Talking to a Former Director of the CBO, the Health-Care Fight’s Most Important Government Agency

Doug Elmendorf, lover of charts and data, prior to testifying about fiscal 2015 legislative branch appropriations for the Legislative Branch Subcommittee of the House Appropriations Committee. Photo: Tom Williams/CQ-Roll Call

Approximately 22 million people would lose health insurance under Trumpcare, the nonpartisan Congressional Budget Office (CBO) estimated in its most recent report on the Republicans’ attempts to repeal and replace the Affordable Care Act. The CBO score for every GOP attempt at health-care reform has found millions losing coverage, which has produced round after round of bad headlines and spooked Republican lawmakers.

Because of the agency’s continued prominence as the health-care fight unfolds, Daily Intelligencer spoke with Doug Elmendorf, director of the CBO from 2009 to 2015. Elmendorf led the agency during the battle over Obamacare, and he explained how the CBO operates and adjusts (or doesn’t) to demands of politicians.

This week, Office of Management and Budget director Mick Mulvaney characterized CBO’s previous analysis of Trumpcare as “almost … not a fair analysis.”
Director Mulvaney’s attacks on CBO are well beyond previous sorts of criticism. They are unfounded, in my view, and destructive for the country. Remember, he’s the OMB director who criticized the Bureau of Labor Statistics for not calculating the unemployment rate honestly. He has a very unfortunate tendency to undermine facts that are inconvenient for the administration’s purposes. I think that is very damaging for policy discussions in this country.

That does not mean that CBO analysis is beyond criticism. But one has to bring evidence and explain why that evidence contradicts or undermines what the CBO has done. Mick Mulvaney has emphatically not provided evidence of that sort. If you go back to our estimates with the Affordable Care Act, we did not get everything right. There’s no doubt about that. But our estimates were much more accurate than most of the wild guesses by both opponents and proponents of the Affordable Care Act at the time.

The question is, do you want imperfect estimates from knowledgeable, independent people, or do you want to rely on totally partisan guesses by people without the same level of expertise? I think the answer to that question is obvious.

Mulvaney cited CBO’s use of the “Jonathan Gruber’s methodology.” What is that?
Jon Gruber is an economist at MIT. He is one of a number of people who built models of the health-insurance system more than decade ago. As CBO was building its own model of health insurance, Gruber was one of the experts with whom the CBO consulted. But we mostly stopped that in 2009 when we were trying to analyze the Affordable Care Act. We realized we were moving from a general model [of health insurance] into working on specific proposals which were confidential and we could not share.

So, there are certainly commonalities between the CBO’s and Jon Gruber’s modeling. You expect that, in that there’s been a lot of work by smart people on this. CBO’s models today are different than when I was there just because new information about what’s going on in the world requires adjustments.

How do the people inside the CBO take public criticism of their scores?
The people at CBO are used to being criticized because the work they do is important and receives a lot of public scrutiny. I thought it was very important when I was director, as all the directors have thought, that CBO be quite open to new information, to new evidence, to new arguments about how to analyze some proposals. But, at the same time, completely resistant to political pressure. That is a standard I firmly believe has been met by CBO consistently for 42 years.

The analysts and managers at CBO just focus on the work they have to do. [But] you can’t block it out entirely. When I was director, we changed our estimates on a number of things and we were very explicit. We wrote letters and reports that said, “This is a different estimate than we have done before because we have read the following studies or have done the following analysis or something had changed in the world.” We are not averse to changing our minds. It was important to be open to that. But only in response to evidence — not yelling.

What has been your impression overall of the CBO’s role in Republican efforts to repeal the Affordable Care Act?
I think the folks at CBO have done an outstanding job during this past six months of working on ACA repeal. The caliber of the analysis has been extremely high and the clarity of the presentation of the analysis has been extremely high. Those are both hard to do, and they’re especially hard to do at the pace at which CBO has had to work.

Speaking of fast pace, this health-care process moved much more swiftly than the ACA. How do you think the CBO managed?
CBO adjusts to the needs mostly by working harder and longer hours. It’s very hard to take shortcuts in the analysis itself. To analyze the effects of a health-insurance proposal like the one they’ve been doing, it takes a lot of people, with a lot of models, a lot of time with data, a lot of interactions with state officials. You really can’t do an analysis without all of those steps. So when the time is shorter, CBO and their colleagues at the Joint Committee on Taxation, they just work more hours of the day.

When I was there and we were working on the ACA, there were times when the night owl on the staff would work on a document, say from 10 p.m. to 2 a.m., and the early birds, like me, would get up at 4 and take over the document at 4 a.m., then do four hours of work for 8 a.m. and by that point other people would be in the office and they can the document from there. We ended up with a nearly round-the-clock process.

