February 3rd, 2012
As we embark upon a presidential campaign season, we can anticipate many lively debates on the topics of taxation and spending in this nation. As health spending in the Unites States accounts for 18 percent of our gross domestic product – a rate often called unsustainable – it is critical that we be clear-eyed in our understanding of the tradeoffs between tax revenues and a sustainable rate of national health spending.
Few have tried to paint a comprehensive scenario of what a sustainable rate of health spending might look like in the United States. That is what the research team at Altarum institute’s Center for Sustainable Health Spending has endeavored to do in this article. Only by having a well researched and objective picture of sustainable spending growth can we set measurable targets and track progress toward reaching what should be a national goal of this highest priority.
In an earlier blog on Altarum Institute’s Health Policy Forum, I defined sustainable health spending as a projected growth path of spending that is within what the nation is willing to pay. I argued that, under the Affordable Care Act (ACA), the protections provided by the federal government reduce the effects of health spending growth on other stakeholders. (See note 1) Consequently, the sustainability of a given rate of health spending growth largely rests with the federal government’s ability to meet its various health spending commitments, including those added under the ACA. This, in turn, rests upon the willingness of the citizenry to allocate the necessary tax revenues. Given the inherent uncertainties regarding the public willingness to pay future taxes, this research is partly hypothetical, but I think it provides insights into the tradeoffs between tax revenues and health spending that contribute to informed decisions about targets for taxes and health spending.
My goal is to estimate the average annual rate of growth in health spending that is consistent with a long-term balanced federal budget in which tax revenues, and the portion allocated to health, elicit public support – a sustainable rate of health spending growth. I target the year 2035 as it is far enough out to represent the long-term but near enough to still be relevant. Following the established convention, I focus on tax revenues as a share of Gross Domestic Product (GDP). I follow three steps:
1) Estimate the share of GDP that the public is willing to provide in tax revenues;
2) Estimate the share of GDP that the public desires for federal spending on non-health items (the remainder represents what is available for federal health spending); and
3) Estimate the federal percent of national health expenditures in 2035.
If we know federal health spending in 2035 (step 2), and the portion of national health spending it represents (step 3), then we also know total health spending in 2035 that is consistent with a balanced budget. With the 2035 health spending target, we can solve for the average annual rate of growth that produces it – the sustainable growth rate.
My analysis relies heavily on CBO’s 2011 long-term budget outlook. While they reach to 2085, I use their 2035 projections provided under two scenarios:
- Extended-baseline scenario: adheres closely to current law.
- Alternative fiscal scenario: incorporates changes to current law that are widely expected to occur or that would modify some provisions of law that might be difficult to sustain for a long period. According to CBO, many analysts believe this scenario presents a more realistic picture of the nation’s fiscal policies. (See note 2)
1. What is the public willing provide in federal tax revenues?
Historically, tax revenues have averaged around 18 percent of GDP, but, due to high levels of unemployment, revenues in 2011 are estimated by CBO to be 14.8 percent of GDP. Under the extended-baseline scenario, this share is projected to grow to 20.8 percent in 2021 and 23.2 percent in 2035. The alternative fiscal scenario projects the share at 18.4 percent for both 2021 and 2035. Prior to the recession, in 2005-2007, revenues averaged 18.0 percent of GDP. (See note 3) I will consider a range of tax revenues between 18 percent and 25 percent.
2. How much of these revenues must be set aside for non-health care spending?
I assume that, by 2035, the nation must balance revenues with “primary” spending (excludes interest payments). (See note 4) The share of GDP available for federal health spending would thus equal the share raised in tax revenues minus that needed for non-health federal spending. Under the extended-baseline scenario, 2035 federal spending other than on the major health programs (Medicare, Medicaid/CHIP, ACA subsidies) is 13.9 percent, while under the alternative fiscal scenario it is 14.6 percent. I use the 13.9 percent figure to emphasize that non-health spending is being kept to a minimum in this analysis. The 13.9 percent includes health spending outside the major health programs, and also includes an offset for Medicare premiums. (See note 5) I shift these items into the health spending category which reduces non-health federal spending in 2035 to 13.5 percent of GDP. (See note 6)
The percent of GDP available for federal health spending in 2035 is equal to the percentage of GDP collected in tax revenues minus the 13.5 percent set aside to meet minimal requirements for other uses. If tax revenues are set to 18 percent, federal health spending is limited to 4.5 percent of GDP. (See note 7)
3. What is the federal share of total health spending in 2035?
The federal spending shares of GDP can be converted into national health expenditure shares if we know the federal health spending share of national health expenditures in 2035. In 2011, I estimate that federal health spending was 34.5 percent of total health spending. (See note 8) This share should be higher in 2035 due to the rapid increase in the federally-funded Medicare population, and the federal premium subsidies and federally supported Medicaid expansion under ACA. Under my baseline scenario, I estimate that the share in 2035 would rise to 41.5 percent, as shown in the bar chart below. (See note 9)
This baseline assumes that there is a single underlying cost trend (determined by payment and utilization rate trends) that affects all payers (Medicare, Medicaid, commercial insurers, etc.). Control of health spending is most feasible through systemic changes that affect all segments of the health care system. Also, this scenario assumes no changes in the share of Medicare beneficiary costs borne by the government, and no change in the Medicaid and exchange/subsidy share of spending by the non-Medicare population (following full ACA implementation).
