Here is David Cutler over at Cato Unbound:
"We observe that areas that spend 30 percent more than other areas do not get better outcomes. It is possible that randomly cutting 30 percent of spending in high spending areas would not affect outcomes. But this is not likely. Cutting spending by 30 percent would almost surely eliminate some valuable services as well as some less valuable ones. One could make up for the loss of valuable services by providing other services that are low cost but not currently provided. Overall health might not be affected, but money would be saved. Better still would be to selectively eliminate the care that has little value and provide the other services that are valuable but are not currently provided. This would leave us spending less and with better health. I believe we can do this, but the task is harder than it seems at first pass. We don’t do ourselves any favors by pretending it is easy."