Your Humble Blogger has been meaning to write something about this Politifact business. For those Gentle Readers who don't frequent Left Blogovia or scour the web for political misinformation, Politifact is a web site set up by the Tampa Bay Times to grade the truthfulness of politicians and pundits. Left Blogovia has been hating on them recently because (first) they chose as their Big Lie of 2011 the claim in a DNCC ad (and elsewhere) that Republicans voted to end Medicare, and (B) they graded Our Only President's SotU claim that In the last 22 months, businesses have created more than 3 million jobs first as half-true and then as mostly-true.
First, there's this: Politifact's ratings stink.
There is, however, some useful or at least informative stuff in the articles that accompany the ratings. If you were to read the article about the SotU claim, for instance, you would find the information that “During that 22-month period, the number of jobs grew by almost 3.16 million”. Sure, if you just look at the meter, you would think that the statement was only mostly-true, but if you read the article, you would find out the actual number. So that's all right, Gentle Reader.
And furthermore, they did the exact same thing when they fact-checked Mitt Romney's ad about Rick Perry's record as Governor of Texas—he said that there were a million Texans out of work, that the unemployment rate was the highest in twenty years, and that unemployment doubled, and they called it half-true, despite all three of those statements being entirely accurate. Again, if you read the article, they will tell you that the ad is accurate; if you don't, you won't find that information out.
So I don't think that the problem is entirely a structural bias against Democrats. That argument is somewhat persuasive—in order to maintain an image of non-partisan disinterest, they avoid having a massive imbalance of lies on their site, and since there actually is a massive imbalance of lies, they are compelled to make up lies on the other side of the scale. There may be some of that, and I do think that they probably were concerned that giving the Big Lie to a liar from the Other Party for the nth consecutive year would have been awkward. But I think the structural bias is actually to keep the truth-o-meter needle away from the bright green. After all, if politicians largely tell the truth, there's no need for Politifact. If you go looking for lies and you are willing to include half-truths, omissions, exaggerations and ungrounded implications, you will find them. And you will convince yourself, and possibly others, that you are doing a great service, because there are so many to find. So having decided that any politician taking credit for anything that happened while he was in office is telling a half-truth, they can happily post that truth-o-meter over and over again—and the actual data is in the article, so they are conveying all the right information. Right?
The question, then, is what possible service Politifact is engaged in. I mean, their page simply states that they “help you find the truth in politics”, as if that was obviously sufficient. You want to find the truth in politics, don't you? So Politifact is here to help.
Only—why is it important that people know the truth in politics? Is it, for instance, that people need a certain amount of accurate information to be free and self-governing? Is it that people mean less if the elected officials do not do what they promised as candidates? Is it that power cannot be checked if its use is secret? What's actually going on?
I believe that the Politifact project was started as an attempt to shame politicians into telling fewer lies. That's my guess; I don't know for sure. The project creates negative consequences for lying, which should mean that politicians will only tell lies if the gain is greater than the loss, and the higher you can make the loss, the fewer lies. I think that's the idea, and it's not a stupid idea in the first place. I don't think it works for a variety of reasons (mostly that politicians can delegate the outright lies to surrogates in the media who actually reach a lot more people than the politicians do) but as an idea, it makes sense.
I'm not sure they see, though, that having an expansive view of lying, which is exactly what is going on with their truth-o-meter, works directly against that purpose. If your meter only comes up orange when I give a completely accurate statistical statement, then why would I bother? I'm not going to get a green anyway, and frankly everybody around me is constantly getting oranges and reds and flamey-wameys without experiencing any real retribution, so seriously, what do I care if your truth-o-meter will call my statement half-true or mostly-true? I mean, Mitch Daniels got flamey-wameys for saying that nearly half of all persons under 30 in this country didn't go to work last Tuesday, when the statement was completely accurate, if utterly irrelevant. Of course, you would know that it was accurate if you skipped the truth-o-meter and read the article. So that's all right.
Tolerabimus quod tolerare debemus,