When we released The 2014 Long-Term Budget Outlook on July 15th, a number of people asked for more information about how much we have revised projected federal health care spending during the past several years. This blog post tries to answer their questions.
CBO now projects that, if current laws remained generally unchanged, net federal spending for the government’s major health care programs in 2039 would equal 8.0 percent of gross domestic product (GDP)—1.6 percentage points, or about 15 percent, less than the 9.6 percent the agency projected in 2010 (see the figure below). That revision stems in large part from the observed slowdown in health care spending in recent years, but it also includes the effects of other factors; some of those factors reduced projected spending, and others increased it.
The programs included in the calculations are Medicare, Medicaid, the Children’s Health Insurance Program, and subsidies for insurance purchased through exchanges. The projections are part of CBO’s “extended baseline,” which generally reflects current law by matching the agency’s baseline projections for the first 10 years and extending that concept into later years. (See Appendix C of the July 15th report for a summary of the changes in CBO’s long-term projections over the past two decades. All of the long-term projections from 2010 discussed in this blog post incorporate the enactment of the Affordable Care Act in March 2010 and have been adjusted to account for changes in the estimation of GDP made by the Bureau of Economic Analysis in July 2013.)
The revision to CBO’s projections over the course of the past four years reflects three types of changes: legislation enacted since 2010, modifications to the agency’s economic outlook during that period, and technical changes (which encompass everything apart from legislation and changes in the economic outlook). CBO does not attempt to allocate the revisions to its long-term projections across those three categories as the agency does for its 10-year projections. However, in CBO’s judgment, the net downward revision to projected health care spending as a percentage of GDP in the past four years is more than accounted for by technical factors—that is, downward technical changes were greater than the net downward revision because they were offset, in part, by other types of changes in the other direction.
Changes in projections for technical reasons may encompass changes made for a wide variety of reasons, including new information from survey and agency data on programs’ enrollment and spending, improvements in the methods CBO uses to project spending, and the effects of administrative actions and judicial decisions.
In the case of the major health care programs, actual federal spending in recent years has been lower than we had anticipated, and analysis by CBO and others suggests that such spending will grow more slowly in the future than we projected several years ago. (For more discussion, see pages 32 to 37 of our recent report; a CBO working paper from last August, Why Has Growth in Spending for Fee-for-Service Medicare Slowed?; and a presentation I made last September, Comment on ‘Is This Time Different? The Slowdown in Healthcare Spending’.)
CBO has also revised down projected spending for those programs because of other technical factors. For example, revisions to projected enrollment have led to reductions in projected Medicaid spending. The two largest changes in projected enrollment since 2010 have been: a reduction in estimated enrollment stemming from the Supreme Court decision that made the expansion of Medicaid eligibility under the Affordable Care Act optional for states; and a reduction in projected growth in enrollment by elderly and disabled individuals. Projected Medicaid enrollment in 2020 is now lower by about 5 million people than CBO anticipated in 2010, a reduction of roughly 5 percent.
Taking all of the technical revisions together, CBO has substantially revised down its 10-year projections of federal spending for Medicare and Medicaid from those made four years ago, as shown in the following table:
|Technical Revisions to Projected Federal Spending Between June 2010 and July 2014|
(Net of Offsetting Receipts)
|Billions of Dollars||Percent||Billions of Dollars||Percent|
|Source: Congressional Budget Office.
Notes: Numbers are rounded to the nearest $5 billion.
The Long-Term Budget Outlook issued in June 2010 used CBO's March 2010 projections for the first 10 years, updated to incorporate the effects of the Affordable Care Act. The July 2014 volume uses CBO's April 2014 projections for the first 10 years.
Thus, apart from the effects of legislation and revisions to our economic forecast, CBO has revised down projected federal spending for Medicare (net of premiums paid by beneficiaries and other offsetting receipts) and Medicaid between 2011 and 2020 by roughly $1.1 trillion, or about 11 percent. (In a presentation last September, I showed a similar table for revisions between March 2010 and May 2013. Those revisions summed to the slightly larger figure of $1.2 trillion; we have made some methodological changes in these calculations since last May that have reduced the revisions a little.) In 2020, the reduction for technical reasons in projected spending for the two programs comes to about 16 percent.
New legislation and changes to CBO’s economic outlook since 2010 have also led to revisions in our projection of spending for Medicare and Medicaid. For 2020 (with numbers rounded to the nearest $5 billion):
Taking those pieces together, CBO’s projection of federal spending for Medicare (net of offsetting receipts) and Medicaid in 2020 is about $160 billion (or roughly 12 percent) lower now than it was in 2010; that difference consists of an upward revision of about $40 billion because of legislation and revisions to our economic forecast and a downward revision of about $200 billion because of other factors.
In the past four years, CBO has also revised down projected growth of private health insurance premiums, which reduces projected federal spending for the subsidies being provided through insurance exchanges; however, the agency has not produced a tally of those revisions at a level of detail comparable to what is provided here for Medicare and Medicaid. Overall, after numerous changes affecting enrollment, the distribution of income among enrollees, and the average subsidy per subsidized enrollee, projected outlays for exchange subsidies and related spending in 2020 have not changed significantly since 2010.
The downward revisions to CBO’s 10-year spending projections of federal health care spending have reduced the agency’s longer-term spending projections by lowering the estimated level of spending in 2024, to which estimated long-term growth rates are applied. In addition, because of the slowdown in health care spending during the past several years, CBO has reduced the long-term growth rates that it uses. The anchor for CBO’s long-term projections is a weighted average, since 1985, of growth rates of spending per capita in the entire U.S. health care system relative to growth rates of potential GDP per capita (with various adjustments, as described on pages 34 and 35 of the recent report). That historical weighted-average rate is now 1.4 percent; applying that same methodology to data from 1985 through 2007 yields a weighted average of 1.7 percent (as mentioned in footnote 43 of the report).
Some people have commented on the revisions to CBO’s long-term projections of federal health care spending since 2009. The revisions to those projections between 2009 and 2010 primarily reflected two factors—enactment of the Affordable Care Act in March 2010, and changes in CBO’s methodology for projecting long-term growth in health care costs. As described on pages 33 to 35 of The Long-Term Budget Outlook issued in 2010, those methodological changes were not principally a response to slower observed growth in health care spending—although the slowdown was, in fact, underway—but instead stemmed from changes in CBO’s judgment about what historical period best captured features of the health care and health insurance systems that were likely to endure. Therefore, in CBO’s judgment, the revisions to the agency’s long-term projections since 2010 provide a clearer illustration of how the effects of the slowdown in health care spending that the country has experienced have affected our expectations for future spending.