The causes of the disinflation
Whether very low inflation is problematic depends in part on why it is low. Two complementary approaches can be used to determine the source of the falloff. The first might be characterized as a top-down approach: focus on overall inflation and disentangle the causes of the fall. The second might be characterized as a bottom-up approach: focus on more detailed measures of inflation and identify individual factors that pulled down important inflation components.
Our recent research using these approaches points to several key forces at play in the 2013 disinflation. First, the sluggish pace of the US economic recovery broadly helped to limit most price pressures. Second, enough slack remains in labor markets to keep the growth rate of labor costs very low by historical standards. Third, there is little pressure on consumer prices coming from commodities such as energy. Finally, some special, temporary forces—such as a deceleration of medical care inflation associated with changes in laws—have put some short-lived downward pressure on inflation.
The top-down approach
One way to implement the top-down approach is to use a forecasting model that characterizes the relationships among inflation, economic activity, labor costs, import prices, energy prices, and monetary policy. We use one such Cleveland Fed model to assess the contributions of various forces to the decline of core PCE inflation from the second quarter of 2012 through the fourth quarter of 2013.
Three-fourths of the fall in core PCE inflation can be explained by the behavior of the model’s inflation determinants.
Approximately three-fourths of the fall in core PCE inflation can be explained by the behavior of the model’s inflation determinants. Variables that directly capture economic activity—GDP, employment, and unemployment—play the largest role. On balance, over the 2012 and 2013 period, GDP and employment grew more slowly than expected, while unemployment fell more slowly than expected. These shortfalls put downward pressure on inflation. Overall, real economic activity accounts for about one-half of the fall in inflation. Labor costs, import prices, and energy prices account for another one-fourth or so.
The remaining one-fourth of the decline in inflation since early 2012 cannot be explained by the model’s inflation determinants. Instead, this portion of the decline in inflation is the result of unexpected, temporary events that are specific to inflation. Some of these forces are evident from a more bottom-up analysis.
The bottom-up approach
One way to implement the bottom-up approach is to split the price indexes into broad components. We split inflation into several parts—food and energy prices, core goods, and core services—and look at their recent behavior and key drivers.
A key driver of overall inflation is volatile movements in food and energy prices. The results of the decomposition show that for much of 2012 and 2013, inflation has been held in check by flat commodity prices. For example, retail gasoline prices trended down for much of last year (figure 5). This trend helped pull overall inflation below core inflation at times.
For much of 2012 and 2013, inflation has been held in check by flat commodity prices.
Inflation in core PCE goods prices, which exclude food and energy goods, has fallen by about 2 percentage points since early 2012 (figure 6). In an accounting sense, goods comprise approximately one-fourth of the core basket of consumer spending. Thus, the deceleration in goods’ prices implies a reduction in core PCE inflation of about one-half of a percentage point, all other things equal.
Because a large number of goods are either imported to the United States or have some imported content, there is a significant connection between import prices and goods prices. Over the last few years, the deceleration of core goods prices has been driven in part by an even sharper deceleration in prices of imported goods (figure 7). The falloff in import prices likely reflects slow growth in the global economy and the strength of the dollar.
Compared with inflation in goods prices, inflation in services prices has been much more stable. However, even core PCE services inflation has slowed since early 2012, from 2.3 percent to 1.9 percent at the end of 2013 (figure 8). Services comprise about three-fourths of the core PCE basket, which means that this more modest reduction in services inflation implies a roughly one-third percentage point reduction in core inflation.
Because the primary cost in the provision of services is labor, the low rate of services inflation is likely attributable to the historically low rate of growth of labor costs that has prevailed since the end of the recession (figure 9). From the mid-1990s through 2007, labor costs as measured by the Employment Cost Index (ECI) increased at an average rate of more than 3 percent. But since 2009, the ECI has been rising only about 2 percent per year.
Using CPI instead of PCE inflation measures paints a mostly similar picture, with a few key differences. Since early 2012, inflation in the CPI for core goods has fallen sharply, in lockstep with the corresponding PCE-based measure (figure 6). However, inflation in the CPI for core services has edged down only slightly since early 2012, while the PCE-based measure slowed more significantly (figure 8).
Cleveland Fed research has attributed the widening gap between PCE and CPI services inflation to several forces. One important source of the gap is shelter costs, which have a larger weight in CPI services and whose inflation rates have been running above those of some other components of services. Putting a relatively large weight on a component with a relatively high rate of inflation—even if the rate of inflation is not high in an absolute sense—causes inflation in the services component of the CPI to run above inflation in PCE services.
Another important source of the CPI–PCE gap in services inflation is medical care costs and the larger weight they receive in PCE services. Inflation in medical care costs as measured by the PCE index has slowed sharply, partly as a result of downward pressures on Medicare prices associated with the Affordable Care Act. As a result, the deceleration of medical care costs has put more downward pressure on PCE services inflation than on CPI services inflation.