Shaky Ground: Congressional Budget Office Estimates of Major Health Care Reforms

by Pete Sepp, Demian Brady / /


Shaky Ground: Congressional Budget Office Estimates of Major Health Care Reforms

NTUF Issue Brief #169
By Demian Brady and Pete Sepp


Since 1975, the Congressional Budget Office (CBO) has been tasked with providing Congress with detailed analysis on the budgetary and economic impact of proposed laws. Because of its “scorekeeper” role, CBO wields an immense power over the legislative process, with some calling it “the most powerful agency you have never heard of.”

However, there are many examples – including scores of farm bills, liabilities in federal student loans, and air traffic control reform – where CBO has been criticized for providing a distorted picture of federal fiscal policies. Many of CBO’s most crucial and problematic analyses have been in the area of health care policy. The fate of major overhaul packages can hinge on CBO’s cost projections. Its figures will be at the center of the current fiscal and policy debate of repealing and replacing the Affordable Care Act (ACA). As the estimates are being discussed in Congress and in the media, lawmakers – and taxpayers – are well advised to consider CBO’s track record of providing estimates that relied on assumptions that later proved faulty.

Congress’s Scorekeeper: Small Agency with an Outsized Role

CBO was established in the Congressional Budget and Impoundment Control Act of 1974. This Act set in place the modern budget process: it established the standing House and Senate budget committees, required Congress to approve an annual resolution specifying levels of spending and tax revenues, put in place a timeline for developing and passing the budget, and enacted a reconciliation process to help ensure that outlays and revenues adhere to the targets in the budget resolution. CBO was to provide budgetary and economic analysis to Congress. Previously, lawmakers were dependent on the White House’s Bureau of the Budget (renamed the Office of Management and Budget in 1970) for fiscal and budget projections. Now, Congress’s own budget arm employs roughly 235 employees across eight policy divisions, with a budget that has grown from $5 million in 1976 to $45 million in FY 2016.

The agency’s primary task is to conduct cost estimates of proposals under consideration in Congress. Even as a relatively small entity, the influence wielded by CBO over the legislative process is enormous. Its projections can determine the fate of bills. Legislation is drafted and revised, or dies, because of the figures reported by CBO. And although many of the Budget Impoundment and Control Act’s provisions have come under criticism for their flaws, comparatively little scrutiny has been given to CBO. The purpose of this National Taxpayers Union Foundation (NTUF) Issue Brief – and those that will follow – is not to denigrate all of CBO’s work, which is often of high quality and unassailable standards. Rather, it is to provide a constructive assessment of those areas in which either CBO’s methods could be improved, or policymakers’ expectations must be adjusted. It is also clear to NTUF that the entire budget process could benefit from greater outside oversight and thoughtful input on matters of budget estimates and legislative scoring.

Deciding the Fate of Major Health Care Reform

While the extent of CBO’s influence may come as a surprise to some, its role in past health care reform debates demonstrates its importance. The best illustration of this can be found in CBO’s score of the Clinton Administration’s comprehensive health care reform in the early 1990s known as “HillaryCare”. Contradictory claims that it would add to the deficit landed a mortal blow to the proposal. Because then-President Clinton was dedicated to tackling the budget deficit, CBO’s analysis turned the administration’s reform plan into a massive political liability, and therefore made it impossible to pursue. That led Harvard University political scientist Theda Skocpol to state, “[CBO] has by now become virtually a sovereign branch of the US federal government, comparable in clout in relation to the executive and the Congress to the courts back in the Progressive era and the New Deal.”[1]

Yet, even as CBO’s projections of HillaryCare led to its failure, its score of President Obama’s Affordable Care Act was essential for the bill’s passage. Due to CBO’s indication that the ACA would shrink the deficit, then-Senate Majority Leader Harry Reid (D-NV) was empowered to utilize the budget reconciliation process, which reduced the number of votes necessary to advance the bill. If CBO’s score of the ACA had reflected its budgetary predictions of HillaryCare’s impact – a real possibility if the individual mandate had been scored as a tax – the bill would have languished in the Senate due to political and procedural obstacles.

Subsequent to the bill’s passage, lawsuits were filed and conservative Members of Congress rallied support for the repeal of “Obamacare” due to expectations of decreased efficiency, poor incentives in administration, and the skyrocketing costs of health care services. NTUF was the first organization to thoroughly address CBO’s flawed analysis of the economic impact of the repeal of the ACA. We published an issue brief just days after the CBO report in question entitled, “Repealing ‘Obamacare’: A Look Beyond the Media’s Misguided Deficit Focus”, exposing faulty research that argued the repeal of the ACA would have severe negative impacts on the federal deficit.[2]

