Figure 2 illustrates excess growth in real Medicare spending on physician and clinical services. During the 1970s and 1980s, excess growth was quite high, generally ranging between 4 percent and 8 percent. Beginning around 1990, excess growth moderated somewhat and since then has generally been between 0 percent and 4 percent. There was a substantial uptick in excess growth in 1999 through 2003.
This is from a working paper, not from an official CBO document. His point is that the cost control methods for Medicare enacted in 1997 worked (he also documents other periods where cost control worked). I was surprised that this was effective. Of course, one can still be skeptical that any long-term reduction in health care spending under Medicare can be achieved.
Yesterday, some folks at CBO invited me to lunch to talk about health care. The thing is, they already know a lot. Typically, I'll have a semi-educated guess about something, and they'll have an 80-page background paper on it. Estimating the impact of medical tort reform on health care spending, for example; or estimating the impact of consumer-directed health plans on health care spending.
One question that stumped me was, "How do you explain the regional variation in usage of medical services?" I was familiar with the finding, by Dartmouth's John Wennberg and crew, that even after controlling for patient condition you see a lot more services used in some regions than in others.
It's hard to come up with a good theory to explain this variation. For example, suppose you think it's supply driven--if there are twice as many doctors in a region, then there will twice as many procedures. How exactly do the doctors in the high-supply region manage to keep their calendars full?