Facts In The Dark, Part III: “Poligraph” And Selection Bias

For years, now, I’ve had questions about how politicians’ statements get selected for MPR’s “Poligraph”.

If you Google the feature, one might be forgiven for thinking the feature should be named “MPR’s Michele Bachmann Bureau”.  That’d be unfair; Poligraph reporter Catherine Richert does spread some of the fact-checking love around among parties.

But I do seriously wonder what a pol has to do to get a statement picked up by Poligraph.

Huge Gaping Factual Hole, Ready For Occupancy

For example, I’ve wondered for years why Richert’s crew have never once checked up on Heather Martens, who has yet to speak her first significant truthful thing about the gun control issue.   This blog has spent years shredding everything Martens has ever said on the public stage.

It’s a big issue to me, naturally.  If I were a real cynic, I’d say it’s because MPR has invested some of its own credibility in Martens, airing an op-ed of hers in which every single one of her fifteen factual assertions – every one – was untrue.

But Martens isn’t an elected official?  Okie-Dokey – Dakota County Attorney Jim Backstrom is an elected official, and every single word he wrote about last years’ “Stand Your Ground” bill over the past four years was a lie.  Every single word.   And he’s elected, ya?   And both of them had Governor Dayton’s ear last session, when he vetoed the “Stand Your Ground” bill, which had passed with a bipartisan majority in the legislature.

Is it because MPR’s target demo doesn’t care about the issue?  I could almost understand it if that were the rationale.  But I don’t suspect a news organization would get behind that as an official alibi, do you?

Checking The Facts

So I wrote Richert over the weekend.  She emailed me back bright and early Tuesday morning.  Since I didn’t specify anything would be on the record (it was late), I’ll paraphrase; she referred me to the “About Poligraph” page, and noted the feature’s ground rules involve picking one statement from each party, each week (or, at times, two from one party one week, and two from the other the next).

The “About” page also notes:

PoliGraph puts the findings into short, clear explanations accompanied with a rating — accurate, misleading, false or inconclusive.

– Accurate: These claims are entirely or mostly true. They include important details and are supported by the facts.

-Misleading: These statements that leave out key information, are exaggerated, or have been taken out of context.

– False: These claims are not true or misleading to the point of being false.

– Inconclusive: This rating typically applies to projections or estimates. While such claims could be true under certain circumstances, more information is needed.

Well, that explains a few things, anyway.

One might hope that this next bit, however…:

If this reminds you of PolitiFact.com, the Pultizer Prize-winning from the St. Petersburg Times, you’d be right. We know good ideas when we see them.

…does not.  The political bias of Politifact (and those of the Pulitzer committee, as well) are a matter worth discussion; if those are “good ideas”, MPR News may see it; I do not.

More about the “Fact-checking” industry tomorrow.

Anyway – that brings us to my question from last Tuesday.

Selection Bias?

Yesterday, we looked briefly at “Poligraph”‘s take on an Amy Klobuchar claim to bipartisanship during her debate with Kurt Bills.    While Klobuchar’s statement was accurate as far as it went – the numbers literally supported the exact letter by letter intent of the Senator’s statement – Richert’s “fact check” focused to exclusion on the numbers, while ignoring the larger context Klobuchar’s statement seems to have been meant to hide.  This earned “Poligraph” a rating of “Cherry-PIcked”

That was their weekly “Democrat” fact-check.

But today’s installment will go back to this past Tuesday’s installment, in which Poligraph hit its self-imposed weekly “GOP” quota.  There, we looked at the “Poligraph” “fact-check” of a Tony Hernandez statement linking the bank bailouts to the unemployment rate.  While Tony oversimplified the issue, there is considerable debate about the question, and Richert herself focused excessively on refuting Hernandez’ words and ignored the broader context of the remark.   Calling Hernandez’ statement “Misleading” rather than “Oversimplified” earned “Poligraph” a rating of “Obtuse”.

But I wondered:  if “Poligraph’s” quota is one article per party per week, why pick the fairly innocuous Klobuchar quote about her record of co-sponsored bills?  The claim was almost as innocuous as the Senator herself (although it covered, I maintain, a much more important context).

But let’s go back to another moment from the State Fair debate.

