PolitiFact's shell game with claim selection

There they go again.

We've pointed out the bias inherent in PolitiFact's choices about what parts of a claim to rate. And they're at it again at PolitiFact, this time at PolitiFact Wisconsin:

PolitiFact Wisconsin based its "Pants on Fire" judgment solely on the source of the money.

• Cost: about $50k (true)
• Source of funds: tax dollars (false)
• Rock considered a symbol of racism by some (true)
  

So guess where PolitiFact puts its story focus? Take it away, PF:

For this fact-check, we’ll be focusing on her claim that Wisconsin taxpayers were on the hook for the rock removal.

So PolitiFact didn't consider the amount spent on the rock removal or the reason it was moved.

Totally legit? No. It's one of the easy avenues for bias to enter fact-checking, which some people hilariously believe is strictly the telling of facts.

We've brought up in the past the "Mostly True" rating Barack Obama received during the Democratic presidential primaries when he claimed his uncle had helped liberate Auschwitz.

Here's that set of claims, for comparison:

• Uncle among Allied troops liberating concentration camp (true/truish)
• Auschwitz: (false--Soviet troops liberated Auschwitz)

In Obama's case, PolitiFact downplayed a claim it could have chosen to make the focus of its fact check. Instead, it prioritized everything else in the claim to justify the "Mostly True" rating.

To avoid that manifestation of bias, a fact checker needs to employ the same standards consistently. Picking and choosing story focus counts as yet another subjective aspect of fact check ratings.

It's a scam. And it's a lie to call it unbiased.

Yet that's what PolitiFact does.

Obama could have received a "Pants on Fire" rating with a story focus on whether his uncle liberated Auschwitz.

Campos-Duffy could have received a "Mostly True" with a story focus taking her whole claim into account and giving her credit for the true elements.

And we want these people partnering with Facebook to help decide what get throttled down?

Updated seconds after publication to tag the PolitiFact writer Laura Schulte.

PolitiScam: It's Not What You Say, It's How PolitiFact Frames It

PolitiFact's Oct. 29, 2018 fact check of President Trump gave us yet another illustration of PolitiFact's inconsistent application of its principles.

The same day as shooting at a Pittsburgh synagogue resulting in multiple deaths, Mr. Trump justified not canceling his campaign appearances by saying that terrorizing acts should not alter daily business. Trump used the New York Stock Exchange as his example, saying it opened the day after the Sept. 11, 2001 terrorist attacks.

But it didn't open the next day. Trump was flatly wrong.


PolitiFact:

Note that PolitiFact spins its description. PolitiFact says Trump claimed he did not cancel the political rally simply because the NYSE opened the day after Sept. 11. But the NYSE opening was simply an example of the justification Trump was using.

This case involving Trump carries a parallel to a fact check PolitiFact did in 2008 of then-presidential candidate Barack Obama. Both Trump and Obama made false statements. But PolitiFact found a way to read Mr. Obama's false statement favorably:

Obama claimed his uncle helped liberate Auschwitz. But Obama's uncle was never particularly close to Auschwitz. That most famous of the concentration camps was located in Poland, not Germany, and was liberated by troops from the Soviet Union.

One might well wonder how Obama received a "Mostly True" rating for a relatively outrageous claim.

PolitiFact Framing to the Rescue!

It was very simple for PolitiFact to rehabilitate Mr. Obama's claim about his uncle. The uncle was a real person, albeit an uncle in the broad sense, and he did serve with American troops who helped liberate a less-well-known concentration camp near Buchenwald, Germany.

PolitiFact explains in its summary paragraph:

There's no question Obama misspoke when he said his uncle helped to liberate the concentration camp in Auschwitz.

But even with this error in locations, Obama's statement was substantially correct in that he had an uncle — albeit a great uncle — who served with troops who helped to liberate the Ohrdruf concentration/work camp and saw, firsthand, the horrors of the Holocaust. We rate the statement Mostly True.

See? Easy-peasy. The problem? It's pretty much just as easy to rehabilitate the claim Trump made:

There's no question Trump misspoke when he said the NYSE opened the day after Sept. 11.

But even with his error about the timing, Trump was substantially correct that the NYSE opened as soon as it feasibly could following the Sept. 11 terrorist attacks. The NYSE opened the following week not far from where the twin towers collapsed.

PolitiFact only used two sources on the reopening of the NYSE, and apparently none that provided the depth of the Wall Street Journal article we linked. Incredibly, PolitiFact also failed to link the articles it used. The New York Times story it used was available on the internet. Instead, the sources have notes that say "accessed via Nexis."

All it takes to adjust the framing of  these fact check stories is the want-to. Trump was off by a week. Obama was off by a country. Both had underlying points a fact checker could choose to emphasize.

These fact checkers do not have objective standards for deciding how to frame fact checks.


Related: "Lord knows the decision about a Truth-O-Meter rating is entirely subjective"