Will health care reform raise premiums?

Debate continues over what will happen to health insurance premiums if something like the Senate bill becomes law. One rather frustrating strand of the controversy centres on what the CBO's recent letter to Evan Bayh on the subject actually means--which doesn't say much for its clarity. Mickey Kaus took on Ezra Klein and others over this, Kaus arguing that the CBO report says premiums will go up, Klein arguing the opposite. I'm with Klein, though I think it's something of a phony argument. What both sides are saying is true, just not the whole truth. The squabble is easily resolved by ignoring Kaus and Klein and just looking at what the CBO says.

After I've done that, I'll turn to a more important issue. Is the CBO right?

Look at table 1 on page 5 of the CBO's note. It plainly says, first, that most Americans will see their insurance costs go down at least a little. It is not "insane", as Kaus charges, to regard that as the headline.

Most Americans are in group insurance schemes--either small group or large group. The change in premiums for these people will be minor (between plus 1% and minus 2% on average for small group; 0 to minus 3% on average for large group). In addition, 12% of small-group customers would get subsidies to help pay for their insurance under the bill; they would be significantly better off.

The position in the individual market is more complicated. There, the CBO says, premiums would rise substantially--by between 10% and 13%. However, as the table shows, this is because the rules would oblige people who are already covered to buy better insurance, on average, than they do now. (Premiums for the same policy, if you were allowed to buy it, would fall by 14% to 20%.) In other words, outlays on insurance would go up, but premiums per unit of coverage would go down. In addition, 57% of individual customers would receive subsidies. Their premiums, after the subsidy, would fall by nearly 60%--and of course by even more, on a per unit of coverage basis. 

Very well. The disagreement between Klein and Kaus is thus about semantics.

(1) What has happened to the price of potatoes if the store charges an extra 10 cents a pound and the government pays a subsidy of 20 cents a pound? Kaus is unhappy if you say the price has gone down. I am fine with saying that--so long as you make clear that it has gone down only for people receiving the subsidy, and not everyone will be so lucky.

(2) What has happened to the price of potatoes if the store says you must spend 10 cents more than you did last week (whether you like it or not), but receive in return an extra 20 cents' worth of potatoes? Again, I'd say the price has gone down--though if that is all you say, you are not telling the whole story. 

As I say, mostly semantics. Much more important is to think about the substance of the CBO's findings. This new report by the consultants Oliver Wyman says the CBO is wrong and that premiums would rise a lot. It was paid for Blue Cross and Blue Shield, and so was reflexively trashed as biased (including by the White House, which called it another sham report). I don't know whether the findings are right, but the analysts emphasise an important point. It is one previously stressed by Martin Feldstein. It's one I've mentioned too--in this FT column on health care reform, and in this blog post (scroll down to the section on Massachusetts at the end). 

The issue is adverse selection. If you have guaranteed access to health insurance regardless of pre-existing conditions (which the bill would provide), why buy insurance at all until you actually need it? Healthy people will opt out, and the per-person cost of servicing the remaining customers will go up. If that happens in a big way, premiums will soar. The proposed solution is subsidies and the mandate, to pull healthy people in and assure the widest possible risk pool. But are the subsidies going to be high enough, and the penalties for defying the mandate severe enough? The CBO implicitly says yes: it expects adverse selection to pose no great problem. Oliver Wyman disagrees: it expects adverse selection to play a key role.

The Senate bill's subsidies are pretty generous; but not everybody gets them. On the other hand, the penalties under the mandate are mild. Neither group of analysts has explained its reasoning in detail--though Oliver Wyman points to the unhappy experience of individual states that have combined guaranteed issue and rating reform with weak coverage requirements. Those cases are worrying. I'd like to know what the CBO thinks of this study.