Check out this segment from the Hernandez-McCollum debate:

(Video courtesy MN CD4 Conservatives blog)

Here’s the money quote from Rep. McCollum:

“The Ryan Budget does nothing to move this country forward.  It only protected tax cuts for the wealthy…[when presented with a putative Democrat budget proposal]…the Republicans said “No, if we can’t have tax cuts for the upper 1%”, which by the way is borrowed money from China, that we couldn’t have the middle-class tax cuts!”

This is an unvarnished lie.  The GOP and Ryan’s plan have been all about tax cuts across the board all along, combined with broadening the tax base so that a broader share of the people are actually paying something.  The Democrats want to use “tax cuts” as a class-warfare-baiting wedge, and seek to jack up taxes on the “wealthy”.

This McCollum statement was devoid of fact.  It contains an absolute absence of truth.  There is no validity to it in any way shape or form.

And yet it passed, while Richert spent a solid day or two vetting Hernandez’ off the cuff oversimplification about the bailout, and giving Amy Klobuchar’s blandishment about her “bipartisanship” a pass.

Why was that?

So I’ll give Poligraph a “Huh?”.

Here’s another one:  :

She says there’s lots of “Federal Highway Money” involved in the new St. Croix bridge project.  But there’s actually fairly little direct federal funding involved; it’s a lot more complex than that.

Now – the standard set with Tony’s oversimplification we looked at Tuesday was that, according to “Poligraph”, “too complex to put exactly that way” is “Misleading”.

So what is this?

We give Poligraph a rating of “Double Standard” for this one.

The question is, why does “Poligraph” pick the statements they pick?

More tomorrow.

12 thoughts on “Facts In The Dark, Part III: “Poligraph” And Selection Bias

  1. Mitch: Wondering why you didn’t post the response I sent you Tuesday morning!

    Here’s the entire e-mail I sent Mitch Tuesday morning when he asked how we pick claims to check:

    Hi Mitch,

    First, here’s a link to the “About PoliGraph” section of our website: http://minnesota.publicradio.org/display/web/2010/05/19/about-poligraph/

    Here are our basic criteria for choosing statements to check.

    1. We choose statements that can be verified with sound data, and through research and interviews with experts. We don’t assess statements of opinion.
    2. We choose statements that are on the news in one way or another. For instance, if the Legislature is debating a particularly controversial bill, we’ll look at claims being made about that legislation.
    3. We tend to focus on statements that are central to a political debate and/or a candidate’s political platform or message. We also look at statements that voters are hearing a lot, for instance claims about the health care law.
    4. Finally, claims that simply pique our curiosity are always fair game. We assume that voters are also intrigued by statements that prompt us to say, “Huh. Really?”

    As you’ve probably noticed, we check one Democrat and one Republican every week, and occasionally a member of the IP. Once in a while, we switch the schedule up and check two members of the same party in one week. When that happens, we check two members of the other party the following week.

    Both the Hernandez and Klobuchar claims from last week meet several of our criteria. Both were “checkable” statements, both were made in debates, which are significant news events, and both are central to major campaign issues. Hernandez is adamantly opposed to the bank bailout, which highlights the GOP’s broader campaign theme that government has become too intrusive. And Klobuchar has made bipartisanship a cornerstone of her political persona; her claim about how many bills she has sponsored with Republicans underscores that part of her campaign message.

    Hope this helps.


  2. Ms. Richert,

    Thanks for the response.

    As I noted in my post, I didn’t print your response since I had not specified that the conversation was on the record. I treat emails as private (other than the odd abusive one) unless everyone involved knows they’re on the record.

    And thanks for your response.

    But as I note in today’s installment, I still have a question, here. Your response touches on it in the last paragraph, which clarifies the overall policy – but I’m still curious as to the logic behind the statements that get selected. I maintain that Rep. McCollum was – as always – a boundless font of near-incoherent cynical buncombe, fairly screaming for some “fact-checking”…

    …and yet we get a dissection of Klobuchar’s platitude? (And a dissection that was, I maintain, flawed to boot)

    Thanks again for the response.

  3. Catherine:

    If you’re going to take time to examine one or two statements a week then shouldn’t you pick a very important statement and then judge that on it’s quality.

    Selecting Amy saying she is bi-partisian lets you off the hook of ignoring something else like Mitch Berg suggested in this post.

    Not to mention I don’t think Amy is really bi-partisian on important issues.

    After all if she was she would’ve joined with Senate Republicans to try to pass a budget (which the senate hasn’t done for three years if you weren’t aware of that fact) to vote for a bi-partisan budget.

