For the first time since the mid-1990s, the United States has seen significant reduction in the growth of health expenditures. A critical question for policy makers and the health industry is whether the slowdown is sustainable.

Our newly published Health Affairs article with Marty Gaynor, David Newman, and Robert Town  puts the reduction into the context of a long arc of cost growth in employer-sponsored insurance (ESI), the largest source of coverage in the nation. Specifically, the article, “Trends Underlying Employer-Sponsored Health Insurance Growth for Americans Younger than Age 65”, provides a comparison of the recession to initial recovery numbers on medical and prescription trends for people with ESI. The data came from the Health Care Cost Institute (HCCI).

During and immediately after the recession, we found that ESI spending per person grew at an average annual rate of 4.9 percent, with spending growing slower in 2010 and 2011 than during the recession. We found the same pattern for medical and prescription spending, although medical spending grew faster than prescription spending in all years.

We looked at the underlying trends for medical and prescription health spending – examining differences in spending for consumers and insurers. We determined that out-of-pocket medical spending was recession-proof, as consumers bore more of the medical bill in 2011 than in 2007. In contrast, consumers paid a smaller share of prescription drug spending in 2011 than in 2007.

We also compared our findings to the relevant literature on prices and utilization. We found utilization grew more slowly for medical and prescription services after 2009. As a result, most of the growth in spending came from rising health care prices, although the directions of medical and prescription price trends were different.

Adding 2012 Data To The Picture

The question remains, is the United States experiencing a sustainable reduction in health care expenditures in the ESI insurance market? Fortunately, HCCI now has 2012 data that were not available at the time the Health Affairs article went to press.

The new HCCI report presents revised health care cost benchmarks primarily from 2010 to 2012. One important change is a revision of spending growth for 2009 up to 6.3 percent, and down for 2010 and 2011 to 2.9 percent and 4.1 percent, respectively. Why the change? Revisions are a natural consequence of using claims data, which is usually not “final” (that is to say, paid and adjudicated), for up to two years. But generally, there were few top-level health spending trend changes for people with employer insurance between 2011 and 2012.

We had found that ESI spending per person grew at relatively slow rates between 2007 and 2011, and 2012 continued this trend. Where we estimated in the article that the average annual spending growth was about 4.9 percent per year, with the addition of the 2012 data, we now estimate that the annual average rate of health spending between 2007 and 2012 was 4.6 percent.

We had also found that ESI spending per person outpaced the GDP per capita and the growth of the National Health Expenditures (NHE). The new HCCI report confirms these findings, except for 2010 when ESI growth matched GDP and grew slower than the NHE.

The new data also confirms our finding that spending during the economic recovery grew slower than spending during the recession. We estimated in our paper that spending per person grew at 5.8 percent per year between 2007 and 2009. In that same period, medical spending grew by 5.9 percent and prescription spending grew by 4.8 percent. According to the new HCCI data, during the economic recovery (2010 -2012), ESI spending grew at an average annual rate of 3.7 percent. During the recovery period, ESI medical spending grew by 3.9 percent per year, while prescription spending grew by 2.7 percent per year. In all years, medical spending growth outpaced prescription spending growth.

We had observed for 2007-2011, and now can confirm for 2012, that out-of-pocket medical spending grew faster than insurer medical spending growth. We estimated in our paper that during the recession, medical spending out of pocket grew at about 7.2 percent per year, whereas insurer spending grew at 5.7 percent. The new data suggests that medical spending out-of-pocket grew at an average rate of 7.6 percent between 2010 and 2012, and that insurer spending slowed to an average annual rate of 3.3 percent per year.

Unlike medical services, prescription out-of-pocket spending growth was slower than insurer spending growth. As observed in the Health Affairs paper, prescription spending grew at 6.3 percent per year during the recession. The new HCCI data suggests the insurer portion of prescription spending grew by 2.7 percent per year during the recovery. The 2012 data confirms our previous finding that the average annual growth in out-of-pocket spending per person on prescriptions remained flat between 2007 and 2012, due to declining growth after 2010.

The Overall Picture: Fundamental Change Persists

One advantage of the HCCI data is that it contains the underlying utilization, price paid, and intensity of services that can be used to determine the drivers of health care spending. The new 2012 data allow a restatement of some of the utilization, price, and intensity trends for 2010 and 2011. The revised data did not change the long run (2007-2012) utilization or price trends for prescriptions or medical services. Some interesting shifts did occur in 2012 such as the rise in generic prescription and outpatient facility prices and the rise in professional procedure use.

Our Health Affairs paper points out the important fact that slower aggregate spending growth was not reflected in the medical spending by consumers. These patterns continued in 2012. That a third year of recovery has resulted in slower health care cost growth for over 50 percent of the country suggests that substantial, continuing effects have persisted since the end of the recession. Although neither our paper nor the new HCCI report offer enough evidence to a single cause, we believe we have at least pinpointed when the changed occurred, and whom it has affected. We intend to continue examining the HCCI and other data to explore further the implications of this historic slowdown in health care spending.