Some quick responses:
1. Competitive barriers: Allowing for full interstate competition is just one example. Differing tax treatment (individual purchasers vs. employer purchases) is another. Rules that preclude multistate pools, along with other restrictions on pooling is another.
2. The findings of empirical research should be applied to determine whether licenses in various medical professions actually add value or restrict supply (number of practitioners). My guess is that there is a mixed situation there. Some licenses make sense e.g., physicians. Others don't.
3. Empirical literature on disruptive technologies has shown that at the very early stage, the cost of such technologies is high and the benefits vis-a-vis existing technologies are small, even somewhat negative. Medical technology is just one subset. Accelerated depreciation rules for tax purposes create an incentive for firms to providers to be the first to deploy new technologies (at the least efficient/effective end of the curve: least value added per dollar of cost). Payment practices may also contribute to such an outcome. Health services, like any other service, is confined by constraints. Tax code and payment practices should be adjusted to consider the full macroeconomic picture involved with technology purchases and other practices. Indeed, were robust productivity measures required and regularly reported, one would get a much better understanding where value is added and where it is not. For now, the BLS has no such measures so a large part of the health system operates in the "dark" with respect to understanding productivity. In that context, the existence of cost issues is not at all surprising.
4. Basically, consumers are not permitted to reimport pharmaceutical products. That amounts to legal protection of price arbitrage situations. For example, if gold were selling for $1,130 per ounce in London and $1,140 per ounce in New York, investors would buy gold in London (pushing up the price there) and sell it in New York (pushing the price down in NY) and the price disparity would be eliminated. Barring drug reimportation perpetuates a situation where American consumers cannot eliminate price premiums that might exist for some medical products. As a result, it builds in a degree of extra expenditures on such products that might not exist in a competitive market.
5. Currently, it is very difficult for foreign doctors to practice medicine in the U.S. The H-1 visa rules should allow for a much larger number to enter the U.S. Unless there is statistically significant empirical data that requires additional licensing for such doctors (who are licensed elsewhere e.g., in the EU) to practice in the U.S., those professional entry barriers should be eliminated. That is just one example of barriers that make it difficult for foreign providers to aggressively expand in the U.S. One has to recall that in many countries, the costs of foreign medical staff are less than those of their American counterparts, sometimes quite substantially. Hence, again one finds a de facto protection of wage premiums that might otherwise disappear in a competitive market. In the long-term, growth in consumption of medical services is projected to exceed growth in providers. An aggressive push to increase the supply of providers would alleviate the cost premiums attributed to a demand imbalance.
6. The Bureau of Labor Statistics would need to work with the health industry to construct such indexes and their components. Research would be needed so as to ensure that the measures are representative and reliable. There is no intellectual reason why such measures are not possible for the health industry. The task would involve identifying outcomes, matching them with costs, addressing any cyclicality that might exist (probably modest relative to most industries), and standardizing reporting. Under such a framework, BLS would report a variety of productivity measures for the health sector (hospitals, physicians, medical products, etc., essentially the same categories that are used with respect to the CPI). The providers would be required to publicly disclose their productivity data (productivity number and constituent components).
