Wednesday, April 23, 2014

PolitiFact Wisconsin's version of the gender pay hack

Context matters – We examine the claim in the full context, the comments made before and after it, the question that prompted it, and the point the person was trying to make.
--"Principles of PolitiFact, PunditFact and the Truth-O-Meter"

It's often difficult to take seriously PolitiFact's claim that it examines context.  We have our latest example of this PolitiFailure via PolitiFact Wisconsin, on the subject of--what else?--the gender pay gap:
"Women deserve equal pay for equal work. It's just that simple," Burke, a Madison School Board member, said in an April 8, 2014 news release.

"In Wisconsin, a woman only earns 80 cents for every dollar a man earns--and pay discrimination doesn't just hurt our families, it hurts our economies, too."
What's so complicated about doing this fact check correctly?  If Democrat gubernatorial candidate Mary Burke is arguing that women deserve equal pay for equal work, then the statistic she's using is highly misleading.

PolitiFact Wisconsin explains why, in the same fact check:
In rating a number of pay-gap claims, we and our PolitiFact colleagues have found that wording is crucial. Two of those fact-checks help put Burke's claim into perspective.

Former Dane County executive Kathleen Falk, while campaigning for the Democratic gubernatorial nomination to challenge Walker in the 2012 recall election, said Wisconsin women "are paid 81 cents to the dollar of a man doing the same job."

The key phrase was "same job."
PolitiFact Wisconsin rated Falk's claim "False." And in rating similar claims before this year, PolitiFact gave them no higher than a "Mostly False" rating.

This year, however, the national PolitiFact gave President Obama a "Mostly True" for a claim closely parallel to Burke's. Apparently a Democrat politician can say it's important that we establish laws to help ensure equal pay for equal work--that's for men and women "doing the same job"--then throw in the statistic for men and women working full-time at different jobs and like magic the "Truth-O-Meter" rating rises to "Mostly True."

It's ridiculous.  This is an intentional deception.  PolitiFact overlooked it.  PolitiFact Oregon overlooked it.  And now PolitiFact Wisconsin has signed on to sell the Democratic Party's lie du jour.

The stat doesn't fit.  And if the stat doesn't fit, PolitiFact must acquit.

Monday, April 21, 2014

PunditFact's PolitiMath

Here at PolitiFact Bias, we keep some tabs on what we call "PolitiMath"--the mathematical indicators that correspond (or not) to various positions on its "Truth-O-Meter" scale.

This week offers us another potentially informative case, as PunditFact looks at whether Michael Eric Dyson was accurate in claiming that Sunday morning political talk shows "usually" feature conservative white men.

In terms of math, this case is fairly simple.  PunditFact counted 25 percent of the Sunday show guests as conservative white males--a little short of a plurality.

PunditFact also shared parallel figures compiled by the left-wing Media Matters organization.  Media Matters put the figure for conservative white men at about 29 percent, which did count as a plurality.

"Usually" means more than half the time, so using PunditFact's count Dyson was off by 50 percent.  Using the Media Matters count, Dyson was off by about 42 percent.

PunditFact rated Dyson's claim "Mostly False":
Dyson described the Sunday shows as having been "given over" to conservative white males. While that phrase isn't exact, it does suggest a dominant presence. The numbers don’t back that up. Conservatives outman the liberals but by the time you drill down to white, male, conservatives, they lose much of the edge.

Dyson pushed too far on his adjectives. We rate the claim Mostly False.
We don't understand PunditFact focusing on "given over."  It is the extended phrase "mostly given over" that provides the basis for the fact check.  "Mostly" communicates a "given over" figure exceeding 50 percent.

Going by the "Principles of PolitiFact, PunditFact and the Truth-O-Meter," a "Mostly False" statement contains an element of truth.  We are unable to identify what PunditFact thinks is the element of truth in Dyson's statement.

This case featuring Dyson compares very naturally with PolitiFact Florida's rating of Sen. Marco Rubio's statement claiming Americans are mostly conservative.  PolitiFact Florida gave Rubio a "Half True" for that one, though only one of three polls had conservatives self-identifying in majority numbers.

We certainly think there's much to criticize in the way PolitiFact's fact checkers went about rating both of these claims, but in terms of PolitiMath we can set those concerns aside and simply look at how PunditFact's numbers correlated to the "Truth-O-Meter" rating.

By the two measurements PunditFact offered, Dyson was at least 42 percent in error.  That resulted in a "Mostly False" rating.  By three measurements, Rubio was correct on one and off by a maximum of 36 percent by the two polls that showed conservatives as the plurality.

It's easy to see how Rubio's statement could count as at least partly true.  One poll unambiguously supported him.  But Dyson?  Not so much.

That's how PolitiMath works, for what it's worth.

Tuesday, April 15, 2014

PolitiFact's shaky and debatable conclusions (Updated)

On April 10, 2014, PolitiFact graded Sen. John E. Sununu "False" for a statement he made about the health care reform law's effect on the number of uninsured Americans.

