In defense of the Congressional Budget Office
© Getty images


As Senate Republicans lurch to repeal and replace the Affordable Care Act, the Trump Administration has undertaken an unprecedented campaign to undermine the credibility of a nonpartisan office strictly devoted to even handed evidence based evaluation of policy: the Congressional Budget Office (CBO).

This turn of events is both ironic and dangerous. CBO was established by Congress in 1974 because Congress was skeptical about the integrity of legislative cost estimates produced by the Nixon administration. Every member of Congress, Republican or Democrat, should want the best possible information in voting on legislation so consequential for all Americans. Only CBO can provide that kind of rigorous and balanced evidence.


CBO brings unique legal, technical, and institutional strengths to the estimation of the costs and consequences of a legislative proposal. First, as an arm of Congress – rather than a branch of the administration or an outside organization – CBO (working with the Joint Committee on Taxation) has exceptional access to government records and information.

These records include the most current information on the distribution of taxable income and taxes paid by all Americans. This information is critical to determining who will receive tax credits or use tax-favored savings accounts under the new health bill, and only CBO has those data. CBO also has the latest Medicaid and Marketplace spending and enrollment data. Of course, it also has access to all the publicly-released information which all other modelers rely upon. 

This unparalleled access to information means CBO has unmatched and unmatchable information on how new policy will operate. From the start, its analysis is more complete, current, and accurate than that of any other entity.

One cannot overstate the importance of this credible and consistent baseline information. When we ran the office in the Department of Health and Human Services charged with modeling outcomes of the Affordable Care Act, we were frequently stymied by a lack of data that CBO had easy access to. The Executive Branch has two fairly comprehensive models of the U.S. healthcare system, one in the Treasury Department and one at the Centers for Medicare and Medicaid Services. Because of restrictions on data-sharing within government, these models produced distinctly different results – so different that we could not adequately reconcile their estimates and generally used CBO estimates instead of our own. Only the CBO had all the information used by both these agencies.

CBO has enormous technical strength. The CBO healthcare team of over 30 analysts has analyzed hundreds of healthcare options. It includes economists, actuaries and legal researchers, expert in budget rules and the interpretation of congressional language. Modeling policy proposals is all they do – all day, every day, year after year, now for over 40 years.

While reputable outside modelers publish technical manuals describing the details of their models, CBO’s documentation is the most comprehensive, transparent and well-sourced. And CBO’s healthcare model is just one part of the Office’s responsibilities. CBO’s health team can rely on other parts of the Office to develop estimates of the economic growth, the population, the GDP, taxable income and inflation. This capacity is so unusual that virtually all modelers outside of CBO rely on the CBO forecasts of these key parameters for their own analyses.

CBO is institutionally unique. It exists to provide Congress with the best, non-partisan estimates of the consequences of health proposals. As Congress recognized back in 1974, any modeler working for the administration has an incentive to generate estimates that will please the president. Outside modelers, however fair, rely on their funders for business. Reputable modelers won’t lie – but they have the option to turn down a contract. The CBO has no such escape hatch. It must make estimates, whether it likes a proposal or not, and it cherishes its reputation as a bipartisan “wet blanket”. 

The CBO, under directors appointed by Democrats, doused the Clinton administration’s hopes for health reform back in 1993 and created numerous challenges for Democrats trying to craft their health reform proposal in the summer of 2009. Today’s CBO, under a Republican-appointed director, has drawn the ire of congressional and administration Republicans. But as Dan Crippen, a former Republican director of the CBO put it back in 2003: “It would be naive not to recognize that some of our critics have a preferred policy prescription and a preferred outcome.”  

The proof of CBO’s edge is in the fact that its estimates have been consistently more accurate than those of any other modeler. Recently, the Trump administration released its own analysis of one element of the Republican repeal proposal. It exemplifies the weaknesses of the worst non-CBO models. It utilizes inadequate data, makes rookie technical errors, and includes grossly inadequate documentation, modeling malpractice the administration didn’t even use the two existing models it had access to. Any half-decent card shark can pull the card she wants from a deck, and the administration has now shown that it is as adept as most at that parlor trick. But Americans deserve a Congress that considers the best available evidence about the impact of healthcare proposals, and that evidence will come from CBO. 

Richard G. Frank, Ph.D., is the Margaret T. Morris Professor of Health Economics in the Department of Health Care Policy at Harvard Medical School, and Sherry Glied, Ph.D. is Dean of the Wagner School of Public Service at NYU.

The views expressed by this author are their own and are not the views of The Hill.