Administration officials and many others on the left who talk about slowing health costs in the coming years never really attribute that expectation in any concrete way to the new law. Rather, they point to the fact that the growth of health costs has slowed a bit during the recession and the painfully slow recovery of the past few years, and they simply expect that slow rate to continue even as they simultaneously expect the economy to recover much more robustly in the coming years. …

Until this year, the CBO has always assumed that these families just wouldn’t drop their coverage, but in its latest score of Obamacare, the agency for the first time projects that the number of people in the exchanges will actually begin to drop after 2018, declining by almost a tenth over the subsequent five years even as the population grows. And since the people who remained in the exchanges would tend to be poorer and sicker, the costs of providing them subsidies would grow very quickly (by almost 6 percent annually), since the exchange pool would become more risky. (And this projection, remember, is still based on rosy expectations about overall health-cost growth.) This nightmare scenario, too, is pretty unlikely to happen, since the people involved would be middle-class families. They’re not going to accept the enormous downside of Obamacare without even the modest upside of exchange subsidies, and they’re not going to like being forced to go uninsured. The politics of this just wouldn’t hold.

In both cases, it is only possible to imagine that Obamacare might be sustained if we assume very low growth in health costs. That assumption is absolutely critical to liberal fiscal and health policy today. But of course, Obamacare doesn’t really offer any serious mechanism to achieve such low costs — in fact, it’s actively hostile to the kind of consumer incentives and competitive pressures it would take to achieve it.