Following a recent hearing, we were asked by a Member of Congress to provide an updated estimate of the budgetary effects of the Affordable Care Act. Here is our answer (provided as part of a set of answers to questions for the record):
In March 2010, just before the Affordable Care Act (ACA) was enacted, CBO and the staff of the Joint Committee on Taxation (JCT) estimated that changes in direct spending and revenues under the legislation would reduce federal budget deficits by $124 billion over the 2010–2019 period and by roughly one-half of 1 percent of gross domestic product (GDP) over the ensuing decade (see the cost estimate for H.R. 4872, Reconciliation Act of 2010 [Final Health Care Legislation], March 20, 2010). In the four years since those estimates were produced, there have been significant changes in the economic outlook, in the health care and health care financing systems, in CBO and JCT’s estimating methodologies, in provisions of law that relate to the ACA, and in the implementation of the ACA as guided by judicial decisions and administrative actions. All of those changes could affect the impact of the ACA on budget deficits, potentially in significant ways.
In response to the request for an estimate of the net impact on the deficit of the ACA, the following points are important:
Consistent with their statutory responsibilities, CBO and JCT can continue to estimate the effects of prospective legislative actions, such as proposals to modify provisions of the ACA or to repeal the law entirely. Because of the complexities involved in implementing a repeal of the ACA, the budgetary effects of repealing the act at this time would not simply be the opposite of the budgetary effects of the ACA itself.
The principal obstacle to producing a new estimate for the ACA is that CBO’s cost estimates represent the budgetary effects of legislation relative to the current-law baseline. Because the ACA is part of current law, its budgetary effects would now need to be estimated relative to a counterfactual benchmark that excluded the ACA. CBO does not construct such a counterfactual benchmark for all of the ACA, and attempting to do so would raise significant challenges.
Under CBO and JCT’s normal procedures, the agencies still produce separate estimates of the effects of the ACA’s provisions related to insurance coverage, in part because those provisions established entirely new programs or components of programs and in part because those provisions are mostly being implemented in 2014 or later. In particular, the subsidies to be provided through insurance exchanges and the costs of expanded Medicaid eligibility are not part of the flow of budget data for preexisting programs. Hence, their budgetary consequences can be isolated and reassessed, and the counterfactual—what would have happened to those components of the budget in the absence of the ACA—is clear: Those amounts would have been zero. CBO and JCT have published updated estimates of the effects of those provisions numerous times since 2010 (see Updated Estimates of the Effects of the Insurance Coverage Provisions of the Affordable Care Act, April 2014). Over time, the effects of the coverage provisions that are not separately identified in flows of budgetary data will become increasingly difficult to isolate.
By contrast, the ACA’s provisions that are not related to insurance coverage largely modified existing federal programs and made changes to the existing tax code, so CBO and JCT cannot identify the incremental effects of many of those provisions. Consider the ACA’s substantial changes to the Medicare program, many of which have taken effect during the past four years. CBO does not produce baseline projections for Medicare that are based on the program’s statutes as of February 2010 to which the current baseline projections can then be compared. Moreover, the basis on which the agency could try to construct such a counterfactual baseline is unclear. With respect to the way Medicare pays certain providers, for example, CBO cannot determine the program rules and payment rates that the Centers for Medicare & Medicaid Services would have established over the past four years in the absence of the ACA. Moreover, CBO cannot determine how those program rules and payment rates under prior law would have affected the behavior of beneficiaries and providers—which in turn affects what federal spending would have been in the absence of the ACA. The basis for developing a counterfactual receipts baseline is also unclear because JCT cannot determine how taxpayers would have organized their financial affairs over the past four years in the absence of the ACA.
That problem is by no means unique to the ACA, nor is it related to developments regarding the implementation of the ACA that have surprised CBO and JCT. Indeed, judicial decisions and numerous administrative actions have caused the ACA’s provisions related to insurance coverage to be implemented differently than CBO and JCT had initially expected, and, as noted above, the agencies continue to update their estimates of the budgetary effects of those provisions. Rather, the problem is common to all legislation that changes existing federal programs or tax provisions with results that cannot be clearly distinguished from what would have occurred under previous law.
Last year, in response to a question from a Member of Congress about whether CBO goes back to review its estimates of legislation in order to improve the accuracy of its methods (see a March 2013 blog entry titled “The Accuracy of CBO’s Budget Projections”), CBO wrote:
CBO routinely monitors the budgetary effects of enacted legislation to help improve projections of spending and receipts under current law, as well as to improve cost estimates for new legislative proposals. However, it is often difficult or impossible to determine, even in retrospect, the incremental impact on the budget of a particular piece of legislation …
The prescription drug program known as Medicare Part D is a relatively rare example in which actual spending can be directly compared to the projections contained in the CBO cost estimate. In most cases, legislation modifies existing programs; it is often not possible after enactment of such legislation to determine how spending for a modified program has changed specifically as a result of that legislation, or how much of future spending would have occurred even without the change in law. In contrast, the legislation that created Part D established a new component of Medicare with a system of new benefit payments, associated administrative costs, and payments from premiums and states. The actual net cost of Medicare Part D has been much lower than CBO originally projected.
