WASHINGTON — The big news this week turned out not to be a health-care vote on Thursday or Friday, but the concession by GOP Senate leadership on Tuesday that they don’t yet have the votes for the bill and won’t bring it to the floor before the July 4 recess. It’s not exactly the news I hoped to be reporting upon getting to town today, but being here has been worthwhile all the same.
For example: In a scrum with reporters right after the caucus luncheon today, Sen. Marco Rubio of Florida pointed to a new criticism of the Congressional Budget Office’s estimate of the Senate bill that I’d overlooked:
“The one place I think everybody feels like we need to know more — look, CBO just told us, the baseline they’re basing these coverage losses off (is) a baseline that isn’t real. They’re basing it off 2016 numbers that in some cases are off by 200 percent. Five million of the 22 million of losses are people they think might be added to the Medicaid rolls by new states that decide to expand in the future. But on, basically, guesswork, they’ve admitted.
“I think the first thing we need is an actual CBO score on what the actual insurance marketplace looks like right now. Because the baseline they’re using to compare this bill to is March 2016 numbers that are way obsolete at this point.”
You may recall I’ve previously explained the CBO’s estimate of an Obamacare repeal-and-replace bill was bound to be fraught with error because it uses as a baseline its projections of what it thinks would happen in the future under Obamacare, and compares any new bill to those. The problem is that the CBO has consistently overestimated how many people would buy insurance on the Obamacare exchanges, but it hasn’t lowered its projections for future years to take account of its past mistakes. Until it does that, it is bound to overestimate — perhaps by many millions — the number of people who would “lose” insurance under a reform. (I’m not going to rehash the entire argument here, so if you haven’t read it please click here.)
What I didn’t catch in my initial perusal of the latest CBO report was this paragraph, which may have been what Rubio was talking about:
“In the projections published in January 2017, the direct spending and revenue effects of the ACA’s insurance coverage provisions and the total number of people projected to be uninsured were similar to those in the March 2016 baseline, but the number of people projected to purchase subsidized coverage in the marketplaces was smaller, and the average subsidy per person was larger. If this legislation was evaluated relative to the January 2017 baseline rather than the March 2016 baseline, it is unclear how different categories of insurance would be affected and whether the budgetary effects would differ noticeably.”
Emphasis added. The big question is why the CBO wouldn’t have used the January 2017 baseline instead of the March 2016 one, which obviously lacks an entire year’s worth of data and experience with Obamacare — including even higher premiums, even larger deductibles, and even more disappointing exchange enrollment. The closest the report gets to providing an explanation is this line: “The agencies have not had time to undertake a follow-on analysis of the effects of this legislation under that baseline.”
Or, as Rubio put it: “They said they’re busy.”
That’s unacceptable. And not even wholly rational: The Senate bill just came out this week; how was there time to analyze it with the March 2016 numbers but not the January 2017 numbers?
More to the point: Why wouldn’t we have even fresher data at this point, with the 2016 enrollment period now completely over? We know now, for instance, that rather than the 15 million exchange enrollees the CBO projected last year for 2017, there are 10.3 million, a difference of almost 33 percent from the projection and a decline from last year. That was not known in January. Why is anyone still basing a projection for a new bill on such an obviously faulty and out-of-date baseline number?
Senate leaders deserve some fault for this. They have been working on the bill for weeks now; why have they not in the meantime directed the CBO to update its projections with the most recent data? If the most recent data caused the number-crunchers to adjust their projections for future years to levels more realistic, based on past experience, that could have a huge effect on the projection for the Senate bill — and thus public opinion of the bill. Senate leaders had waited this long to release a draft. It surely would have made sense to wait longer if the CBO needed more time.
Overall, the sense I’ve gotten in just a relatively few hours here is that while there’s some political and procedural value in putting a bill text out in the public — if nothing else, it makes people start to get more serious about what they would or wouldn’t vote for — the bill wasn’t truly ready for prime time. The CBO estimate only underscores that.
ADDED Wednesday at 11 a.m.: In case you were wondering how much those CBO estimates have changed over the years, and specifically between March 2016 and January 2017:
For you math majors out there, that’s a decline of 5 million to 7 million per year. Taking account of that one change in projections would eliminate one-fourth to one-third of the coverage “losses” estimated for the Senate bill.