- requiring greater employee contributions to pension and health care -- Will is right that they would still leave public employees “better off” than those in the private sector. We rate Will's comment True.

So, PolitiFact's True rating is based on the belief that “Will clearly stated that he was talking about just the benefits side of the equation.” But he didn't. Here's Will's statement, as quoted by PolitiFact:

Will responded, “Donna, what you call the grassroots is a tiny minority of this tiny minority of Wisconsin people who work for the government. Three hundred thousand public employees in Wisconsin went to work -- while the teachers were clutching their little signs that say it's all about the kids, they're abandoning their classrooms, lying to their supervisors, saying they were sick, and going off to protest in defense of perquisites, which if the governor cuts them as much as he plans to do, would still leave them better off than their private sector” counterparts.

Will said that cutting state workers “perquisites … would still leave them better off than their private sector” counterparts. The simplest reading of that sentence is that the word “them” refers to workers, not perquisites. In which case, Will was not “talking about just the benefits side of the equation” -- he was saying even if the benefits are cut, the workers will be better off than their counterparts. You can, I suppose, argue that the word “them” referred to “perquisites,” though the word “they” had twice been used to refer to teachers in the same sentence, and though it would be odd for a wordsmith of Will's calibre to describe “perquisites” as “better off.” You can argue that was Will's intent -- but it certainly isn't “clear,” as PolitiFact claims.

So, basically, the entire premise of PolitiFact's assessment is flawed. There's little reason to reason to give Will the benefit of the doubt that he was referring to how well off workers are in terms of benefits only. And there's no reason to think he was “clearly” doing so. Finally, even if he was doing so, he was being misleading -- as PolitiFact notes, "[I]t's important to consider an employee's total package of wages and benefits when comparing public and private sector compensation." In so noting, PolitiFact refers to a previous fact-check by its Wisconsin bureau, which indicates that when considering both wages and benefits, state employees make less than private sector counterparts.

That PolitiFact Wisconsin entry was deeply flawed, too -- and also in a way that benefited critics of unions. PolitiFact Wisconsin awarded an entirely unjustified “false” rating to Wisconsin AFT president Bryan Kennedy's statement that state workers earn less than comparable private sector employees. We needn't go into much detail here -- a simple glance at PolitiFact's conclusion is enough to see that it is not justified:

In sum, Kennedy did not provide evidence that supports his claim, which itself provides only a partial look at the debate. If any more evidence emerges, we'll review it. In the meantime, we rate Kennedy's claim False.

That paragraph makes clear that, at worst, Politifact should declare Kennedy's claim unproven, not “false.” Indeed, by PolitiFact's standards, its assessment of Kennedy's statement is itself false. See, PolitiFact did not prove its claim, therefore -- by PolitiFact's logic -- its claim is false. But I reject PolitiFact's flawed logic, so I won't declare it's assessment “false.” Just unjustified.

But that isn't the only flaw in PolitiFact's reasoning. Let's take a look at Kennedy's statement, as presented by PolitiFact:

On the “Upfront with Mike Gousha” public affairs program, Gousha asked Kennedy if state employee unions “have a perception problem with the general public.”

“Well, I think that there's been talking points used by those who don't like government to continually bash us and make us look as if we're the haves and we're really not,” Kennedy replied during the Dec. 5, 2010 show.

“If you look across the board, we're averaging about 8 percent less than if we all worked in the private sector. Some of our people make half or a third as much as they could make if they worked in the private sector.”

Now, here's PolitiFact's summary of Kennedy's statement and the supporting evidence he provided PolitiFact:

He said Wisconsin state employees are “averaging about 8 percent less” in pay than if they worked in the private sector. But the figure he relied on was from a national study that combined pay and benefits.

Notice that the words “in pay” aren't in quotes? That's because Kennedy didn't say them. PolitiFact is hitting Kennedy for supporting a specific comment about “pay” by providing information relating to “pay and benefits” -- but Kennedy's statement did not specify that he was referring only to pay. PolitiFact just inserted that on its own. Here, let's look back at the Kennedy statement in question, as quoted by PolitiFact:

“If you look across the board, we're averaging about 8 percent less than if we all worked in the private sector.”

See? He didn't specify “pay” as opposed to “pay and benefits.”

Finally, PolitiFact complains that though Kennedy provided a study showing that state workers make 7.6 percent less in pay and benefits than comparable private sector employees, and 11 percent less in pay alone, the study was national in scope. Ok… but Kennedy never said the 8 percent figure referred to Wisconsin state employees alone, rather than national figures for state employees.

In short: PolitiFact altered Kennedy's actual statement, pretending he was more specific than he actually was. Then it completely failed to disprove even that altered statement. And then it declared the altered statement to be false, based on an apparent belief that “unproven” and “false” are synonymous.

Note also that both PolitiFact pieces baselessly narrowed the scope of discussion to a portion of compensation rather than total compensation, even though it acknowledges that it is more useful to compare total compensation, and even though it wasn't clear in either case that the statement being fact-checked referred to only a portion of compensation. In other words, PolitiFact assessed less relevent claims than the ones actually made. Weird.