Thursday, August 30, 2012

A Little Bit More and We'll Be Inches Away from Hacks

Sometimes PolitiFact's assault on consistency is so overwhelmingly obvious, it makes my brain hurt. Do these people even keep track of what they write?

A few weeks ago PolitiFact published an article analyzing an Obama campaign ad describing the president's tax plan:

Image from
The ad attempts to differentiate between the president's plan and Mitt Romney's tax plan. The ad claims millionaires, under Obama's plan, will "pay a little more." PolitiFact's article goes on to describe different figures and analyze different metrics and concludes millionaires, on average, would end up paying roughly $189,000 more in taxes. Seems like a simple thing to rate. What could go wrong?
Supporters of Obama’s tax plan are free to argue that the tax hike on high earners is wise policy or morally justifiable. However, we think that even for a millionaire, an extra $189,000 in taxes on average -- resulting in a decline in after-tax income of 8.8 percent -- goes well beyond chump change.
So how does this rate on the trusty ol' Truth-O-Meter?
We considered putting this to the Truth-O-Meter, but we decided that "a little more" is an opinion, not a checkable fact.
That's right. The phrase "a little more" is beyond the scope of objective facts. It's an opinion. I agree with PolitiFact. So what's the point of this post? Have a look at this:

Image from

When Barack Obama tells you millionaires will only pay a little more, it's an opinion. When Mitt Romney uses a common metaphor, it's a lie. In fact, it's a lie twice:

Image from

Twice PolitiFact found enough verifiable facts to assign a value of honesty to a Mitt Romney opinion. Obama? Well, heck, they don't check opinions.

All three articles were written by Louis Jacobson. Both Romney articles were edited by Martha Hamilton.


  1. It's "Pants on Fire" because we're actually 7,333 kilometers from ceasing to be a free market economy.

    That's a big difference.

  2. They're both opinions, but they both could be converted to factual statements that can be evaluated for truthfulness without much difficulty. I'm guessing they reacted to the perceived magnitude of the "free market" statement and felt compelled to respond.

    What is it that scientists say?
    "Extraordinary claims require extraordinary evidence."

    While I agree that both statements should evaluated equally. I'm guessing PolitiFact viewed the Romney statement as alarming, and thus felt it required significant evidence to even HINT at such an idea. If their leanings are indeed liberal then I'd expect this sort of defensive response. I'm curious if it is intentional or subconscious.

    1. Thanks for the comment, Chris.

      "Extraordinary claims require extraordinary evidence."

      I believe you're thinking of skeptics, not scientists. Scientists, if they're doing science, require the same strength of evidence for any claim.

  3. Of course they're both opinions. That's the problem. PolitiFact takes conventional oration and judges it against(supposedly) concrete standards. Notice that Romney's claim earns a Pants on Fire rating, which means Louis Jacobson considered it "ridiculous." What's the objective measure for that? There's isn't one, just as there isn't a definitive guide for how much "a little more" tax is. Frankly, the crowd I run with would argue we're *not* a free market economy now. But that's an opinion. It's not a verifiable, tangible fact. But that isn't mentioned when PF puts out a "report card" on Mitt Romney's credibility.

    We agree that PolitiFact tends to "react to the perceived magnitude" of a claim. But that's not fact checking, that's editorializing, and it should be labeled as such.

    Bryan and I discuss our thoughts on whether or not PF is intentionally distorting these ratings on our About/FAQ page (see upper right corner). Thanks for stopping by and leaving a comment!

  4. I do hope you try a 'Nothing to see here' effort to provide examples of stories worth fact-checking that liberal-biased fact checkers weren't interested in. It would be great to do a few and ask politifact and others to take a look at doing them, see what the response is.

    Your emphasis on the selection process as a source of bias is a very important, overlooked aspect of fact checking- in the academic world, self-selection is universally recognized as the primary source of bias in research. Some standards need to be found, no matter how difficult, or we will start turning away from the very good idea of evaluating truth in politics. There's this Romney "inches away" stmt, where fact checkers jumped in eagerly to address potential exaggeration and went with 'pants on fire', which seems ridiculous. There's also the RNC Ryan speech plant closure protest re (I love this) it being 'technically correct but phrased in a way that might leave listeners with the wrong impression.' (Kessler- two pinoccios!) I'm still reeling from that. Whether or not the President caused the closure had nothing to do with Rep. Ryan's point (Rep. Ryan is, btw, rated slightly higher in honesty than the President by the Post- quite an achievement, in this context).

    These non-fact impressions are where fact checkers are particularly weak and prone to ideological bias. I sympathize: it's not an easy task to handle exaggeration and deception even-handedly. And you can't ignore them, because one can use omission, emphasis, context, and all kinds of things to effectively lie in an important way (that's the President's game, for the most part). Maybe your point is to be reasonably 'conservative' around such: to be much, much more careful with that stuff. It's important to prevent liberal bias in these efforts because when actual conservative bias comes up to be evaluated, conservatives should not be afraid to see and learn about mistakes. But liberal bias makes it impossible to trust fact checking feedback so that it's useful, even if they have it correct.

    There's powerful evidence from solid research that conservatives are more biased in general than liberals due to genetic/neurological and cultural differences. This is often seen as an end-all be-all fact by the left. In a bit of a shell game move, they use that evidence of higher bias to assume their ideas are better. In actuality, bias levels have nothing whatever to do with who's got the correct ideas: that'd be like saying I have 20/20 vision, so I always drive to the right places in life. I address this a lot in my blog at Overall bias levels in conservatives should be considered the neurologically unavoidable but relatively minor side of a very useful coin: the desire to further conservative values of stability, good tradition, order, self-sufficiency, and a tighter focus on family and community. And anyway, bias levels in conservatives are mere percentages higher in most dimensions: they're not double, triple, or nine times as high. That kind of reporting differential in fact checking is very strong evidence of liberal bias, probably mostly in the selection process.

    I applaud the work you're doing here as volunteers, and hope your efforts will continue to focus on story selection and other weakness, so that we can see when bias truly is involved, and we can get a fairer sense of the impact of such bias.

    1. Great comments, Scott, and thanks for the praise.

      I'm less convinced than you that good evidence exists for higher levels of bias for conservatives. Perhaps that has to do with which studies we're looking at--I'll be sure to look into what you've written. Academia rivals journalism in the degree to which it slants left, so soundness of a study design far outweighs publication in a peer-reviewed journal.


Thanks to commenters who refuse to honor various requests from the blog administrators, all comments are now moderated. Pseudonymous commenters who do not choose distinctive pseudonyms will not be published, period. No "Anonymous." No "Unknown." Etc.