• This doesn’t help

    A few people sent me links to an AP story fact checking Nancy-Ann DeParle’s blog about the report from last week on the PPACA.  So I went straight to the source and read the post.  Here’s the questionable bit:

    Today’s report by the Office of the Actuary confirms a central point of the Affordable Care Act passed by Congress and signed by President Obama:  The Act will make health care more affordable for Americans. In fact, the Actuary’s report indicates that total health care spending per insured American will be more than $1000 lower thanks to the provisions of the new law than it would have been if Congress and the President had not acted.

    What is “spending per insured American”?  I’ve seen spending per capita.  I’ve seen spending as a percentage of GDP.  But when I google “spending per insured American”, there are an amazingly small number of hits.

    I get why Director DeParle might want to use this new statistic.  You see, total spending is going up. That means total spending per capita is going up, too.  So the only way to make it look like costs are coming down is to increase the number of people you’re dividing costs by.  You can do that by making the denominator the number of insured.

    Think about it – costs are going up, but if we spread those over the 32 million newly insured, it looks like spending is coming down.  “Spending per insured American” that is.

    This is only important because someone made Americans think that the bill was somehow going to save money.  Uh, no.  It might reduce the deficit, but I don’t know anyone serious who thought it was going to result in our spending less money.

    I saw the report last week, and noted that while PPACA might cost us money, it seemed like a decent bargain to get 32 million more people insured.  I think there are better ways to do it, more cost-effective ways, but it was actually better than I expected.  There were legitimate, and intellectually consistent, ways to acknowledge the findings of the report and defend them.

    You don’t need to invent new statistics.  That makes it look like you’re hiding something.  It weakens your position.  And it makes what could have been good news into a bad story

    • Isn’t that just a way of saying that the rate of insurance coverage will increase faster than the cost of insurance/healthcare?

    • emdash,

      There are lots of ways to talk about this that make sense. I’m not even sure that’s correct, though. Those are two separate things.

      For instance, the latest KFF survey showed that employer based insurance is getting more expensive. I don’t blame that on PPACA, nor do I think PPACA is improving that. I’m nost going to waste time crediting or blaming PPACA for that, and that group accounts for most insured Americans.

      Bottom line is that if you’re trying to spin this, you can get inventive, probably without having to invent new statistics. But it’s probably more straightforward to acknowledge (as the President did) that no one expected to insure 32 million more Americans and have spending go DOWN. The best hope is for it to go up as little as possible.

    • Oh, sure. I happen to agree with you. I was just trying to come up with a charitable interpretation.

    • Aaron –
      I’m bit behind on my reading, but I have to profoundly disagree with you here. I think Ms. DeParle’s measure, spending per insured, is the appropriate one if we’re interested in measuring affordability of coverage. I believe most of the health spending in this country comes from those with insurance, whether it be a private or public payer. Typical insurance policies have an actuarial value between 70 and 90%. That is to say, they pay that percentage of beneficiaries health costs. So, while we might measure the insurance people are getting access to in terms of actuarial value, a shorthand might be that people are getting insurance.

      Then the question is how much does that insurance we’re buying cost. One way to think of this is if we put everyone with insurance into a pool and divided up the costs. Would the cost of insurance go down with health care reform? Yes, says DeParle. DeParle’s measure is a way of quantifying the cost while controlling for the extent to which people suddenly get access to care. And if the service we’re measuring the cost efficiency of is access care, it’s misleading to include those who don’t have access (i.e. the uninsured) in a measure of cost efficiency.

      A few caveats. Obviously we’re simplifying to say that uninsureds don’t access care. With the single insurance pool example, that has to be invented because we have different insurance pools (this might not be the case with single payer). Also, this is for a very different conversation that talking about national health expenditures. This is only appropriate in a conversation about efficiency, not about total expenditure. An invented analogue might be seen in electrification from the 1880s until 1940. Imagine we had more expenditures on electricity by providing electricity to all, but the efficiencies of larger power plants serving entire regions permitted the cost per person receiving electricity to actually go down.

    • Ok, sorry, I just understood the AP article. It implies that DeParle included uninsureds in the denominator but not in the numerator, i.e. it took total national health expenditures for everyone and divided them by the number of insured people. That would be inappropriate. From the article, I’m not sure she did that and AP certainly didn’t correct for it. The appropriate measure, as I argued above, would be cost for the insured population divided by number of insureds. You can take the difference to of it after vs. before health care reform to tell you if you’re able to get some efficiencies of scale.

    • GrandArch

      Actually, I think she did just that – adjusting the denominator. That’s my issue with it.