Thursday, March 31, 2016

Fact checker avoids checking facts

A summary article by PolitiFact Wisconsin's Tom Kertscher contributes new evidence supporting our claims that PolitiFact checks facts poorly and applies its standards inconsistently. We'll address the latter point in a later post.

Kertscher's article reviewed statements made in Wisconsin by Democratic presidential front-runner Hillary Rodham Clinton. The following example from Kertscher's story grabbed our attention:

"The Republican governor of Florida has forbidden any state employee ever to use, either orally or in writing, the words ‘climate change.’ "

news report in March 2015 made that assertion, though the governor, Rick Scott, denied it.
The story Kertscher linked in support, however, dealt only with Florida's Department of Environmental Protection. The State of Florida has quite a few employees outside of the DEP. On the face of it, Clinton exaggerated wildly and the best PolitiFact Wisconsin can do in response is prop up a he-said/she-said facade supporting Clinton.

PolitiFact Wisconsin reported falsely. The March 2015 news report made no assertion that Gov. Scott placed a "climate change" gag order on every state employee in Florida.

We sent a message to Tom Kertscher on March 30, 2016 pointing out the error.

We'll update this item if we receive any response.

Wednesday, March 23, 2016

Have Democrats ''never held up a Supreme Court nomination"? (Updated)

We said before that PolitiFact does not hold itself to the same standard it applies to others. Though perhaps PolitiFact's scarce adherence to any consistent standard makes that inevitable. Our example comes from a March 20, 2016 fact check of Senate Minority Leader Harry Reid (D-Nev.).

Let's start with the misleading headline:

PolitiFact noted in its fact check that Democrats did, in fact, hold up the Supreme Court nomination of Robert Bork, whom Reagan nominated in 1987. So how does it work out that Reid's statement is "mostly true" anyway?

It's tricky, in keeping with PolitiFact's tradition of convoluted and selective justification.

In context, Reid stipulated that he was talking about lame duck cases. That makes his statement literally true false (see Update below) under an expansive interpretation of "lame duck," since Bork was nominated before 1988, Reagan's final year in office.

Up through this point, one might argue PolitiFact is treating Reid unfairly by rating his true statement only "Mostly True."

But there's much more to this story, and PolitiFact leaves out important parts.

First and foremost, there is no historical parallel to the current situation with the Supreme Court. The Bork nomination is considered a prime turning point in the politization of the confirmation process, and there is no example of a lame-duck nomination since Reagan.

And the current situation during Obama's last year in office sets a new precedent because his choice would not replace a liberal justice but a conservative justice. With his choice of Bork, Reagan was trying to replace the Nixon-appointed Justice Lewis Powell.

PolitiFact's effort to help us understand the truth in politics completely omits any information about differences in the ways these nominations would affect the political balance on the Supreme Court. It's apparently unimportant context in PolitiFact's eyes.

PolitiFact almost makes it look like Democrats rolled out the red carpet for Bork compared to Obama's hapless nominee Merrick Garland:
Bork did face a hearing and a Senate vote, which he lost, but his confirmation process made the rules of the game more contentious.
What's left out? The Democrat-controlled Senate Committee that sent Bork's nomination to the Senate recommended the Senate reject the nomination. And Democrats had enough of a majority that Bork had no chance with 52 of 54 Democrats voting against him (four Republicans often associated with the acronym RINO also opposed Bork).

Why is this background important? Let's revisit the context of Reid's reply to Meet the Press host Chuck Todd. Todd played a clip from 2005 of Reid saying the Senate has no constitutional duty to vote on a Supreme Court nomination. Now Reid says the Republicans have that duty. Todd asked Reid what changed. What changed, Reid said, is that the Democrats have never opposed a lame-duck nominee--a history running right up through 1988, before Reid ever claimed the Senate has no duty to vote on a nomination.

