PolitiFact, the fact-checking arm of the Tampa Bay Times, has really outdone itself with this checkof a recent Crossroads GPS ad. Their 1,100 word fact check focuses inon just 14 words and still manages to completely confuse the issue. It winds up judging a true statement “False” based on nothing but the author’s intuition of the advertiser’s bad intentions.

This certainly looks like an egregious example of political bias, but rather than assume theworst about the author, I emailed her and her editor to see if they would agree to explain themselves. After a back and forth with PolitiFact’ssenior editor, which I’ll share below, my efforts led to the addition ofan update which does absolutely nothing to address the problem. So, isthis rampant bias or just incompetence? I’ll let you be the judge.

Here’s the Crossroads GPS ad which is the focus of PolitiFact’s attention:

As you can see, it’s a list of promises the President made whichthe narrator claims were broken. In fact, the title of the ad ischeck comes at the 28-second mark. Obama says, “If you like your health careplan, you will be able to keep your health care plan.” The narrator thensays, “Broken” and adds this key line of description:

Millions could lose their health care coverage and be forced into a government pool.

After introducing the ad, writer Angie Drobnic Holan starts out witha promising look at the history of Obama’s promise. She notes that theAffordable Care Act “will change the decisions employers make abouthealth coverage, at least to some extent.” She’s already confirmed thecore of the statement she’s reviewing but, as you’ll see later, shedoesn’t know it.

The next section begins with a heading in bold “Millions losecoverage?” Holan notes that the source for the ad’s claim is areport in The Hill which reads in part:

As many as 20 million Americans could lose their employer-providedcoverage because of President Obama’s healthcare reform law, thenonpartisan Congressional Budget Office said in a new report Thursday.

The Hill story goes on to say that, at a minimum, three million peoplewill likely lose the insurance they have by 2019. And Holan clearlyunderstands this. She writes, “The law will cause some shifting around ofhow people get their insurance, though. That’s what the CBO reportdocuments.”

There’s one more part of the ad that getsfact checked. It’s the last bit wherein the narrator says people will beof this, since Obamacare is based on a mandate to purchase insurance.But what is a “government pool” exactly? PolitiFact knows exactly whatit is:

The exchanges are new virtual marketplaces where people who buyinsurance on their own will be able to comparison shop for plans. Andtechnically speaking, this is a pool, because people who buy on theexchange will be able to buy plans regardless of any pre-existing healthconditions.

At this point, PolitiFact has proventhe ad’s claim is completely “True.” Let’s just review the statementbeing checked, word for word:

Butyou’ve probably guessed by now that PolitiFact rated this absolutelytrue statement as “False.” How is this even possible? Well, it’s prettyinstructive actually, so let’s take a look.

How PolitiFact Renders a True Statement “False”

Here’s how they did it. Watch thisamazing bit of sleight-of-hand:

CBO and other independent groups project that the number of uninsuredwill actually decrease because of the health care law. Estimates vary,but the CBO has projected that between 29 million to 31 millionuninsured Americans will gain access to coverage…

[T]he Crossroads ad doesn’t mention any of those specifics.Its point seems to be simply that a lot of people will lose coverage.This is deeply misleading, because the law will increase the number ofpeople who have coverage.

The ad is clearly about the President’s promise that you could keep your insurance, not someinsurance. Instead of staying on that point, PolitiFact’s introduces anovel new interpretation of the ad’s meaning. Suddenly, it’s not aboutthe President’s promise at all, rather ” Its point seems to be simply that a lot of people will lose coverage.“Really? Where does it say that? In fact, that’s not the point of thatad and isn’t even a possible interpretation given the context. But byintentionally misreading what the ad says, PolitiFact changes thesubject to something which it can then label false.

I wrote to Holan and PolitiFact editor Bill Adair to ask them to explain this bit ofmisdirection. I pointed out in detail, similar to what I’ve writtenabove, that Holan seemed to get confused about what the ad wasclaiming. I got a response from Adair as follows:

We rated the claim False because we found the ad was inaccuratelydescribing the health care law. As you can see in our fact-check, wewere largely focused on the part that “millions could lose their healthcare coverage and be forced into a government pool.” We foundexchanges, where people will be able to choose from various plansoffered by private insurers. On balance, we felt that was such adistortion that it made the claim False.

Still, we recognize that readers won’t always agree with our conclusions and I’m glad you’re taking time to analyze our work.

Cheers,

Bill Adair
Editor, PolitiFact

You may have noticed, as I did, that Adair didn’t argue with my analysis of the claim that millions of people would lose theircoverage. Instead, he retreated to the last defensible position, thepart about government pools. As I’ve already noted, PolitiFact admits inthe course of its fact-check that the exchanges are “technically” agovernment pool. So how is this claim misleading? Once again, it’s timeto move the goal posts:

Under the terms of the plan, some of the people will be buyinginsurance with a government subsidy, but some will be buying with theirown money. But they’ll be buying plans from private insurance companies,to be treated by non-government doctors and hospitals.

Given the contours of the health care debate over the past two years, the ad seems like it’s trying to confuse people into thinking they’ll be signed up for a government-run health plan like Medicare. That’s actually not the case.

And there you have it. What the ad says is completely true andtechnically accurate, but PolitiFact rated it “False” because it “seemslike it’s trying to confuse people.” In other words, the author of thispiece, about whom we know nothing, has decided that this true statementcould be misleading to someone.

There may be a place for fact checkers to point out when claims aretrue and still misleading. But that’s very different from labeling acompletely true statement “False.” If PolitiFact wants to rely on theintuition of its authors about the intent of the advertisers, it oughtto make clear that is a separate question from the truth or falsehood ofthe statements themselves. It also needs to offer a lot more in the wayof justification of those intuitions. Saying something “seems like”isn’t enough to base an entire conclusion on.

Rather than acknowledge the serious shortcomings of this piece, Adair decided to add an update. You can read it yourself here. As you’ll see, it clarifies nothing that wasn’t already clear from the text. The rest of the fact check piece is unchanged.