Joe Paduda writes:

It appears the answer lies in declining reimbursement rates. These hard-working docs are spending plenty of time (over 45 hours a week) with patients, but their reimbursement rates have not kept pace with inflation. For example, Medicare has increased fees by 13% during the study period, while the underlying inflation was 21%. And, private payers’ reimbursement declined from 143% of Medicare’s rate in 1997 to 123% in 2003.

So, clearly physician income is not a driver of medical inflation. One driver appears to be the increased volume of tests performed; utilization in this area was up at a 6% annual rate over the study period.

But the real driver appears to be higher utilization of physician services (more docs doing more stuff), and, slightly less important, a significant increase in hospital and facility costs.

Like I wrote in a comment thread below, there is zero disincentive to overtest.