Can CMS be a venture capitalist?

Lisa Suennen, a venture capitalist, writes this post about the provision in the national health care reform act that created the Center for Medicare and Medicaid Innovation (CMI). This agency has $10 billion to “research, develop, test and expand innovative payment and service delivery models that will improve the quality and reduce the costs of care" for patients covered by CMS-related programs. Lisa notes, "What is great about CMI is that they have the authority to run their programs much more like a business would without many historical governmental constraints. "

I don't want to be a stick in the mud, particularly as my able friend Don Berwick takes charge of CMS, but I want to point out that previous efforts by the government to be innovative in other fields have failed because:

(1) Venture funding embodies risk-taking. Government usually does not do this because there is a political imperative never to be blamed for misspending taxpayer money. The bureaucracy, therefore, systematically eliminates ideas that are untested.

(2) Alternatively, the leaders of such agencies get seduced by good-sounding ideas that have not been able to meet a market test and whose efficacy is subject to the normal variations in markets. Then, they will persist with unsound investments because the concept of sunk costs is politically difficult to acknowledge. Private firms consider past investments as fiscally "gone" for purposes of evaluating future investments. But the government often behaves as if investments have a carry-forward risk of embarrassment, so it is more likely to throw good money after bad.

(3) Worse yet, agencies are encouraged to invest in those ideas that have political connections.

My favorite example of all of this was the Synthetic Fuels Corporation, created by Congress to move research and development of synthetic fuels out of the Department of Energy and into a public-private partnership that would hurry along new energy sources. And here is another more recent Massachusetts case. Let's hope these examples are inapposite to the vision for CMI, but history does provide warnings.