Reid's answer to Todd was complete baloney, in context. PolitiFact's fact-check does nothing to emphasize that context to its readers. Instead, PolitiFact readers get a misleading headline sending the message that Democrats hardly at all obstruct the Supreme Court nominations of Republicans.

PolitiFact: Putting the Clintonian "is" in "nonpartisan" since 2007.

Update: "Lame Duck" Lameness

Jeff D. points out the elephant in the room.

Before we rule, we wanted to note a slight error in the second part of Reid’s statement that "since 1900 in a lame-duck session, there have been six (nominees) that have all been approved." We have found in a previous fact-check that since the early 1900s, there have been six Supreme Court nominees in election years, and all were confirmed. However, only one was clearly a "lame-duck" nominee, meaning the president making the nomination was no question on the way out (Reagan). The others were nominated by presidents running for re-election to serve another term (Herbert Hoover, William Howard Taft, Franklin D. Roosevelt and Woodrow Wilson, who nominated two people in 1916).
Reid's stipulation that he was talking about lame ducks makes his statement literally false, contrary to the charitable reading I gave it in the post above. Reid's statement was flat wrong, but PolitiFact arbitrarily determined that Reid's exaggeration of 500 percent (of the number of "lame duck" Supreme Court nominations) was a "slight error" that does not appear to count against Reid's eventual rating.

PolitiFact used tweezers to pull out the most truth it could from Reid's statement, leaving behind plenty of falsehood.


Correction March 23, 2016:  Replaced "Mostly False" with "Mostly True" in referring to the rating Reid received from PolitiFact. March 31 update: Added strikethrough of "true" and added "false" to clarify the meaning.

Thursday, March 17, 2016

Ted Cruz fully to blame for giving Obama too much blame?

Does PolitiFact show a left-leaning bias in the blame game?

We thought PolitiFact went a bit easy on President Obama in a State of the Union speech some time ago. Obama said businesses had created so many jobs. PolitiFact said Obama's claim was "Half True" but then later elevated the rating to "Mostly True" because the president did not take as much credit as PolitiFact had first believed.

No, of course there was no concrete explanation for why PolitiFact changed its opinion.

PolitiFact played the blame game again on March 16, 2016, this time with Ted Cruz.

Here's how it looked:

PolitiFact said Cruz said President Obama has been presiding over U.S. jobs going overseas. PolitiFact reasons that Cruz gives Obama too much blame and so rates Cruz's claim "Mostly False."

Whatever plausibility PolitiFact's rating carries from its headline and deck material ought to fade pretty quickly once readers stumble over what Cruz actually said (bold emphasis added):
[Meet the Press host] Chuck Todd played a clip of Obama saying the Republicans are significantly to blame for the angry tone of politics today.

Cruz responded, "You know, Chuck, Barack Obama's a world-class demagogue. That language there is designed to divide us. No, Mr. President, we're not angry at that. We're angry at politicians in Washington, including you, who ignore the men and women who elected you, who have been presiding over our jobs going overseas for seven years."

The part of Cruz’s comment that caught our eye was that Obama has "been presiding over our jobs going overseas for seven years." We decided to take a look. (Cruz’s staff did not respond to inquiries.)
To factually conclude that too much blame was placed, the fact checker needs a blame baseline. Knowing whether Cruz blamed the president too much requires the fact checker to reasonably gauge how much blame Cruz placed on the president.

We think Cruz made that very difficult for PolitiFact with the wording he used, for Cruz did not single out the president. Cruz first mentions anger at "politicians in Washington" and after that makes clear Obama is included in the group ("including you").

So how much blame is Cruz placing on Obama, based on what Cruz said? How is the blame divided up between "politicians in Washington" and President Obama?

We don't see any way for PolitiFact to make that determination without simply making an assumption. Cruz offered no guidance. There's nothing in the context that helps. At least in the earlier case featuring President Obama we have the context of the State of the Union address. Presidents use that address to implicitly play up the benefits of their policies.

