Did I Accidentally Prove that Obamacare is a Success?

New
York’s Jonathan Chait
needles me for having “accidentally” shown “how Obamacare is
succeeding.” To make his point, Chait summarizes and briefly quotes
several posts I wrote between January and the second week of March
of this year noting low or overstated enrollment figures in various
parts of Obamacare, and then jumps forward to a post I wrote
earlier this week arguing that a recent estimate combining all of
Obamacare’s coverage expansion provisions is almost certainly too
high. 

“We have gone from learning that the law has failed to cover
anybody to learning it would cover a couple million to learning it
would cover a few million to learning that it has probably insured
fewer than 20 million people halfway through year one,” he
concludes, as if the items he links to are all discussing
comparable components of the law’s coverage expansion.

But Chait’s argument relies on on selective quotations from
posts of mine that examine different aspects of the law that aren’t
really comparable, stringing them together in a way that implies
they are all referring to the same thing. And he fails
to mention other posts I wrote, between March and July, noting the
law’s late-March enrollment surge and effects on covering the
uninsured.

Mostly, then, his item serves to demonstrate is that enrollment
in Obamacare’s insurance exchanges was, as was widely reported,
sluggish from the beginning of the year until the middle of
March, and that it grew afterwards. (Until the very end, it lagged
behind the administration’s own projections.)

Throughout the year, I noted the low enrollment figures as they
came in, arguing that the numbers weren’t promising. And when the
enrollment did pick up at the end of March, I also wrote items on
the surge and its effects on multiple occasions—including a March
31
post noting a “last-minute sign-up surge,” an April 1
post noting the White House’s announcement that 7 million had
signed up for coverage, an April 17
post noting the administration’s report that 8 million had
signed up for coverage, and a May post
specifically saying that early speculation that the law might have
no or negative effect on coverage was not plausible.

Chait does not mention any of these items in his post, which
does not quote or link to any post I wrote between March 11 and
this month. He also says my posts are useful as a “lagging
indicator of Obamacare’s progress.” But all of these posts were
written within a day of the information becoming public.

Chait also fails to quote my own caveats in the posts he does
cite, or to provide important context for comparing the posts he
quotes. For example, he quotes me
saying on January 21 that “it appears possible that there has
been no net expansion of private coverage at all.” Note that this
is presented as a possibility, not a certainty, and that I also
wrote in the same post that that even if this were true, “there’s
still time for that to change. As the administration is keen to
remind us, people who want coverage have until the end of March to
sign up for coverage this year.”

The
March 11 post Chait quotes from, meanwhile, dealt specifically
with the question of how much effect Obamacare was having on the
uninsured rate, based on a Gallup survey, which turns out to be a
completely different metric from what’s being looked at in the next
post he selects.

The Gallup survey, I said, was the “best evidence” that
Obamacare was reducing the rate of the uninsured. This effect,
which was measured prior to the late-March surge, is separate from
the question of how many people now have coverage through some
coverage-expanding provision in the law. But the total-coverage
question is what the next post he quotes from (a July
item taking issue with a New England Journal of
Medicine (NEJM) study estimating 20 million total covered
by Obamacare, through the exchanges, Medicaid, and other
provisions) discussed.

The authors of that study specifically note that, unlike the
Gallup survey, it was not an attempt to judge the law’s effect on
the uninsurance rate. As the authors of the study say, “We do not
know yet exactly how many of these people were previously
uninsured…”

Chait pairs the two posts without much context, implicitly
suggesting that they are looking at the same thing. They are
not.

The July post on the NEJM‘s total-coverage estimate is
also not really comparable with the
the February 24 post that Chait quotes from, which specifically
deals with President Obama’s claim that almost 7 million people had
gained coverage under Obamacare just through the law’s Medicaid
expansion.

Obama’s claim was wrong at the time. It’s still wrong. I was not
the only journalist to say so at the time. (The Washington
Post’s fact checker
gave Obama’s statement four Pinocchios.) By now even the
administration has backed off that number for the Medicaid
expansion, saying only that total enrollment in the program (which
additionally counts the surge month after Obama made the 7 million
claim) has increased by 6 million following Obamacare’s coverage
expansion, and that
not all of that increase is directly attributable to the law.
If you read that post on the day that it was published, you got an
accurate and real-time impression that President Obama was wrong.
If you read it now, you get an accurate impression that Obama was
overstating the health law’s coverage effects.

Obamacare has, without question, enrolled far more people since
January, and the evidence is pretty strong that it has cut the rate
of the uninsured by several points. But even as the coverage
figures have increased, what you also see is that the president,
his administration, and the law’s backers have consistently relied
on dubious, misleading, or incomplete metrics to overstate what can
be known about the law’s impact based on the information
available.

Chait may see my posts noting this as inadvertent proof that
Obamacare is actually a success, but to me it looks more like
evidence that the administration and its supporters are desperate
to convince people that it is more of one than the evidence
supports.