There's plenty wrong with PolitiFact's reasoning in that fact check, starting with its assumption that federal spending on the ACA has to do with an increase in the number of people who sign up for insurance through an employer.  Contrary to PolitiFact's skewed focus, Sununu was making the point that the insurance exchanges were providing little bang for the buck in reducing the number of uninsured Americans.

But rather than writing an opus dealing with the entire misguided fact check, we're going to narrowly focus on one particular PolitiFlub:
A Rand survey that debuted a couple days after Sununu’s comments put the previously insured rate at about 36 percent of new marketplace enrollees. The Rand survey did not account for people who signed up for insurance in the final days of March.
If 36 percent of the marketplace enrollees were previously insured, then that leaves the balance, 64 percent, as enrollees moving out of the ranks of the uninsured.  That's a big win for Obamacare and the insurance exchanges!

Actually it's a big blunder by PolitiFact.  Our unbiased, Pulitzer Prize-winning (2009, not 2014) fact checkers left out a little "un" that reverses the percentages.  Here's how the RAND study put it (bold emphasis added):
Our estimates suggest that only about one-third of new marketplace enrollees were previously uninsured. While this percentage seems low in absolute terms, it is slightly higher than an earlier figure reported by McKinsey & Company.
So PolitiFact was way off.

We tried to give PolitiFact warning that might result in a correction.  PolitiFact solicits fact checks with the #politifacthis Twitter hashtag.  We obliged with the following:
The error in PolitiFact's reporting persists.

Update 4/15/2014 8 p.m.
PolitiFact issues a correction, less than 12 hours after we posted.  It probably didn't hurt that we repeated the above tweet @KatieLSanders earlier today.

We appreciate PolitiFact making the correction, particularly since it doesn't always happen.

Monday, April 14, 2014

No 2014 Pulitzer for PolitiFact (VIDEO)

Adolf Hitler learns PolitiFact failed for the fifth consecutive year to win a Pulitzer Prize

We're not kidding that the Tampa Bay Times' chief executive, Paul Tash, chairs the Pulitzer Prize committee this year.

We're not kidding that PolitiFact's founding editor Bill Adair had a life-sized cutout of President Obama standing in his office.  We don't know how long it was there or if it ended up in the office of Adair's replacement, Angie Drobnic Holan.

Why are we tweaking PolitiFact over its failure to win a Pulitzer this year?  Because PolitiFact uses its 2009 Pulitzer to burnish its reputation for reliability.  But PolitiFact has never been particularly reliable.  The 2009 Pulitzer was likely awarded mostly in recognition of PolitiFact's innovation, specifically producing an attractive and popular online outlet for what journalists think is serious journalism.  The Pulitzer was awarded in 2009 for a set of 13 stories.  That's what the Pulitzer committee was judging, not overall reliability.

Don't forget that hogwash has won Pulitzers in the past.

Friday, April 11, 2014

More gender wage gap shenanigans from PolitiFact Oregon

Back in February, we pointed out PolitiFact Oregon's blunder in ruling it "Mostly True" that women in Oregon earn 79 cents on the dollar compared to men for doing the same job.

PolitiFact Oregon is back today with another misleading take on the gender wage gap.
The National Partnership For Women & Families, a Washington, D.C.-based advocacy group, released a new analysis April 8, 2014, to coincide with Equal Pay Day. It cited U.S. Census data showing that women who work full time in Oregon are paid 79 cents for every dollar paid to men -- a claim PolitiFact Oregon has investigated previously and found solid.

But it went further. If the wage gap were eliminated, it said, "a working woman in Oregon would have enough money per year for 2,877 gallons of gas, 72 more weeks of food for her family or nearly 12 more months of rent.["]
PolitiFact Oregon apparently doesn't remember that its earlier fact check examined a version of the gender wage gap claim that had men and women doing the same work.

So we've got a fact check that focuses on a wage gap that exists primarily because men and women work at different jobs and men tend to put in longer hours.

This is a little like hearing "If you step on a crack, you'll break your mother's back" and then focusing intently on whether somebody stepped on a crack.

The National Partnership for Women & Families drew PolitiFact's attention with an April press release expressly designed to encourage support for paycheck fairness measures--measures that would do next to nothing to reduce the 21-cent gap that creates all the would-be disparities PolitiFact carefully verifies.

OMG!  We stepped on a crack!  Our poor mothers!

PolitiFact Oregon rules it "True" that we stepped on a crack the average woman in Oregon could afford to pay for more stuff if the wage gap was eliminated.  PolitiFact Oregon ignores the fact that the paycheck fairness measures the NPFWF supports would narrow the gender wage gap fractionally if at all.

Looks like PolitiFact Oregon is too busy fact checking to worry about little details like that.


We liked The Oregonian's added touch of promoting its gender wage gap story with teaser headlined with "That 79-cent gender pay gap and what it will buy."

A 79-cent gap!  What happened to the 21-cent gap?

