08/16/2009 05:12 am ET Updated May 25, 2011

AP Pimps Poorly Sourced Price Tag On Health Care [UPDATE]

If you haven't heard about it already, the Associated Press is running report after report on the health care reform package, in which they've decided that the plan will cost $1.5 trillion. Just for the record, the Congressional Budget Office estimates the cost over the next ten years to be just over $1 trillion. Their source for the massive gap is an unnamed "House Democratic aide" who appeared magically in an initial report, and has not been heard from since.

Naturally, multiple sources indicate that Time Magazine's Mark Halperin, who could not bring value to the world even through the decomposition of his biomass, was one of the many pre-eminent journalists to parrot back AP's error all over the world.

Now I accept, as a fact of life, that unnamed sources are going to worm their way into journalism, but I'd be very averse to bridging the divide between a trillion dollars and a trillion and a half dollars on the back of a phantom. This is not a figure that the AP should be continuing to run, caveat free. Steve Benen has some suggestions:

I don't mean to sound picky, but reporting like this not only misinforms news consumers, it also has the potential to adversely affect the larger policy debate. If the AP is intent on using the $1.5 trillion figure, it could at least add some language to reflect the concerns, such as "a number Democratic leaders dispute," or, "though the CBO puts the figure closer to $1 trillion." Something.

I realize that the exact price of the proposal is unclear at this point. We're dealing with estimates on legislation that's still being shaped, and it's possible the final figure may exceed, or not, the current figures. But in light of the published CBO score, the AP's reporting is sloppy and incomplete, and runs the risk of undermining reform efforts.

I'd also suggest that this is one of those occasions where, unnamed or no, it would be useful for reporters to get in the habit of asking their sources to, as my teachers would say, "show their work." If we cannot hold the source accountable, we can at least account for the numbers.

UPDATE: The AP has defended their use of the $1.5 billion dollar figure to Greg Sargent. (I've also just received a similar email, from the AP's Paul Colford:

The Congressional Budget Office score of $1.04 trillion that the Democrats cite is the figure for the new health insurance "exchange."

However, that is a net figure, including about $237 billion in revenue raised from employer and individual mandates -- fees paid by those who don't provide or purchase care. Thus, if you look at costs, the score on that is about $1.27 trillion.

There is also a separate piece of the bill covering Medicare. It includes about $350 billion in new spending (the biggest single piece is for the so-called "doc fix," which involves the payment rate to doctors under Medicare).

Now, this is, at least, "showing your work." But yes, readers, I know what you are thinking, and Greg has anticipated your objection:

But again, the problem is that we don't yet know what the bill will cost in the end. Estimates differ. House Dems argue that it's reckless to assign a hard and fast cost before the CBO has completed its score. Yet the AP keeps describing the bill as a "$1.5 trillion plan," without registering the Dem objection -- and without including the CBO's initial analysis.

Even if you agree that the bill is likely to cost this in the end, it's still reckless of the AP to keep treating this number as established fact, when it simply isn't any such thing.

The thing I'd add is that it's not clear whether these conclusions are the Associated Press' or their original unnamed source. Looking at Colford's math, he's got a guesstimate that totals up around $1.65 trillion. If that's how they figure it, why not use that figure? Why use the unnamed source's figure? Also, why continue to use the unnamed source's figure, after you've gone and given up on citing the source, at all?

Nevertheless, what the AP is doing, essentially, is what is best known as "guessing," and they should report it as such instead of intimating that it is fact. There's still nothing wrong with using Benen's cited suggestions. Moreover, anyone else citing this number should be upfront about citing it as the AP's guesstimation.

[Would you like to follow me on Twitter? Because why not? Also, please send tips to -- learn more about our media monitoring project here.]