Wednesday, March 27, 2013

Manufacturing New ACA Myths

Here's how myths get started.

Over the weekend, Alyene Senger at Heritage ran an item called "Obamacare at Three Years: Increasing Cost Estimates." Her claim:
Over the last three years, the Congressional Budget Office (CBO) has revised its cost estimates for Obamacare’s new entitlements—the Medicaid expansion and exchange subsidies—many times, and they have more than doubled since 2010.
And she has a chart showing estimated new costs rising from $898 billion in March 2010 to $1.6 trillion in February 2013. The chart is titled "Obamacare's New Spending Estimates Keep Rising," and includes text saying "The Congressional Budget Office as made several estimates of the 10-year cost of Obamacare's new spending on the Medicaid expansion and exchange subsidies, and the costs keep rising."

The chart is important, because it's easily exportable. So David Frum today ran an item which was just the chart; the only thing Frum added was a one-sentence introduction, "The estimated costs of Obamacare keep rising."

I haven't looked, but I'm sure the chart will get plenty of play.

Except...it's entirely phony. The real story from the CBO charts she cites says the estimates haven't really changed at all.

Senger's own article undermines, to some extent, the story the chart tells. As she (quite honestly) points out, the March 2010 estimate was a ten-year projection in which the ACA only kicks in for the last six years. For the February 2013 estimate, which she again (very honestly) points out, is an eleven, not a ten, year estimate, the number she uses for the chart subtracts out the 11th year (2023). So it's nine years (plus one pre-expansion) compared with six (plus four pre-expansion).

But there's more! Neither the subsidies or Medicaid is projected to kick in all at once. In the original March 2010 estimate, for example, subsidies were expected to grow from $15B in 2014 up to $75B in 2017, and then increase more slowly ($82B in 2018, $87B in 2019). So the original estimate wasn't really for 6 full years of ACA; it was really four years before costs kick in, three partial years, and three full years. The February 2013 estimates, then, are for one pre-ACA year, three partial years, and six full years.

What we really need is year-to-year comparisons of the coverage expenditure estimates. These appear to be the numbers Senger is using from the CBO tables:

Year     2010 est     2013 est
2014         $49B        $47B      
2015          100           96
2016          158         153
2017          184         186
2018          197         206
2019          210         217

Um...that's not doubling. Indeed, since I assume that the 2010 estimate is in 2010 dollars while the 2013 estimate is in 2013 dollars that the small increase is pretty much all inflation-related. Indeed: the CBO reports also list four categories of payments (including for example penalties for individuals who don't have insurance), and the estimated offset from those goes from $42B in the 2010 estimate for 2019 to $68B in the 2013 estimate for 2019 (although some of that may be that the revenue effects were changed by other tax changes). By the way, that means that the net cost of coverage expansion now appears to be lower in 2019 than it appeared three years ago.

Now, if you want to use these numbers to say that keeping the original cost estimate under $1T was phony...I'll be right there with you. It was. Once fully phased in, the ACA has costs far greater than $100B per year. It pays for them, but yes, that number was phony.

But if you want to use these CBO reports to claim that "the costs keep rising"....well, that's not what CBO says. A better description of the data would be "the cost estimates have not changed at all over the last three years."

So everyone who runs Sanger's chart should know: it's massively misleading. Or, I should say: the bar graph is misleading; the text is just plain dead wrong. All in all, it's garbage.

No comments:

Post a Comment