Austin Frakt has a very smart post up about what it would take for him to support a vouchered approach to Medicare. The most important of these is that there would have to be an actual competitive bidding process run as a pilot to see if competition could actually reduce the cost per beneficiary (risk adjusted). In other words, there has to be some evidence that a private bidding process would actually reduce costs, when trying to provide the same benefit package.
The current Medicare Advantage plan is not this, and I am not a big fan as I have written. MA has been a complete failure as a way to 'save money' for the past 30 years, mostly because of cherry picking, but also because of politically driven increases in premiums amounts to be able to say more beneficiciaries were choosing private plans.
Austin's suggestions for what would be required to 'be for' a vouchered approach to Medicare make sense, and least in moving that direction. I am still not sure it is possible. I wonder if you could actually make a private voucher approach work if you didn't somehow have entire geographic areas included in the vouchering? If not, you are always so worried about cherry picking and adjusting premiums either before or after the fact. The true test of competitive bidding is to show that it can lower per capita costs in setting in which everyone in a geographical area has their care financed in that manner.