That is the question asked in a new working paper by David Cutler, nicely summarized here in the NBER's Bulletin on Health and Aging. This assessment is consistent with just about everything I have observed in health care markets:
The obvious question is why the health care market has not evolved to become more efficient. To answer this, Cutler explores the incentives for providers to make investments that will improve quality. He notes that providers cannot charge a higher price for providing higher quality care because prices are typically fixed (for example, by Medicare). Providing higher quality care might result in a greater volume of patients if consumers had ready access to information on quality, but generally this information is not available. In fact, providing higher quality care may lead to lower patient volumes and revenues, since higher quality care may mean patients need fewer services and providers are often reimbursed a fixed amount for each service.
Cutler also examines why insurers and other payers do not impose changes that would improve productivity, such as requiring providers to use electronic medical records. One issue is that any individual payer has both limited ability to change provider behavior on its own and limited incentive to do so because it will not reap the full benefits. A lack of public information on insurer quality (which greatly reduces the insurer’s ability to benefit from quality improvements) and turnover in a plan’s patient base may also create weak incentives for insurers to improve quality.
To this assessment, I would add the following stream of consciousness: Why be an entrepreneur if you cannot attract new customers? You cannot attract new customers unless they have meaningful choice. Meaningful choice only happens when you can have new producers enter the market. New producers are often small producers. Can you be a small insurance company? Not unless you are protected from adverse selection. Could the government do this? Yes. Is it doing it? Not yet. Will it be doing it? Sort of. Read more here.