Does Knowing the Price Make Right?

The EconLog has a recent post salivating over the Singapore health care system. It was an interesting piece, but I was actually more intrigued by James A. Donald's comment about how there seem to be no set prices in the U.S. health care system. Implicit in the comment is the contention that providers don't seem to know what anything costs. He's got some great vignettes from recent conversations with various hospital reps. (Here is some additional evidence from the academic literature.)

The lack of prices, transparent or otherwise, is something market-oriented reformers point to as a tremendous problem with the U.S. health care system. Obviously, prices play a big role in consumer to seek care and provider decisions to offer it, especially in the context of a third party payer system. The argument then is that nonexistent or opaque prices mean nonexistent or opaque incentives, which in turn leads to inefficient outcomes and high costs.

Here is a related, but slightly more focused, line of inquiry: at the physician level, what are the consequences of providers lacking knowledge about the true costs of their actions? Does knowing the cost of care lead to changes in what sorts of treatments, diagnostic tests, etc are provided?

Martin Andersen, a former Yalie and good friend of mine, now doing his PhD in health economics at Harvard, did some interesting work on this subject last year for his final project in PLSC 512: Experiments in Social Science, a course he and I took together. In particular, he randomly assigned clinical vignette-based surveys to a group of fourth year medical students. Half of the questionnaires had a clinical scenario followed by a choice of biologically equivalent treatments. The other half were exactly the same, but this time the cost of each treatment drug was provided as well. Martin found that the information to be powerful: providing information on costs or prices led the average student to pick the cheaper options and the difference was practically and statistically significant even in the small sample he was using.

If this result holds more generally, what can we take from it? After decades of innovation on the cost-control front - capitation, pay for performance, utilization review, etc - perhaps one of the simplest methods is to keep things streamlined is to educate physicians about how much their actions cost. At the very least, making costs and prices more transparent to providers might serve as an effective complement to the myriad of other strategies we (quixotically?) employ to achieve cost-control. In any event, I'd definitely like to see more research in this area.