In recent years, CBO has failed to produce the complete scores necessary for carrying out an informed debate on several major pieces of health care legislation. In May of 2013 and again in January of 2015, CBO stated it was unable to provide an estimate for a proposal to repeal the ACA.[3] In June, CBO did provide a macroeconomic and budgetary analysis on the ACA, stating that repealing the Act would increase the federal deficit. However, the estimate was incomplete, as it only showed how the repeal would impact the federal budget. Since ACA was designed with so many fees and taxes, including the individual insurance mandate “tax,” it would raise more in revenues than it would disburse in benefits over CBO’s 10-year budget window. Even though it projected that repealing the ACA would result in an increase in GDP of 0.7 percent over the 2021-2025 period, CBO neglected to budget for the positive economic effects associated with repealing burdensome taxes. In a 2012 issue brief, NTUF found that repealing Obamacare would actually reduce federal spending by $319.5 billion over five years. A more recent estimate of repealing the law did account for macroeconomic feedback and determined it would save $470 billion over five years, and $1.4 trillion over a decade.[4] While there are additional discretionary savings that would be achieved, details have not been fully reported.

Methodological concerns aside, CBO even admitted a great deal of doubt over the score it had provided, stating, “The uncertainty is sufficiently great that repealing the ACA could reduce deficits over the 2016 - 2025 period – or could increase deficits by a substantially larger margin than the agencies have estimated.”[5] Given the limits CBO conducted in its assessments, the range of impact is not entirely inexcusable. It is also clear that CBO’s equivocation should have been of little use to policymakers seeking data before taking future legislative steps. Instead, mainstream media, egged on by proponents of the ACA, continued to treat CBO’s estimate as gospel without noting serious caveats.*

Components of ACA Prove Problematic

In addition to the coverage and topline budget figures, CBO’s assumptions regarding components of the ACA proved to also be problematic. Some of these helped to ease passage of the law because of the offsets that were expected yet failed to materialize, and others that cost significantly more than projected.

Community Living Assistance Services and Supports (CLASS) Act

The CLASS Act played a key role in the ultimate passage of the ACA ...  thanks to CBO’s score. The plan, originally introduced by Senator Ted Kennedy, would have established a long-term care entitlement program. As written, beneficiaries would have signed up for the program and begin paying premiums – scored as offsetting receipts which are akin to negative spending – in the early years of the ACA’s budget window, but benefits would not begin to be paid out until after an individual has been vested for five years. The CLASS Act was packaged with a serious flaw: the law prohibited medical underwriting, which is the use of medical or health information in the evaluation of an applicant for coverage. This meant that the premium costs for younger, healthier individuals would be uncompetitive with options that could be obtained in the private market.

To an unbiased outside observer, this represented a major obstacle to the long-term sustainability of the new long-term care entitlement. Senator Kent Conrad (D-ND) called the CLASS Act “a Ponzi scheme of the first order, the kind of thing that Bernie Madoff would have been proud of.”[6] But, CBO estimated that the CLASS Act’s premiums would collect $40 billion over the first five years and $102 billion over ten, providing a significant offset to the CBO’s deficit projection of the ACA.[7]

This estimate was soon proven to be ungrounded in reality. When it came time to roll out and implement the CLASS Act, the government had to face the facts of the program’s dim prospects for sustainability: in October of 2011, the Secretary of Health and Human Services announced that she did not “see a viable path forward for CLASS implementation at this time.”[8] The CLASS Act was shelved, and eventually repealed by Congress.

Health Demonstration Programs: Scoring Savings before They’re Advanced

The Center for Medicare and Medicaid Innovation (CMMI), another program included in the ACA, was established to facilitate demonstration projects for payments and services within these programs. The intent was to “test innovative payment models” that would potentially lead to savings within the two entitlements. The scoring assumptions from CBO on the agency’s mandatory Medicare Part B reimbursement models raised a number of concerns, over behavioral changes to the economics of the health care and the viability of cost estimates. In testimony before a House hearing, Joseph Antos, of the American Enterprise Institute noted:

The most important issue is CBO’s decision to account for savings that might be lost for ‘initiatives CMMI is undertaking (or is expected to undertake)’ [emphasis added] in assessing legislative proposals. This remarkable decision to score lost savings for demonstration projects that have yet to be announced is a sharp break with past practice. Just as CBO does not score legislative proposals that have not yet been advanced, it seems unreasonable for them to score actions by CMMI that have not yet been advanced.[9]

CBO believed that CMMI’s unelected bureaucrats would have saved tens of billions of dollars from Medicare and Medicaid, but if Congress should wish to set policy instead, it will “cost” taxpayer dollars. This is not only bad scoring; it’s an inappropriate weakening of Congress’ right to make entitlement policy. CBO also assumed that legislation to repeal CMMI’s $10 billion appropriation would not save as much as anticipated because CBO’s baseline had baked in $38 billion in savings from CMMI’s regulatory acts.

