Baumol's Cost Disease

Moneybox:

The most important disagreement between Romney and Obama that keeps not getting covered properly is the dispute over Medicaid. The Obama administration wants to spend a lot of new money to provide Medicaid coverage to a much larger set of the working poor. Romney, by contrast, wants to spend drastically less money on the program... Notionally, Romney's idea is that by not merely implementing huge cuts in funding levels but also offering states more flexibility about the rules that he'll be able to usher in a bright new era of innovation.
In the real world, what happens if you're given less money to cover people with and more flexbility about what to do with it what you're going to do is simply cover fewer people. In particular, care for the elderly and disabled are likely to get hammered over the long-term thanks to productivity issues. Medical technology does advance and that sometimes leads to great new things—pills that treat illnessness, machines that help us test for disease—but the foundation of long-term care for the elderly and the disabled is human attention. People who need help caring for themselves on a day-to-day basis need help from other human beings. If you take the main public program for financing that kind of care and insist on its budget shrinking as a share of GDP, then the people who need help are out of luck. As economy-wide productivity rises, it'll get harder and harder to "afford" programs that provide labor intensive care even in the real world society will be richer than ever.

 Not only will labor-intensive human care get relatively expensive as productivity increases elsewhere, but Medicaid administrators will be increasingly tempted to shift resources away from labor-intensive services and towards things like cheap new drugs that do see costs declining as technology advances. A similar dynamic is happening with military spending:

Many... are upset that military spending as a share of U.S. GDP is far lower today than it was during the Cold War era. A conventional... reply is to note that the objective threat environment is very different. But perhaps a better way of looking at this issue is through the lens of productivity. If you have an industry where productivity advances slower than average then spending as a share of GDP needs to rise to keep quality constant. ...In the military sector you see some interesting dynamics. Thanks to technology, we're way better at a task like "make this guy's house explode" than we used to be. First airplanes and then precision munitions and now unnmanned aerial vehicles have made this vastly cheaper and easier than it was in 1912.

But for other kinds of tasks, technology progress hasn't helped much. In particular "boots on the ground" occupation-and-administration of foreign territory suffers from a similar [labor-intensive] dynamic as kindergarden. ...If we want to maintain a roughly constant level of boots on the ground operations then spending as a share of the economy has to steadily rise, while if we're content to explode things from afar it can steadily fall.

Which is perhaps a roundabout way of saying that on both the military and civilian sides of the government, I think there's too much talk in D.C. about "budget math" and not enough talk about what specifically we're trying to accomplish and why.