Saturday, July 14, 2012

PolitiFact and the unimportance of accuracy in fact checking

If you disagree with a ruling, we encourage you to e-mail the writer or editor with your comments about our ruling.
--Principles of PolitiFact and the Truth-O-Meter

We strive to make our work completely accurate. When we think we make a mistake, we correct it if we feel like it and note it on the original item.
--The (annotated) Principles of PolitiFact

Is accuracy important in fact checking?

Most would say it is, I suspect.  My many criticisms of PolitiFact are peppered with the message that using the fact check label raises the necessary journalistic standard.

PolitiFact often seems strangely ambivalent about its standard of accuracy.  I occasionally send a message to the writer and/or editor of a PolitiFact story to point out problems.  Occasionally I get a warm reply.  Somewhat more often I get a defensive reply that does not address the issue.  And most commonly I receive no reply and the problem in the fact check goes unresolved.  Sometimes I post about my attempts to alert PolitiFact about its errors under the tag "Piquing PolitiFact."

And here's the latest example, in response to a PolitiFact story titled "Nancy Pelosi says 'everybody' will get more and pay less under the health care law":
Dear Jon Greenberg,
cc Angie Drobnic Holan

Perhaps my favorite thing about PolitiFact is the way it asks readers to point out problems with its stories while it proceeds to ignore the problems readers point out at an apparently very high rate.

But enough about me.  Let's talk about you.

You wrote:
PolitiFact has looked at Republican allegations that the law is bad because it will drive up premiums for most people and found those statements False.

Are you sure about that?  I found one example (rated "Pants on Fire," which I'd count as false to your benefit).  Perhaps you should include the citations in the sidebar if you're going to write such things.

You also wrote:

But in its study, the CBO number crunchers actually predict that a large number of people will want to buy better insurance. They would opt for a lower deductible, lower co-pays and a wider range of benefits. That personal choice would cause their total premiums to rise some 27 to 30 percent. For a policy covering just one person, the difference would be about $600 a year, before accounting for any of the subsidies the law provides; over half the people in the individual market would qualify for those subsidies.

The above obscures the meaning of the explanation you later quote from  Ed Haislmaier of the Heritage Foundation and runs largely counter to the explanation occurring in the CBO report (note the recurring term "requirement"):

The main elements of the legislation that would affect the amount of coverage purchased are the requirement that all new policies in the nongroup and small group markets cover at least a minimum specified set of benefits; the requirement that such policies have a certain minimum actuarial value; and the design of the federal subsidies, which would encourage many enrollees in the exchanges to join plans with an actuarial value above the required minimum. (The excise tax on high-premium plans would also affect the amount of coverage purchased; the impact of that tax is discussed in a separate section of this analysis.) Those provisions would have a much greater effect on premiums in the nongroup market than in the small group market, and they would have no measurable effect on premiums in the large group market.

Specifically, because of the greater actuarial value and broader scope of benefits that would be covered by new nongroup policies sold under the legislation, the average premium per person for those policies would be an estimated 27 percent to 30 percent higher than the average premium for nongroup policies under current law (with other factors held constant). The increase in actuarial value would push the average premium per person about 18 percent to 21 percent above its level under current law, before the increase in enrollees’ use of medical care resulting from lower cost sharing is considered; that induced increase, along with the greater scope of benefits, would account for the remainder of the overall difference.

The CBO counts two requirements and one personal choice.  You only mention personal choice and obscure the reason for it, that being people's well-understood tendency to spend more if they perceive that they're getting a good deal. The good deal comes primarily from subsidies.

Who cares?

You?  Angie Holan?

We'll see, won't we?

Bryan White
I sent the message on July 8.  Five full days later the Pelosi fact check continues to contain the same two apparent errors.

This type of result, in my experience, is typical of PolitiFact.

Correction 7/14/2012:  Supplied missing text just prior to the quoted email text.

No comments:

Post a Comment

Please remain on topic and keep coarse language to an absolute minimum. Comments in a language other than English will be assumed off topic.