The Expanding Costs of ACA’s Medicaid Expansion

The passage and implementation of the ACA has been rife with controversies, but among the biggest and ongoing concerns have been its Medicaid provisions. Originally the ACA required states to expand Medicaid to cover those earning less than 138 percent of the Federal Poverty Level, but the Supreme Court struck this down as unconstitutional. The provisions was subsequently modified so that instead of the federal mandate, states that expand Medicaid will receive a 100 percent match from the federal government for the new enrollees through 2016. Nevertheless, the costs to taxpayers remained significant, and turned out to be much higher than expected. CBO’s estimate badly missed the mark in assaying the ultimate costs of the ACA’s Medicaid expansion.

Comparing CBO’s original estimates before the expansion with what actually occurred, a study by the Mercatus Center at George Mason University found that the number of enrollees was nearly 50 percent higher than expected. This has a significant impact on the ensuing costs: “After adjusting previous projections for current assumptions of state participation, the Congressional Budget Office’s current expectation of federal Medicaid spending between 2016 and 2024 is $232 billion in excess of 2014 estimates.”[10] Taxpayers can only imagine how much greater the magnitude of CBO’s error would be, had a greater number of states decided to participate in the Medicaid expansion.

The problem stems from the fact that CBO overestimated the number of healthy individuals that would sign up for Obamacare’s health insurance exchanges. It was assumed that the individual mandate and the ACA’s “tax penalty” would drive people into the exchanges. Rather, Medicaid saw a much higher level of enrollment than was forecast. Nearly 50 percent more people signed up than expected. Unfortunately for taxpayers, this unexpected boost in Medicaid enrollees was matched by a higher-than-expected level of spending per enrollee. The costs for FY 2015 had been projected to be $42 billion but turned out to be $68 billion – 62 percent higher. [11]

Overstating Costs: The Case of Medicare Part D

Sadly, there is a long list of health care programs whose actual costs were grossly underestimated, many involving CBO.[12] One examples stretches all the way back to 1987, when Congress was drafting legislation for Medicaid’s Disproportionate Share Hospital (DSH) payments program. Utilizing CBO baseline estimates, the House Committee Report on the bill projected that within five years (1992); the price tag for DSH would reach $1 billion. As it turned out, the actual cost in 1992 was $17 billion – a catastrophic overrun.[13]

However, this is not always the case.

In 2003 Congress was crafting what would become the biggest new entitlement since Medicare was created in 1965, by adding a new benefit (known as Part D) to that program for the purchase of prescription drugs. In the previous year, the Republican-controlled House passed a Medicare drug bill that would cost $350 billion over ten years, while the cost of Democrats’ plan in the Senate was closer to $600 billion. The 2003 budget resolution limited the proposed prescription drug program’s costs to $400 billion over its first ten years. If the cost went above the level, the legislation would be subject to a “point of order” challenge that would have likely doomed passage. This increased the importance of the CBO score for the bill, and also encouraged lawmakers to tweak the bill, deferring what appeared to be raising costs outside of that budget window.

That November, CBO officially projected a net cost of $395 billion for the reform package. This was a net of $410 billion for the new benefit, minus reforms in other programs. Because the program would not kick in until 2006, the average annual cost over the first five years was $22.1 billion, while it was nearly $60 billion in years six through ten. The bill, the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), narrowly passed after the House vote time was extended, while leadership put pressure on a sufficient number of Members to switch their votes. The outcome might have turned out much differently if a different, more costly, estimate was available to lawmakers.

That summer, Richard S. Foster, Medicare’s chief actuary, estimated that the new plan would cost $530 billion over its first ten years. Foster’s projection could well have halted the MMA’s progress, but he was pressured to keep the figures under wraps. His estimate was not reported until the end of January 2004, after the law was enacted.[14]

Medicare Part D provides an example where the actual program costs were significantly lower than initially expected.

Several factors contributed to the lower costs. Part D uses private insurer plans to deliver the benefit, encouraging competition.[15] This principle could helpfully guide ACA reformers today. However, enrollment was also lower than anticipated and there was a general slowdown in the growth of national drug spending per capita.[16] In addition, the use of prescription drugs was found to have a small but measurable offsetting effect on overall health care spending. CBO subsequently modified its methodology regarding health care estimates after finding that for every one percent increase in the number of prescriptions filled by Medicare recipients, spending on Medicare and other federal programs that include drug utilization is anticipated to decrease by roughly .2 percent.

Although taxpayers should consider Part D to be a happy ending, they are also left to wonder whether “legislating to the score” over a ten-year window might have prevented the program from functioning even better than it has turned out to be. Once again, it must be granted that health care projections are notoriously difficult to compile, but as the controversy surrounding the Actuary’s report demonstrated, allowing a variety of informed opinions on the cost of legislation could encourage more robust debate and aid in more prudent drafting of legislation.


