The Underlying Path of Health Spending in the U.S.

Monday, July 14, 2014

This blog is the third of three to highlight Altarum's upcoming Symposium, The Quest for ValueSign up today to attend in-person or via webcast.  Find the first and second blogs here.

Before addressing the underlying path of health spending, I would like to say a few words about our symposium on sustainable health spending that will be held on July 15 in Washington, DC. The agenda begins with a look at the latest data on health spending. Why has health spending seemingly decelerated in the first quarter of 2014 when expanded coverage and an economic recovery should be pushing spending up? Are there indications of a rebound in the second quarter? Have your questions ready for moderator Ceci Connolly and presenters Peter Orszag, Larry Levitt, and Stuart Altman.

Next we will examine approaches to increasing quality and value in health care delivery. For a sneak preview, see the recent blogs by Kavita Patel and Harold Miller, who, along with Kate Goodrich, will present in this session.

The third session focuses on the fascinating topic of the value of prevention (both clinical and environmental) and its potential role in slowing health spending. My prevention expert colleague George Miller will moderate this session, and we are eager to hear from presenters Alice Rivlin, Dave Chokshi, and Bobby Milstein.

We conclude the symposium with a discussion of what lies ahead for spending and value. Will health spending rebound from the historically low 4% growth rates seen since 2009, or have we reached a new normal? Our moderator for this session is Joanne Kenan. Presenters include myself; Paul Ginsburg; Michael Kleinrock; and, with the final word, Uwe Reinhardt.

I personally can’t wait to hear the exchanges, which, in past years, have included questions from audience members who could just as well be on the podium. Huge thanks to my colleague Paul Hughes-Cromwick for pulling this all together and to the Robert Wood Johnson Foundation for their ongoing support.

Finally, if you happen to be reading this blog after July 15—which is somewhat likely, since it is being posted on the 14th—we will be releasing a video of the proceedings and an edited monograph.

My part of the proceedings will focus on the underlying path of health spending. By this I mean the path about which actual health spending oscillates but returns to over extended periods of time. If we can identify this underlying path, we can do a better job of predicting the future from a noisy history. This underlying path can also serve as the curve to be monitored for evidence of any “bend.”

Inspired by Gene Steuerle’s work, I have been looking at something analogous to what economists might call the nation’s marginal propensity to consume health care. Specifically, it is the share of the growth in real per capita gross domestic product (GDP) that goes toward health spending. I’ll refer to it as the marginal propensity to spend on health (MPSH).[i] There is a lot of noise in the historical MPSH series, mostly because health spending does not respond immediately to recessions and recoveries. For example, during recessions, when the growth in real per capita GDP is near zero, health spending still grows and the MPSH can be nearly infinite. During recoveries, the MPSH will be abnormally low as real per capita GDP accelerates faster than health spending. Sometimes the noise is due to unusual changes in health spending rather than changes in GDP. The most obvious example is the managed care era of the late 1990s, in which health spending grew at historically low levels. This was followed by 3 years of managed care “backlash,” in which health spending grew at an unusually high rate. Neither period was truly indicative of the underlying path of health spending.

In order to eliminate some of the noise associated with recessions and recoveries, I have plotted real per capita health spending against real per capita potential GDP (rather than real per capita GDP) from 1965 through 2007, the year before the most recent recession.[ii] The slope of the line represents the MPSH. What stands out is the relatively constant slope of 0.26 between 1980 and 2007, a period of 27 years.[iii] The managed care era of the late 1990s pulls actual spending below the line for a few years but this effect is reversed by the managed care backlash of the early 2000s.

Marginal Propensity to Spend on Health

This relative stability in the MPSH from 1980 through 2007 opens the door for identifying the underlying path of health spending during this period. Growth in real per capita potential GDP averaged about 2% between 1980 and 2007, with peaks and troughs but no discernible overall trend up or down. The underlying path in the health spending growth rate can be computed by applying the 26% MPSH on health to a uniform 2% growth in the economy, as shown in the chart below.[iv]

Underlying Health Spending Growth Rate

Note that the underlying rate of growth in health spending falls steadily over time, even though both the MPSH and the economic growth are constant. In 1981, the underlying health spending growth rate is 6%; by 2007, it is just above 3%. This translates into “excess” growth of 4% in 1981 and 1.3% in 2007 and suggests that holding the MPSH constant was sufficient to achieve this major reduction in excess growth.[v]

At the symposium, I plan to project this underlying health spending growth rate forward so that the actual rate of growth in health spending can be compared to its underlying path. This will set the stage for some speculation about the impact of the recession on health spending and the extent to which the ongoing recovery is likely to push spending up.


[i] Official national health expenditures are not limited to consumption spending.

[ii] Potential GDP refers to what the GDP would have been if the nation were at full employment. Thus, it eliminates noise associated with business cycle fluctuations around full employment. I use the most recent series from the Congressional Budget Office. I am purposely stopping at 2007, since I wish to establish the underlying path for health spending just before the recession.

[iii] From 1965 through 1980, the slope was much lower at 0.15. Note that 26% is a bit lower than the 30% figure that I presented last year. The difference is partly due to the government’s upward revision to the historical GDP series last July and partly due to a difference in methods.

[iv] Think of the 2% growth rate in real per capita GDP as “permanent” income growth as defined by Milton Freidman. The underlying growth path for health spending is then defined as 26% of the growth in permanent income.

[v] Excess growth refers to health spending growth in excess of GDP growth. This is commonly denoted as health spending growth of GDP+4 in 1981, falling to GDP+1.3 in 2007.

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.



Subject Matter Experts

Areas of Expertise