Thursday, April 5, 2012

If you have followed Parts One and Two of this series, you will be aware of the key role played by federal non-health spending in my calculation of the sustainable growth rate in national health expenditures. Given the expanded coverage provisions of the Affordable Care Act, I argued that the sustainable rate of growth in health spending is largely determined by what the nation chooses to allocate to federal health spending in future years. And this allocation is simply the difference between what the nation is willing to provide in total tax revenues and the amount of those revenues being set aside for non-health federal spending.

For example, suppose the nation wishes to keep tax revenues at 18 percent of GDP and set aside 13.5 percent of GDP for non-health spending (under a balanced primary budget). (1) Then the amount available for federal health spending would be 4.5 percent of GDP.  Using 2035 as the target year for bringing federal health spending in line with this target, I calculated that the corresponding growth rate in national health expenditures would be 2.6 percent annually, starting in 2012. This is 2.2 percent below the expected GDP growth rate over this period, and lowers the health spending share of GDP to 10.8 percent by 2035—a stark contrast to the 2011 health spending growth rate (4.5 percent) and share of GDP (18.1 percent).

I chose tax revenues at 18 percent of GDP in the example above because that is both the long run historical national average and also the average just prior to the recent recession (see below).  The 13.5 percent of GDP set aside for federal non-health spending in 2035 was my calculation using the June 2011 Congressional Budget Office’s 2011 Long Term Budget Outlook. (2)
Graph of federal tax receipts as percentage of GDP 
Source: Office of Management and Budget

The Ryan Path for Non-Health Spending

CBO has released a new report presenting the federal non-health spending path specified by Congressman Paul Ryan (R-WI), the Chairman of the House Budget Committee. This path puts non-health spending at a significantly lower share of GDP than the CBO extended baseline and I thought it would be interesting to examine its impact. (3) To be clear, this is not a model of the full Ryan budget proposal. Rather it is a model showing the rate of increase in health spending that would be sustainable under the assumptions of my model if federal non-health spending followed the path specified under the Ryan budget as estimated by CBO. (4) As discussed below, one of the distinctions between the Ryan budget and my model is that Ryan assumes the ACA is repealed.

The chart below is from Table 2 of the CBO Ryan report and shows projected federal spending, exclusive of major health programs, social security and interest as a share of GDP. (5) Social Security spending is excluded from this chart because the Ryan path is the same as the extended baseline. (6) The Ryan path starts 1.25 percentage points lower than the CBO extended baseline in 2023. Beyond 2023, the share of GDP under the Ryan path falls steadily because spending growth is tied to the GDP deflator, i.e., it is held constant in real terms. On the other hand, the share of GDP stays nearly constant under the extended baseline scenario. 

Graph of federal spending with exclusions

Source: Congressional Budget Office

Implications for Sustainable Health Spending

From these data, I estimate that the federal non-health spending share of GDP in 2035 under the Ryan path is 10.7 percent. (7) Returning to the earlier example, if federal revenues are held to 18 percent of GDP, this would leave 7.3 percent of GDP available for federal health spending in 2035. Under my model, the sustainable growth rate in health spending would be an annual rate of 4.7 percent starting in 2012—very close to the projected rate of increase in GDP.

It is worth noting that the CBO extended baseline non-health spending estimates in this latest report are a bit lower than those given in the June 2011 Long-Term Budget Outlook Report that I used in my earlier work. Using these newer extended baseline figures, the federal non-health spending share of GDP falls from 13.5 percent to 13.1 percent in 2035. This increases the sustainable rate of health spending a bit compared to the results presented here.

The chart below compares sustainable growth rates under the non-health spending assumptions of the new CBO extended baseline and the Ryan path. The differences are quite large. With tax revenues at 18 percent, the extended baseline sustainable health growth rate is 1.8 percentage points below GDP growth, but under the Ryan path it is nearly equal to GDP growth. Tax revenues close to 21 percent are required to bring the sustainable growth rate up to the GDP growth rate under the extended baseline.

Graph of sustainable growth in health spending by path

Source:  Altarum Center for Sustainable Health Spending


My goal in this series of essays is to develop a useful definition of sustainable health spending that can be quantified and used as a target. To get closer to an actual target, there needs to be some agreement about long-run tax revenues and amounts needed for federal spending on non-health items. (8) Of course, there is no such agreement on the horizon. But I think we can begin to bound the problem.

I see the Ryan path for non-health spending as a bare minimum, and the CBO extended baseline as a more reasonable minimum. Since the Ryan budget specifies federal revenues at 19 percent of GDP, I would also take this as a minimum. I think there is recognition among fiscal conservatives that changing demographics mean we cannot get by with the long run average of 18 percent, and the Ryan budget provides some confirmation. I suggest an upper limit of 21 percent for tax revenues.

