Not sure if I should refer to Vox as journalism; their attempt at understanding the per-capita Part A spending is more along the lines of humor.
But what's definitely clear — and what's driving this trend — is that Medicare is spending significantly less per person than they did two years ago. And this report expects that trend to continue for another two years going forward.

Ms Kliff has a number of guesses of why this is, including Obamacare of course. What she misses is the most obvious one: Medicare growth usually happens when someone turns 65. And 65 year-olds have lower claims cost then 80 year-olds. If you add 1.6 million people at the healthiest end of your spectrum, and the costliest end of your spectrum dies off, you would expect average claims to drop. When that growth is higher then usual, i.e. all the baby boomers, you would expect sustained reduction in per capita cost.

They had 3% growth to demographics in the most favorable spectrum possible. A 2% reduction in cost for that favorable of a 3% change in demographics is not impressive. Looking at CMS's website and their press release, it appears they don't understand the actuarial sciences either.