PolitiFact apparently assumes Cruz is blaming the president particularly for some unspecified role in allowing jobs to go overseas. Cruz doesn't even specify how much blame falls on Washington politicians, let alone the president. It isn't even necessary to assume that the anger at Obama and other Washington politicians is justified anger.

Is this fact-checking? It's hard to see how it qualifies.

PolitiFact has no trouble at all, despite the ambiguous nature of Cruz's claim, finding that Cruz placed too much blame on Obama. And PolitiFact likewise has an easy time assigning blame to Cruz for wrongly assigning blame, ergo the "Mostly False" rating.

PolitiFact considered no Cruz blame on "Washington politicians" other than President Obama.

In a way, it's easy to understand why PolitiFact left the other Washington politicians out of its consideration. Keeping them in consideration makes the fact check even more difficult than doing one that places an unspecified degree of blame on Obama. Pretending Cruz did not spread the blame around makes it easier for PolitiFact to maintain the fiction that Cruz placed too much blame on Obama.

We hasten to point out that such an approach hardly qualifies as unbiased.

Thursday, March 10, 2016

Bernie Sanders, PolitiMath and the price of water in Flint

In our PolitiMath series we look at how numerical errors correlate to PolitiFact's ratings.

PolitiFact's March 7, 2016 rating of Democratic presidential candidate Bernie Sanders (I-Vt.) suits our purposes well, with Sanders claiming Flint residents pay three times for water what Sanders pays in Burlington, Vt.

PolitiFact found Sanders was right if it used outdated water rates:
When we look at average annual bills from January 2015, Sanders’ 3-to-1 comparison is pretty close. But after August, Flint customers were paying a little more than twice as much as Burlington residents.
A judge's order in August 2015 rolled back water rates. Therefore, as PolitiFact notes, Flint residents now pay about twice what Burlington residents pay, counting Flint's charges for the home water meter. Burlington doesn't charge for the water meter.

To us, it's okay if Sanders wants to round up to get to his "three times" figure. So for PolitiMath purposes, we'll calculate how much 2.5 exaggerates the difference in water rates between Flint and Burlington.

Going by PolitiFact's chart, "a little more than twice as much" turned out to be about 2.4, leading to a very modest exaggeration on Sanders' part: about 4 percent. Yes, allowing for rounding up helped Sanders immensely. That's okay. We'd handle this the same way for a conservative.

PolitiFact gave Sanders a "Mostly True" rating, by the way, for a claim that was literally false.  

Deja vu.

Tuesday, March 8, 2016

PolitiFact judges job-killers

What's new from "Objective, Nonpartisan" PolitiFact?

The expert fact-checkers/liberal bloggers at PolitiFact show everyone how to figure out whether legislation kills jobs.

PolitiFact uses two methods. The first method shows whether the Affordable Care Act caused job loss. The second method shows whether the North American Free Trade Agreement caused job loss.

Here's PolitiFact's method for the ACA, from a fact check of Sen. Ted Cruz (R-Texas, bold emphasis added):
Cruz said that Obamacare cost the country millions of jobs and had forced millions into working part-time.

The government’s employment surveys show no sign of that occurring. By every measure, millions more people are working and millions fewer are stuck unwillingly in part-time work since the time the Affordable Care Act became law. The law might have affected part-time work for certain kinds of people, but that didn’t change the improvement in the overall numbers.
PolitiFact apparently reasons that if that overall employment numbers improved then it doesn't count toward Cruz's point that "The law might have affected part-time work for certain kinds of people."

Here's PolitiFact's method for NAFTA, as related in a fact check of Sen. Bernie Sanders (I-Vt.):
Sanders said that NAFTA, which Clinton used to support, cost the U.S. economy 800,000 jobs. There is a report from a left-leaning policy group that reached that conclusion. On the other hand, many other nonpartisan reports found that the trade deal produced neither significant job losses nor job gains. This is a result of competing economic models and the challenges of teasing out the effects of NAFTA from everything else that has taken place in the economy.

