Yesterday an article in the Times profiled new medical schools around the country. Dr. David Goodman at Dartmouth is quoted as pointing out that “When you add more physicians to an area, they just add more services, and their salaries don’t go down anywhere near in proportion to the increased supply.”
It would be interesting to know what that proportion actually is, and I'm no expert in this subfield. Shifting out the supply of doctors along the demand curve should in principal lower doctors' wages, unless the demand curve shifts too, but it clearly depends on the slope/elasticity of demand.