As you said before, the CBO’s analyses are imperfect — they’re forecasts. How does the agency deal with those limitations?
CBO directors and analysts are acutely aware of the uncertainty of their estimates. And CBO reports estimates with precise figures, sometimes with decimal places because the Congress tends to work with point estimates. Budget resolutions say you need to cut taxes or raise taxes by x billion dollars. Not x to y billion dollars.

Where CBO has the time to do more analysis, it tries to provide a confidence region around the point estimate to give a sense of uncertainty. So, when I was director, almost all of the economic analyses we did — effects of deficit reduction, the effects of the economy in different ways, the budgetary effects of different ways of reducing the deficit, or the economic effects of different ways of stimulating the economy, or the effects of the Medicare premium support system — we provided confidence intervals. We said, “Our best guess is six, but it could be between four and eight.” CBO is quite happy to show that uncertainty, but in general, CBO does not have enough time to do that because of the pace of what Congress is doing.

Now, the uncertainty comes from a lot of things. For example, the economy varies a lot of the time. Economists are notoriously bad at predicting overall economic conditions, but certainly if you end up in a recession, or in a boom, that’s gonna have a big effect on the overall budget but also on what companies do about health-insurance benefits and so on. The health-care system itself is very uncertain.

What’s an example of the CBO being wrong in the past?
The cost of the Medicare drug benefit. In 2006, I guess when that was being worked on, CBO had a prediction of what that would cost — but the actuary at Centers for Medicare and Medicaid Services actually had notably higher projections, and they’re very talented, objective people as well. It turns out the cost is much less than the CBO would have expected and even further below what the CMS actuary predicted. The reason is that drug spending slowed very sharply cross the country as a whole. It was not just in Medicare. Those were features of the law, competitive features that held down costs. CBO knew that, and CBO expected a national slowdown in drug spending. The slowdown we got was dramatically slower. And CBO did a set of reports when I was there analyzing what went wrong so it can do better estimates in the future.

Do you think reporters have fairly represented the CBO’s role in the health-care debate?
Members of the press have been very valuable — as I found them to be when I was director — at conveying CBO’s analysis to a larger audience. I felt as director, even for most members of Congress and their staffs, a lot of what they would learn about CBO’s analysis would come from reading a story in the newspaper. They’re not always going to the website and reading the actual CBO report. That’s not true for the people working most closely on a particular bill, but for many other members of Congress and their staffs, they want a condensed lowdown that they get from the Washington Post or what have you.

How would you describe the relationship between lawmakers and the CBO?
In my experience, almost all of the members of Congress have a lot of respect for CBO’s expertise. When I went to meet with members, I found them very interested in our perspective. It’s also true that CBO gets in the way of members when they are advocating some proposal for which CBO’s analysis is not very positive, or when they’re opposing proposals for which CBO’s analysis is viewed as being positive. When members are frustrated, they tend to express that. Sometimes more in public than in private.

What was it like on the CBO side to pass the Affordable Care Act? When the Democrats were trying to pass the ACA they seemed to take CBO’s numbers more seriously than Republicans are now.
I think you’re right about the difference. But it’s not something about Republicans and Democrats intrinsically. When I was the director we did lots of work for Republican proposals and Republicans often changed those proposals in response to initial estimates from us.

I think the difference you’re seeing here is that Democrats had a clear view. They wanted to maximize the increase in health-insurance coverage at a budgetary cost that could be offset by other spending cuts and tax increases. Then, President Obama issued some specific number that he wanted the cost to be below and Democrats wanted to minimize disruption to preexisting health-insurance arrangements. When we produced an estimate that said, “Well, this version doesn’t cover as many people and it costs a little more,” then it was straightforward. Democrats at least knew what direction they wanted to move in from there.

And the Republicans now?
The Republicans don’t have a comparably clear vision. The fundamental problem they faced in the last six months is, what they have been promising the American people for the last seven years is not achievable. They promised a combination of lower federal spending, fewer rules, and yet everybody gets health-insurance coverage and spends less themselves.

That turned out to be an effective political argument, but it cannot be achieved. So, when CBO comes back and says, “Well, if you cut back subsidies a lot, and roll back a bunch of the rules that led to the increase in insurance coverage, you’re gonna lose some insurance coverage.” That doesn’t surprise anybody who’s a health-care analyst. It’s not obvious, then, given the inconsistency of the Republican starting point, how one should respond to that. That’s why you’ve seen a lack of response. They’re stuck, and there’s no way out of the box they’ve created for themselves.

This interview has been condensed and edited.

Talking to a Former CBO Director About the Health-Care Fight