Moreover, determining the budgetary impact of previously enacted legislation that affects ongoing spending programs or tax receipts becomes more difficult over time as the conditions that would have prevailed in the absence of the original legislation become increasingly uncertain. Thus, in its estimate with JCT of the effects of a proposal to repeal the ACA in July 2012, CBO wrote: “Separating the incremental effects of the provisions in the ACA that affect spending for ongoing programs and revenue streams becomes more uncertain as the time since enactment grows.”
The largest changes in the estimated effects of the ACA during the past four years that CBO and JCT have separately identified are those associated with the estimated effects of the ACA’s insurance coverage provisions and the elimination of the Community Living Assistance Services and Supports (CLASS) program. CBO and JCT’s latest estimate of the cost of the coverage provisions is $100 billion lower than the March 2010 estimate for the period from 2014 through 2019 (2019 was the last year of the 10-year budget window used in the original estimate). CBO originally estimated that the CLASS program would yield federal budgetary savings of $70 billion through 2019 (and would have a budgetary cost in later years); however, the Secretary of Health and Human Services announced in 2011 that she did not “see a viable path forward for CLASS implementation.” Combining the reduction in estimated cost of $100 billion and the loss of estimated savings of $70 billion with the original estimate that the ACA would reduce deficits by $124 billion over the 2010–2019 period yields a projected reduction in deficits of more than $150 billion over that period.
The costs and savings that can be attributed to other provisions of the ACA have undoubtedly been affected by many developments in the four years since the law was enacted. Economic conditions during the past four years and CBO’s projections of the economy in coming years are different from what CBO projected several years ago. The health care and health care financing systems have continued to evolve, and health care spending—both in federal programs and in the private sector—has been below the amounts that CBO expected in early 2010. The implementation of the ACA has been guided by numerous regulations, some of which have differed from what CBO and JCT anticipated in their original estimates. For the reasons discussed above, however, CBO and JCT can no longer assess the budgetary effects of all of the provisions of the ACA without constructing a counterfactual benchmark.
Attempting to construct a counterfactual benchmark for the budget that excluded the ACA would raise significant challenges and would go beyond CBO’s traditional role in the budget process. In particular, given the number and complexity of the changes to Medicare under the ACA, as well as the lack of data about what would have happened in Medicare without the ACA, creating a counterfactual benchmark for Medicare would be a very complicated task that would take months of work, and the resulting benchmark would be hugely uncertain and speculative. For example, constructing such a benchmark would require CBO to assess precisely what role the ACA has played in the slowdown in Medicare spending, and the appropriate analytical approach for making such an assessment is unclear. Moreover, the slowdown in Medicare spending has been part of a broader slowdown in national health care spending, and the effects, if any, of the ACA on that broader slowdown also would need to be taken into account in trying to construct a counterfactual benchmark.
Such challenges are why CBO traditionally has not produced retrospective estimates of the budgetary effects of enacted legislation but has produced cost estimates only for forward-looking legislative proposals relative to the current-law baseline. In that way, the agency has focused on its ongoing responsibilities to analyze pending legislation and to build models and other tools in preparation for analyzing future legislation.
Consistent with their statutory responsibilities, CBO and JCT can continue to estimate the effects of legislative proposals to modify provisions of the ACA or to repeal the law entirely. However, the agencies could not produce useful, timely estimates for proposals to repeal the ACA unless those proposals specified what would take the place of the law in areas where it modified preexisting programs and where those modifications have been under way during the four years since the law was enacted (in some cases, with further legislative modifications).
Consider again the ACA’s changes to Medicare. In its estimate of the effects of a proposal to repeal the ACA in July 2012, CBO wrote: “[The proposal] does not specify how to implement the requirement that the provisions of law modified by the ACA be restored as if the ACA had never been enacted—for example, with regard to Medicare’s payment rules and certain changes to the Internal Revenue Code that are already in operation. Because of that ambiguity, H.R. 6079 would cede considerable discretion to the executive branch to implement its provisions.”
That challenge is much greater two years later and will become still greater as more time passes. Because the ACA’s changes to Medicare’s payment rules have now been in place for up to four years, estimating the effects of a proposal that sought to repeal those provisions and largely delegated to the executive branch the task of determining what specific policies to implement instead would involve two steps: first, assessing what policies the executive branch would implement under such broad authority, and second, estimating the budgetary effects of those policies. In some cases, those policies might be quite similar to ones that would have been adopted under the law prior to the ACA; in other cases, those policies might be quite different from prior policies because of changes in the health care or health care financing systems that have occurred because of the ACA.
Given those challenges, projecting spending for Medicare under a proposal to repeal the ACA would be much more complicated and time-consuming—and the result much more uncertain and speculative—if the legislation did not specify an alternative set of policies to take the place of those established by the ACA. Similar considerations would apply to proposals to repeal many other provisions of the ACA, which have affected a wide range of ongoing programs.
Whether a proposal to repeal the ACA specified alternative policies or not, its budgetary effects would not simply be the opposite of the budgetary effects of the ACA itself. In its estimate with JCT of the effects of a proposal to repeal the ACA in July 2012, CBO wrote: “[We] also anticipate that some of the changes induced by the ACA in how public and private health insurance and health care programs are administered would be sustained” even if the law was repealed. Two years later, that conclusion would probably apply even more broadly.