I have previously touched on the issue of a physician shortage here, here (in the context of concierge physicians), here (referring to the distribution of risk), here (in the context of mid-levels) and here (in discussing an influx of foreign physicians) . I am aware of literature that suggests otherwise, that there are too many physicians, who can generate their own demand.
Maggie Mahar makes hay commenting on Buckeye Surgeon's post and The Atlantic's article suggesting there is an oversupply.
The basic premise emerging should be that we are oversupplied with specialists and face a shortage of primary care. If a physician emerges from medical school with a heavy debt and the general perception that primary care physicians are somehow "less than," there is little incentive to dedicate oneself to a career reducing society's overall health care costs. Where's the reward?
Let's take it a step further, what happens when the primary care physician is overworked due to a local shortage? Chances are she'll just refer faster to a specialist who can take the time to figure things out. A referral is fast, easy and an efficient use of her scarce resources in the face of poor remuneration. At least the specialist will be well-paid for an extensive investigation, even though it may not really be strictly required. To take care of it herself would mean an extensive investment of time building the confidence of the patient, and truly assessing the risk of not being sued for non-diagnosis.
If she weren't very busy and really needed some extra income, she would probably learn a procedure like hemorrhoidectomy or laser skin dermabrasion or whatever is the rage, in order to keep up with the Joneses.
Now if a specialist is busy, he'll just do a few extra investigations, get the residents involved and above all, take no chances. It is the purview of the specialty not to miss any pathology, no matter how rare. Testing and further referral as necessary is the only way to go, especially in a litigious atmosphere.
If a specialist is not sufficiently busy, he will ensure that the maximal use of an appropriate package of investigations for the justifiable conditions. One urologist told me the proper response to any patient with prostate symptoms was an IVP, flow studies and a cystoscopy. I know his practice was sluggish, and he desperately needed to be busier, especially given the rent in the hospital's new medical office building.
These vignettes are not universal examples of every physician's behavior, but they illustrate the incentives in different settings. The one situation that does not drive health care costs up is a moderate level of work distributed to a reasonable number of primary care providers.
The sad fact is also, that covering the majority of everyone's health care costs will drive up demand to the extent that patients perceive such care as affordable. There is no cost control possible with expanded coverage.