View: Sticker shock in preview of Obamacare rate hikes

That's only 12 states, of course, and none of the biggest-population ones. But even if we assume that the regulators cut the increases in half, that's a median increase of 5-6 percent, with a mean considerably higher than that. Even if you weight by population — well, actually population-weighting makes that worse, not better, because the states with the lowest rate requests are disproportionately sparse.

Moreover, significant rate increases are what I would broadly expect, because these rates are the first ones set with a full year of claims data, and what we know about the pool is that it is poorer and older — which would also mean sicker — than was projected. Initially, HHS was saying that it needed about 40 percent of the exchange policies to be purchased by people age 18-35 to keep the exchanges financially stable. It was 28 percent in both 2014 and 2015, according to HHS data. The CBO had projected about 85 percent of exchange enrollees to be subsidized, falling toward 80 percent as enrollment grew; instead, that number is 87 percent and actually rose slightly from 2014. It would be pretty surprising if rates weren't increasing faster than inflation, or even than general health care cost inflation.

Eyeing the Journal's list, the most obvious pattern is that states are converging on a price somewhere well north of $300 a month for a 40-year-old nonsmoker seeking a Silver plan; the states with the biggest rate hikes all had premiums under $250, and are asking to be allowed to go near or over $300, while the states that asked for low increases were already over $300, and in some cases well over. (Vermont is at $430 — and asking to go to $476! "Only" an 8.4 percent increase, but wow.) It seems as if states where insurers initially underpriced are now trying to move toward a natural price somewhere between $3,600 and $5,000 a year for a single nonsmoker. If that's the price of providing basic benefits, regulators cannot command it away by fiat; the best they can do is to force insurers out of the market.

I assume that these large insurers are willing to incur some losses in the market for exchange policies in order to stay on the good side of their state regulators and HHS, because overall, those policies are not a large part of their business. But getting those rates down to something more on the order of 10 percent would require some pretty big losses. How long, exactly, will they be willing to carry a product that loses that kind of money?

The good news is that even if we do see big rate hikes for the next few years, that doesn't mean we need expect them indefinitely. Eventually, insurers will figure out the price of providing these products, and then — barring a self-selecting "death spiral" — cost increases will move with the rate of health care cost inflation, rather than wildly gyrating as insurers realize they're losing money. The bad news, of course, is that we don't know how many big increases we might need to get to that price.

See also: Assurant confident about benefits unit sale