Federal Reserve Economic Data

The FRED® Blog

Measuring labor costs

Some economic analysts are looking for signs of faster wage growth (labor costs). In their view, faster wage growth is a sign of building inflation pressures. In October’s employment report (released Nov. 7), average hourly earnings of production and nonsupervisory employees on private nonfarm payrolls rose by only 2.2 percent over the past 12 months. A rather modest increase. Although closely followed, this series excludes most employee benefits, such as employer-paid health insurance and retirement benefits.

Broader measures that better account for these benefits include the employment cost index (ECI) and compensation per hour (CPH) in the business sector, both published by the Bureau of Labor Statistics. In the third quarter of 2014, the ECI increased by 2.3 percent over the past four quarters, while the CPH increased by 3.1 percent. But these two series are also incomplete: The reason is that businesses tend to care more about unit costs: that is, costs of labor and non-labor inputs adjusted for productivity changes. For example, if compensation is increasing solely because of faster gains in worker productivity, then unit labor costs will be unchanged and a firm’s profit margins will be largely unaffected. This can be seen in the graph. After the past recession, compensation per hour was increasing, but because labor productivity was increasing by a larger amount, unit labor costs were falling.

In the productivity and costs report released earlier this month, the Bureau of Labor Statistics reported that unit labor costs in the business sector had increased by 2.4 percent in the third quarter from a year earlier. The modest acceleration in unit labor costs over the past three quarters reflects, on net, slower growth in labor productivity and slightly faster growth in compensation per hour.

How this graph was created: Search for “Nonfarm Business Sector: Unit Labor Cost.” In the “Edit Data Series” function, change the units to “Percent Change from a Year Ago.” Repeat the process by adding these series: “Nonfarm Business Sector: Compensation Per Hour” and “Business Sector: Real Output Per Hour of All Persons” (labor productivity).

Suggested by Kevin Kliesen

View on FRED, series used in this post: HCOMPBS, OPHPBS, ULCBS

Bank failures

The previous recession was clearly associated with substantial problems in the financial sectors. As the graph shows, there has been a significant number of bank failures, as recorded by the Federal Deposit Insurance Corporation (FDIC), which is responsible for managing the closure process and insuring depositors. The number of failures, however, is nowhere near the peak around 1989, the time of the savings and loan crisis. The recession around that time involved different financial problems and thankfully was much less deep than the previous recession.

How this graph was created: Search for “bank failures” and then change the graph type to “Area” under graph settings in the graph tab.

Suggested by Christian Zimmermann

View on FRED, series used in this post: BKFTTLA641N

What’s the “normal” unemployment rate?

As the U.S. unemployment rate inches down, it seems reasonable to ask when it will be back to normal. One measure of “normal” is the natural rate of unemployment, sometimes referred to as NAIRU, published by the Congressional Budget Office. This measure is meant to contain all relevant information except for cyclical factors in the unemployment rate. Thus, when there is no difference between the NAIRU and the standard unemployment rate, the standard unemployment rate should be back to normal. Note that the natural rate is calculated, not measured, and thus is subject to the assumptions made. Some of those assumptions relate to whether structural factors should be taken into account. This question led (temporarily) to two different natural rates during the previous recession.

How this graph was created: Search for NAIRU, select both series, and add them to a graph. Then add the civilian unemployment rate. Finally, change the end date to the current date.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: NROU, NROUST, UNRATE


Back to Top