The accompanying video has PolitiFact Oregon reporter Dana Tims expressing the NPFWF's argument, saying the pay gap would narrow if women were compensated fairly.  Tims also says the claim of a 21-cent pay gap in Oregon pretty much holds up when women and men are doing the same work, which is baloney.

Correction 4/12/2014:  Replaced the "e" in "women" in the next-to-last paragraph with an "a," changing it to the grammatically correct singular form.

Thursday, April 10, 2014

PolitiFact's PunditFact fantasizes about Dick Morris

The title and deck say it all:

When PunditFact is quoting conservative pundit Dick Morris, Morris plainly uses the word "probably."  In PunditFact's deck paraphrase, the "probably" disappears, and Morris is portrayed as saying there is proof that over 1 million people voted twice in 2012.

PunditFact's full quotation of Morris shows that he did not even say there was proof that over a million people "probably" voted twice in the 2012 election:
"It's most important data I've read in a year," Morris said on Fox News’ Hannity. "The elections commissioner there, Kim Strach, did a study of those who voted in North Carolina who also voted in another state in 2012 and she found 35,500 people voted in North Carolina and voted in some other state.

"And only 27 states pool that data. Texas, California, New York and Florida did not pool their data. So you're talking about probably over a million people that voted twice in this election. This is the first concrete evidence we've ever had of massive voter fraud. We’ve talked about it ad nauseam. This proves it."

Morris was obviously talking about proof of massive voter fraud, not proof of any specific number of fraud cases.

It's great having fact checkers who simply make things up, isn't it?

Sunday, April 6, 2014

Twitter critic: '(N)ame one of [PolitiFact's] tweets today which is factually incorrect.'

Our would-be Twitter critic, Matthew Chapman, is at it again.  We're addressing some public criticisms owing to the possibility that the critics' thoughts echo those of left-leaning readers who visit this site.

PolitiFact had another bad week fact checking, and @nextinstinct tweaked Chapman (@fawfulfan) about it.  Chapman responded with the following:
I (@ZebraFactCheck) answered with this:
Mine was a specific answer to Chapman's challenge.  He asked for an example of a factually incorrect tweet, and I gave him one.  This is PolitiFact's tweet about Rubio:
What was it Rubio said?  Here's how PolitiFact quoted it in the body of its fact check (bold emphasis added):
"I mean, the purpose of Obamacare was not to get 7 million people or 6 million people, or whatever the number now is, to sign up on a website," Rubio said. "The purpose of Obamacare, according to them, was to get more people insurance. And by all accounts, it's going to fall woefully short. You're still going to have 30-some-odd million people in this country uninsured."

We wanted to know if Rubio’s claim was correct that the health care law was falling short of its goals.
We posit that any intelligent person ought to be able to consider Rubio's statement and see that he does not say that Obamacare is falling short of its 7 million signup goal.  He's saying it's falling short of its goals for lowering the number of uninsured Americans.  PolitiFact's tweet about Rubio is false.

Chapman doesn't see it.

He's posted a fair number of tweets in response.  He said it's true the goal was 7 million and 7.1 million signed up.  He said PolitiFact was simply checking "Rubio's claim that ACA's signup goal was 30 million in 1st yr."  He said I didn't read the article.  He said Rubio was comparing apples to oranges.  He said I confuse the long run with the short run.  It appears he thinks we're somehow moving goalposts.

Chapman said plenty of things, many of which ought to embarrass him.  And he doesn't address the specific problem we identified with PolitiFact's tweet.  We've reminded him that Twitter is a poor venue for debate and have offered him commentary space in response to our post about the dispute.

Our argument is simple.  When Rubio says the purpose of Obamacare was not to get 7 million people to sign up on a website but rather to get more people insured, we say Rubio is distinguishing between the 7 million signups and lowering the number of uninsured Americans.  When Rubio says the ACA will fall short of its goals for lowering the number of uninsured persons, we assert that Rubio is not talking about exchange signups, which reports indicate are mostly people who already had insurance, but rather he's talking about the ACA's goals for lowering the number of uninsured Americans.  We say that when Rubio says "30-some-odd million" will remain uninsured he's talking about the number of uninsured persons, not the total number of exchange signups.  Nor was Rubio referring to any goal of 30 million signups, contrary to Chapman's assertion.

Considering the above, we think it's perfectly obvious that PolitiFact tweeted falsely when it claimed Rubio said the ACA was falling short of its signup goals.  Rubio was saying the 7 million goal for exchange signups is hardly relevant to the ACA's central goal of lowering the number of uninsured Americans.  Therefore, this serves as a good example to offer Chapman when he asks for an April 4 PolitiFact tweet that was factually incorrect.

PolitiFact could have accurately claimed Rubio said the ACA was falling short of its goals for lowering the number of uninsured Americans.  That would have been a fact.  But that's not what PolitiFact tweeted.

We think all of this is pretty obvious, and we stand ready to defend our view with facts and logic against whatever argument Chapman is able to bring.

Reply below, Chapman.