    Or if she was bi-partisan she would be work with eighth district Congress Chip Cravick on getting jobs for the Iron Range. Gary Gross on his blog talks about all the work Chip is doing and isn’t getting any help from Amy.

    I guess the reason you do it is you just pretend to be a fact checker, but you’re really a partisan democrat who doesn’t care about the truth. You should give your audience the truth about that.

    Mitch shows he posts from a conservative viewpoint and will be slanted.

    So why can’t you?

    Walter Hanson
    Minneapolis, MN

  4. Choosing to test whether Senator Amy Klobuchar is as “bipartisan” as she proclaims does deserve some ridicule. Some “facts” are just not worth checking. The only people really interested in the answer to the question today are folks with very small problems.

    Were there really no other “checkable” statements made by other politicians?

    Consciously or not, true or not, I’m thinking “incumbent wallflower DFL Senator truthfully claims useless fact” and “challenging MNGOP Representative candidate misleads people about banks” are really nice PR for the DFL.

    Surprising? No.

  5. I can’t believe MPR would miss the huge whopper in Congresswoman McCollum’s statement above.
    “No, if we can’t have tax cuts for the upper 1%”, which by the way is borrowed money from China”

    Whether we tax income at the current rate or at the higher rate that Democrats prefer, the income they generated in no way is borrowed from China. No one borrows money to pay taxes (except maybe the Fed). Rep McCollum mistakes the rate of taxation with the amount the federal government spends. Because we are currently spending mare than we take in, we are borrowing to cover the difference. She shows a fundamental misunderstanding of America dn our economic system; we don’t decide how much money people are entitled to keep, we decide how much the government needs to take in order to carry out its legitimate functions.

  6. Next week’s hard-hitting topic: A-Klo – scarf or chapeau? A-Klo’s strategy is to maintain an image as innocuous and above the fray. Her voting record belies the analysis she receives in the press. She is a committed Leftist, but portrayed as a pleasant, non-confrontational moderate. All we seek, Ms. Richert is an accurate portrait. She’s “playing” you.

  7. Catharine’s 4 qualifiers do not address the question of bias on the part of the selector. The “fact checks” begin with an opinion. The “fact checker” says “I think that this should be fact-checked, and I have no rigorous method of avoiding selection bias when I say that”.

    Why not just call it news analysis or editorial comment? That would be the honest approach.

  8. Ms. Richert;

    This type of reporting by NPR is one of the biggest reasons that I asked my Congressional representatives to eliminate the funding for it. Although you liberal Dems think otherwise, if all taxpayers have to involuntarily contribute to your network, then the reporting and programming should not be so slanted to the left. Sorry, but I still have not found any organization besides Fox News that is truly fair and balanced. In my observations, when they give liberals guests air time on their debates, Fox hosts don’t verbally ridicule them and don’t attempt to trip them up by changing the questions.

  9. There’s a problem inherent in any critique of this sort. I understand why so many broadcasters and media people get mad at their conservative critics. Think about it from their perspective — conservatives are pesky and annoying. At the same time, liberals never refrain from doing this sort of thing. One of the things I dislike the most is when a comment troll (cf. Dog Gone) posts something like “I notice that you’ve failed to address [global warming] [Paul Ryan is teh liar !!eleventy!!] [the latest subject of my swift descent into psychosis].”

    MPR knows its audience and also knows its audience likes having its biases affirmed, which is why Michele Bachmann is going to get much more of a gimlet eye from the self-styled fact-checkers than Betty McCollum. This editorial decision is completely rational in that sense. If MPR were to play things more down the middle, they’d probably lose more of their listeners than what they would gain from the Strange New Respect they’d get from conservatives. When cognitive dissonance keeps a roof over your head, you’ll stay the course.

    Where it becomes a problem is that MPR and all other public broadcasters are on the public teat. The only real answer is to cut the funding and let MPR and their friends compete in the marketplace.

  10. You hit the nail on the head, Mr. D.
    Back in the 80’s NPR was losing audience share. They made a deliberate decision to move to a talk-dominated format in an effort to appeal to their core group of listeners — white boomers with college degrees. This group skews heavily left.
    This was purely a mercenary decision by NPR’s management. They wanted to keep their share of a crowded marketplace, “public mission” be damned.

  11. Swiftee, if you are going to mention “papspew” you have to work “withered dugs” into the sentence somehow.
    Otherwise, it just ain’t good English.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.