During the past six months, spending on health care in the U.S. has accelerated substantially, new data on gross domestic product suggest. Naturally, that's led to some breathless commentary suggesting the era of low health-cost growth we’ve been enjoying for the past few years is over. In truth, the case for that conclusion is pretty weak. But that doesn’t mean there's no cause at all for concern. We are missing a crucial opportunity to make sure that the recent slowdown never ends.
After growing less than 3 percent in 2011 and 2012, personal expenditures on health care jumped by 6 percent on an annualized basis in the fourth quarter of 2013, and then 10 percent in the first quarter of 2014.
In assessing these numbers, there are four things to keep in mind. First, as health-insurance coverage expands under the Affordable Care Act, some temporary increase in cost growth is to be expected. People get more health care when they're insured than when they're not. About 8 million people have signed up for coverage through the public exchanges created by the ACA, and another 5 million have enrolled in Medicaid. A reasonable estimate is that about half of these people were previously uninsured, so the insured population has risen by about 6 million, or 2 percent to 3 percent. If all of that increase were to occur within one quarter, and if the increase in costs were proportional to the increase in coverage, the annualized growth rate for that quarter would be in the range of 10 percent. Such an increase, though, would be a one-time event -- and would tell us little about the underlying utilization trend.
Second, the GDP numbers are preliminary. Under the Bureau of Economic Analysis’ methodology, a substantial share of the health-care data in the advance estimate of GDP, which is all we have for the first quarter of 2014, comes from what the bureau calls a “judgmental trend.” In other words, an educated guess.
These guesses are often revised substantially. Consider as an example the third quarter of 2011: The advance estimate suggested an increase in health-care spending of more than 5 percent on an annualized basis. After revisions, the final number showed a decline of almost 1 percent. Substantial caution should thus be associated with the preliminary numbers for the first quarter of 2014, especially since the Bureau of Economic Analysis specifically noted that the effects of health-insurance expansion are still not fully accounted for.
The third thing to keep in mind is that other indicators suggest the slowdown in use of health care is continuing. Medicare spending, for instance, is a more insightful barometer of the underlying trend because it isn't substantially affected by the economy and because coverage increases at a fairly steady rate. Also, Medicare spending data are not estimates; they're based on money out the door. And in the first half of this fiscal year, Medicare spending has risen by only 0.6 percent over the previous year.
Finally, employment in health care hasn't grown rapidly, as it would if health-care companies expected an end to the era of slow growth. The jobs report for April released on Friday showed that job growth in the sector remains low, with particular weakness among hospitals. So either health-care workers' productivity somehow rose spectacularly beginning in the fourth quarter of 2013, employment in health care is about to boom, or the spike seen in the GDP data won't continue in future quarters. My money is on the third option.
We could still learn a useful lesson from this recent scare. In particular, as I have argued, much of the recent deceleration has been caused by expectations among health-care providers that the system is shifting to payments based on quality of care rather than quantity. Unfortunately, because policy makers are not moving quickly on this front, we run a real risk that providers' expectations will not soon be fulfilled. And if those expectations ever come crashing down, we will have lost the opportunity to lock in the recent slowdown.
To contact the writer of this article: Peter Orszag at firstname.lastname@example.org
To contact the editor responsible for this article: Mary Duenwald at 1+212-617-8620 or email@example.com