The report Sanders cited is an outlier, and his use of its findings ignores important facts that would give a different impression. We rate his statement Mostly False.
The biggest difference between the two methods comes from PolitiFact's reliance on raw employment numbers when checking the claim from Cruz. Raw employment numbers were a non-factor in checking Sanders but the key to giving a "Pants on Fire" rating to Cruz.

PolitiFact cited studies supporting and contradicting Sanders, but gave no evidence supporting Cruz. The fact check of Cruz omitted mention of a Congressional Budget Office report estimating supply-side net reductions in labor (workers deciding not to work or to work fewer hours), the equivalent of about 700,000 full-time jobs.

We reason that since PolitiFact is objective and non-partisan, it follows that only a non-objective and non-non-partisan (okay, partisan!) source would mention the findings of the CBO relating to unusually slow job recovery following the 2008 recession. In the world of PolitiFact, supply-side job loss doesn't count and doesn't even apparently affect the economy.

Or, to borrow a bit from Stephen Colbert, the fact[checker]s have a liberal bias.

Jeff Adds: Note that Jon Greenberg wrote both the NAFTA and ACA pieces. This makes it even more difficult to reconcile the use of two different methods in performing the fact checks.

Correction March 10, 2016 (bww): I inexplicably identified Delaware as the state Bernie Sanders represents. The text has been changed to identify Vermont as the state Sanders represents in the Senate.

Wednesday, March 2, 2016

PolitiFingers on the scale

Did you know that PolitiFact is objective and nonpartisan?

Yes, well, it was a trick question.

Yesterday PolitiFact gave us a fantastic example of placing fingers on the scale to change the outcome of a fact check

The example comes from PolitiFact's March 1 fact check of President Barack Obama. Here's the visual:

The problem?

We just don't find it anywhere close to obvious that this:

"None of the GOP candidates have a climate change plan"

is a plausible paraphrase of this:

"There is not a single candidate in the Republican primary that thinks we should do anything about climate change."

Does PolitiFact writer Lauren Carroll make an effort to justify her paraphrase of Obama? Not from what we can tell.

Carroll's conclusion features a representative gloss:
Bush, who was in the race when Obama made his comment, and Kasich both have said they believe human-caused climate change is real and have said pursuing these alternative energy sources could mitigate the problem. But neither has outlined a specific plan.

For some Republican voters, this stance might be a plus. But it doesn’t change the veracity of Obama’s statement.
So advocating increased private-sector reliance on renewable energy sources to mitigate climate change means that Bush and Kasich think we should do nothing about climate change?


One can make Obama's statement somewhat true by (charitably) assuming that by "we" he means the federal government. If we make enough unfounded assumptions, we can find some truth in Obama's statement. But by the time we've made those assumptions we're not really engaged in fact-checking.

Yet that's what PolitiFact did, effectively putting its PolitiFingers on the scale. It's not logical to conclude that lacking a plan for the federal government to act means that there is no desire to address a problem.

This was yet another sham fact check from the fraudsters at PolitiFact.


Here's Ben Carson explaining that it's important to sustain the environment. After that, he condemns Obama's climate change plan for putting a high price tag on accomplishing almost nothing:

Jeff Adds: This editorial from Carroll seems like a rating searching for a quote. PolitiFact openly admits they select claims to rate partly based on what they find interesting.

In this case, it's easy to imagine Carroll wanting to highlight (in her view) the GOP's lax attitude toward the climate change issue and went fishing for a quote to use as a vehicle for her to express that view. That scenario would plausibly explain how she made the leap from Obama's explicit claim to the different (and invented) claim she rated.

This example highlights the selective nature of PolitiFact's body of work. Contrary to the myth that they're dispassionate researchers uncovering the truth of all claims, they're partisan actors promoting or ignoring narratives as they see fit.