This is based, in part, upon some discussions I heard at the outstanding National Association for Business Economics conference on March 26. Alan Blinder allowed that he would be willing to split the difference between the CBO extended baseline revenues (24 percent of GDP) and the alternative fiscal scenario revenues (18 percent of GDP) – which is 21 percent. Lawrence Lindsey, speaking for the conservative side, laughed and said he would split it again – and that is just above the 19 percent in the Ryan proposal.

Given these boundaries, the (maximum) sustainable rate of health spending would lie somewhere between the 3.8 percent and 6.2 percent figures shown in the chart. Coincidentally, this averages to 5.0 percent – very close to my assumed GDP growth rate of 4.8 percent. (9)  I note that the Health Sector Economic Indicators briefs produced by our center show health spending growth well within this range for 2011 (4.5 percent) and actually dropping below a 4 percent annual rate by year end. In terms of the health spending share of GDP, these boundaries translate into a low of 14.2 percent of GDP and a high of 24.7 percent with a mid-range of 19 percent.

I should reiterate that this is not a model of the Ryan budget. Rather, it incorporates one element of his budget into a model that is otherwise very different from his budget. The most obvious difference is his exclusion of ACA, but there are others. Notably, by 2035 his revenues exceed spending exclusive of interest by about 2 percentage points of GDP while my model balances the primary budget by 2035 (revenues equal to spending exclusive of interest payments). In effect, he uses his cuts in non-health spending to reduce the debt while in our model these cuts lead to an increase in the sustainable growth rate in health spending.

We are all, of course, awaiting the Supreme Court verdict on ACA. Should the expanded coverage provisions of the ACA be struck down, however, we will need to revise our method for estimating the sustainable rate of growth in health spending. We do not consider spending sustainable if it fails to address the large, and growing, numbers of uninsured. And there is currently no back-up plan for expanded coverage should those elements of ACA be eliminated.


1. The primary budget does not include interest payments on the public debt. Throughout this report, federal spending refers to non-interest spending.
2. CBO provides two different estimates for future federal non-health spending. I used the smaller figure (from their extended baseline scenario) because I wanted the maximum sustainable growth rate in health spending, and this corresponds to setting aside minimum acceptable amounts for federal non-health spending. The June 2011 CBO report projects federal spending, excluding major mandatory health programs and interest, to be 13.9 percent of GDP. This figure includes the Medicare premium offset as well as some other health spending that is included in the national health accounts. I removed these items to be consistent with my definition of health spending with the net effect of reducing this figure to 13.5 percent of GDP.
3. The report also includes interim updates to the CBO scenarios. Quoting the report: “The extended baseline scenario and extended alternative fiscal scenario reflect projections through 2022 from Updated Budget Projections: Fiscal Years 2012 to 2022 (March 2012) extrapolated into future years using rates of interest and growth rates for revenues and spending from CBO's 2011 Long-Term Budget Outlook (June 2011).”
4. The assumptions of our model include implementation of the expanded coverage provisions of the ACA, no change in the particulars of the Medicare program, and a balanced primary budget.
5. This spending consists of national defense, other discretionary spending (e.g., education, transportation, income security, veterans’ health care, and homeland security), and other mandatory spending (e.g., unemployment compensation, federal civilian and military retirement benefits, “food stamps,” veterans’ benefits, and other income security programs). I do not have data on the Ryan proportions, but for the extended baseline the breakout is: defense (just over 40 percent); other discretionary (just under 40 percent); and other mandatory (about 20 percent).
6. Social security spending is estimated at 5.5 percent of GDP in 2023 and 6.0 percent of GDP in 2030 and 2040 for both the extended baseline and the Ryan path.
7. Spending exclusive of major health programs, social security, and interest is about 5.25 percent of GDP in 2035. I estimate that about one-tenth of this spending is for items included in the national health accounts (e.g., research, public health, military and veteran health costs). Thus the non-health share is reduced to 4.7 percent of GDP. Adding in social security spending at 6 percent of GDP gives the 10.7 percent figure.
8. Not to mention a few key assumptions about the ACA.
9. I project full employment (potential) GDP to grow at 4.5 percent, but, since we are starting at less than full employment, catch-up by 2035 adds 0.3 percentage points to actual GDP.

All postings to the Health Policy Forum (whether from employees or those outside the Institute) represent the views of the individual authors and/or organizations and do not necessarily represent the position, interests, strategy, or opinions of Altarum Institute. Altarum is a nonprofit, nonpartisan organization. No posting should be considered an endorsement by Altarum of individual candidates, political parties, opinions or policy positions.