“Everyone should know that any number will be either too high or too low.” Don Marron[17]

With clear evidence that CBO’s determinations have serious, demonstrable impacts upon the prospects of bills in Congress, and that its major assumptions regarding coverage rages have been questioned, it is imperative that these determinations be accurate and transparent to taxpayers. In 2009, the Joint Economic Committee reported on cost overruns of health care programs and noted:

This is not to say the official “scorekeepers” are bad at their jobs.  On the contrary, they typically exhibit very high levels of skill, integrity, independence, and professionalism – often working under extremely tight time frames and hectic conditions, and not infrequently amidst a din of interested voices attempting to influence their work.[18]

While in some cases this statement may be true, it is clear, especially with the more recent history of the Affordable Care Act, that scorekeeping orthodoxies got in the way of providing complete and realistic scenarios of possible outcomes. In any case, once again taxpayers are reminded that CBO is regarded in the media and among many in Congress as a final arbiter, even when that final arbiter expresses caution over its own findings. This only reinforces the notion that an oversight institution may be necessary not only to thoughtfully evaluate CBO’s work and recommend alternative approaches, but also to hold public officials themselves accountable for misportraying the certainty and validity of some cost estimates.

This is why NTUF will soon announce an ambitious new project intended to bring accountability to CBO, which is one of the least understood, but most influential public policy outfits in Washington. There are precedents for this approach, among them the Shadow Open Market Committee (SOMC), whose members are drawn from academia and the private sector to evaluate the Federal Reserve Open Market Committee’s decisions. Just as SOMC helpfully serves as an institutional balance to the feds, NTUF seeks to provide a measure of balance, oversight, and thoughtful evaluation of legislation and their ultimate impact on taxpayers’ bottom line. In the process, the quality of legislative output could substantially improve.


About the Authors

Pete Sepp is the President of National Taxpayers Union Foundation (NTUF), the research and educational affiliate of National Taxpayers Union.

Demian Brady is NTUF’s Director of Research.



* For examples, see:,,, and


[1] Theda Skocpol, Boomerang: Health Care Reform and the Turn Against Government, W. W. Norton & Company, New York, 1997. Page 67.

[2] Demian Brady, Repealing “Obamacare”: A Look Beyond the Media’s Misguided Deficit Focus, National Taxpayers Union Foundation, July 30, 2012.

[3] CBO, H.R. 45, a bill to repeal the Patient Protection and Affordable Care Act and health care-related provisions in the Health Care and Education Reconciliation Act of 2010, May 15, 2013. While page is no longer available on CBO’s website, a cached version of it has been saved in by Google’s search engine at

CBO, Letter to the Honorable Pete Sessions, January 30, 2015.

[4] CBO, Estimate of Direct Spending and Revenue Effects of H.R. 3762, the Restoring Americans’ Healthcare Freedom Reconciliation Act, as Passed by the Senate on December, 3 2016, and Following Enactment of the Consolidated Appropriations Act, 2016, December 11, 2015.

[5] Congressional Budget Office, Budgetary and Economic Effects of Repealing the Affordable Care Act, June 2015.

[6] Lori Montgomery, “Proposed long-term insurance program raises questions,” Washington Post, October 27, 2009.

[7] CBO, H.R. 3962, the Affordable Health Care for America Act: Revised Estimate, November 20, 2009.

[8] Secretary Kathleen G. Sibelius, “Secretary Sebelius’ Letter to Congress about CLASS,” October 14, 2011.

[9] Joseph R. Antos, “Statement before the House Committee on Budget: Center for Medicare and Medicaid

Innovation Scoring Assumptions and Real-World Implications,” American Enterprise Institute, September 7, 2016.

[10] Brian Blase, Evidence Is Mounting: The Affordable Care Act Has Worsened Medicaid’s Structural Problems, September 14, 2016.

[11] Brian Blase, “Learning from CBO's History of Incorrect ObamaCare Projections,”, January 2, 2017.

[12] Washington Times, “U.S. Health Plans Have a Long History of Cost Overruns,” November 18, 2009. 2009/nov/18/health-programs-have-history-of-cost-overruns/

[13] Joint Economic Committee, Are Health Care Reform Cost Estimates Reliable? History Shows True Costs Are Often Significantly Understated, July 31, 2009.

[14] McKnights, “HHS: Scully pressured Foster to withhold Medicare reform costs,” July 7, 2004.

[15] Kathryn Nix, “Recipe for Reform: Success of Consumer-Driven Principles in Medicare Programs,” Heritage Foundation, August 10, 2011.

[16] Jack Hoadley, “Medicare Part D Spending Trends: Understanding Key Drivers and the Role of Competition,” Kaiser Family Foundation, May 2012.

[17] Peter Buxbaum, “Congressional Budget Office consistently wrong on health-care estimates,” Daily Caller, January 10, 2010.

[18] Joint Economic Committee, Ibid.