A Growing Proliferation of Managers

A while back, we had a dialog with EconBlog about the myth of US health care waste. One issue I had discussed was adminstrative overhead. (See my first previous post here, and follow-up here., and on cost of high-technology, here.)
I argued that my experience as a physician (and discussion with other physicians) suggests that there is a huge administrative and bureaucratic load on physicians, and that this contributes directly and indirectly to health care costs. The best I could do at the time was to cite a study that showed that physicians in practice spend an inordinate amount (a little less than US $25K a year per physician) on "unnecessarily complex or redundant administrative tasks."
I just found some fascinating data along these lines, available from the Center for Medicare and Medicaid Services (CMS) in a series of charts (here, chart 1.13.)
See in particular Table 1.13, Health Care Employment by Occupation.
It shows that from 1983 to 2000, the numbers of health care managers grew at a rate that far outstripped any other kind of health occupation. Taking the numbers off the PowerPoint presentation,

• The number of managers grew from 91,ooo in 1983 to 174,000 in 1990, to 752,000 in 2000.
• That could be compared with the numbers of physicians in those years (519,000 to 577,000 to 719,000), and the number of registered nurses (1,372,000 to 1,667,000 to 2,111,000)

So the growth rates from 1983 to 2000 were 1.39x (39%) for physicians, 1.54x (54%) for nurses, and a whopping 8.26x (726%) for managers.
Another way to look at it is, in 1983 there was 1 manager for every 5.7 physicians and every 15.1 nurses. In 2000, there was 1 manager for every 0.96 physicians and every 2.9 nurses. Again, by 2000, the number of health care managers exceeded the number of physicians. There were more managers than any other species of health care worker other than nurses.
If health care could function in 1983 with one manager for nearly every 6 doctors, why in the world did we need one manager per doctor in 2000?
I would love hear if anyone can come up with a justification for this massive increase in numbers, or show how this proliferation has lead to any improvement in health care.
On the other hand, ecological correlations are not a good way to prove causation, of course, but I would argue that this data suggests that attributing the simultaneous rise in health care costs, decrease in access, stagnation in quality, and dissatisfaction of health professionals like physicians and nurses to the incredible proliferation of managers (and attendant bureaucracy) is not far-fetched.