Source: Altarum Center for Sustainable Health Spending Baseline Scenario
Results: “Sustainable” Health Spending Through 2035
The national health expenditure share of GDP in 2035 can now be derived as the federal health spending share of GDP divided by our estimate of the federal health spending share of national health expenditures (0.415). The table below illustrates these calculations. For example, when tax revenues are set to 18 percent, the federal health spending share of GDP is 4.5 per cent and the national health expenditure share is 10.8 percent.
In 2011, I estimate the health spending share of GDP was 18.0 percent. (See note 10) Assuming GDP growth at an average annual rate of 4.8 percent between 2011 and 2035, it is possible to estimate the sustainable growth rate in health spending for each level of tax revenues. (See note 11) Results are shown in the final two columns of the table, first relative to GDP, and second as a simple annual growth rate.
In the chart below, I present sustainable health spending relative to potential GDP (PGDP) to make historical comparisons more realistic. At the 18 percent tax revenue level, the sustainable health spending growth rate, starting today, would be PGDP minus 1.9 percent! For comparison purposes, the long-run growth rate has been PGDP plus 2.4 percent and, most recently, PGDP plus 0.8. Although this revenue level is the lowest that I show, it should not be interpreted as an outlier. It represents both the long run and the recent pre-recession average, and is also CBO’s alternative (more realistic) scenario. If the nation is truly serious about capping tax revenues at 18 percent of GDP, then it must face the consequences for health spending and, by extension, the health care system. Note that if health spending were to grow at the same rate as PGDP, often viewed as a laudable long term goal, we would still require tax revenues of more than 20 percent of GDP. Should we return to the “bad old days” when health spending grew at GDP plus 2 percent or more, we would be facing tax revenues of 25 percent or more of GDP!
The final chart plots sustainable health spending at its average annual rate as shown in the last column of the previous table. The sustainable rate of growth in health spending associated with the 18 percent tax revenue option is 2.6 percent. The most recent National Health Expenditure Accounts (NHEA) data show record low increases of 3.8 percent in 2009 and 3.9 percent in 2010. The Altarum Health Sector Economic Indicators show an increase of 4.5 percent for 2011 but trending to below 4 percent at year end. A 4.2 percent rate of growth would be sustainable if tax revenues were 20 percent. The ten year projection published by CMS last June showed a long term growth rate of 6.2 percent, which would be sustainable at a tax revenue target of about 24 percent.
If you have made it this far, I commend your perseverance, as this became longer and more technical than originally intended. In the next entry to this series, I will investigate other approaches to quantifying sustainable health spending and evaluate them within this framework.
Note 1. In an idealized version of ACA, individuals are protected by subsidies that are set to keep adequate insurance affordable and employers are protected in that if they drop coverage, their employees will still have access to affordable, adequate insurance. States are largely protected by the federal commitment to bear most of the cost of the ACA Medicaid expansion. If the federal government keeps Medicaid affordable to state governments, then the protection of other stakeholders (payers) is essentially complete. A given rate of growth in health spending is then sustainable if, and only if, the public is willing to provide the tax revenues needed to support it.
Note 2. CBO’s 2011 Long-Term Budget Outlook, June 2011, Summary, page x.
Note 3. CBO LTBO Table 6-2. Note that holding to the 18 percent figure over time has required periodic tax reductions since inflation (“taxflation”) and higher earnings would otherwise drive it up.
Note 4. Following a common convention, I exclude interest payments from federal spending requirements. If the primary budget is balanced, interest on the debt will continue to grow but, as a share of GDP, will fall if GDP growth exceeds the interest rate on the debt.
Note 5. The additional spending includes federal components of public health, medical research, structures and equipment, as well as federally operated health care systems for DoD, the VA and a variety of other items.
Note 6. This is computed as 13.9 percent, minus the 1.1 percent that is really health spending, plus the 0.7 percent offset attributable to Medicare premiums.
Note 7. For comparison, the current federal spending share of GDP is 6.2 percent. This is not, however, the figure you would find in the CBO documents because that includes just Medicare and Medicaid. It also differs from what you would compute from NHEA because I include the Medicare premium offset.
Note 8. I calculated the federal health spending share of GDP from the 2010 NHEA (released January 9, 2012), updated the estimates to 2011, and incorporated an estimate of the Medicare premium offset for 2011. The federal share of total health spending rises to 37.1 percent if the premium offset is not included.
Note 9. The growth in the Medicare population alone increases the share to 39.2 percent. The rest of the growth is due to the expanded coverage provisions of the ACA. This estimate was derived using a fairly simple model that employed my baseline scenario assumptions. Details are available upon request.
Note 11. The 4.8 percent growth rate in GDP assumes a long term growth rate in full employment GDP (“potential” GDP) of 4.5 percent, with an added 0.3 percentage points, because the 2011 starting point is well below full employment GDP. The 4.5 percent growth in potential GDP is based upon 2.3 percent real growth and 2.2 percent growth in the GDP deflator. These are roughly equivalent to the assumptions described on page 24 of the CBO’s 2011 Long-Term Budget Outlook report of June 2011.Email This Post Print This Post
Don't miss the insightful policy recommendations and thought-provoking research findings published in Health Affairs.