The Bureau of Labor Statistics of the U.S. Department of Labor today reported revised productivity data--as measured by output per hour of all persons--for the second quarter of 2002. The seasonally-adjusted annual rate of productivity growth in the second quarter was 1.5 percent.

The increase in productivity is larger than the 1.1% reported on August 9, due primarily to upward revisions in the level of output. The productivity number referred to is that of the non-farm business sector, the most commonly used gauge of productivity.

The original press release is available at:

Definitions of Productivity

The broadest measure is productivity in the business sector, which comprises 77 percent of GDP. The business sector excludes government and nonprofit organizations, employees in private households, and the rental value of owner-occupied housing. The non-farm business sector excludes all of those activities plus farming and accounts for about 76 percent of GDP. Productivity in the non-farm business sector is the most commonly used measure in studies of productivity. The reason agriculture is removed is because output and therefore productivity are significantly influenced by weather changes.

Importance of Productivity Changes For Economic Growth

Our capacity to produce goods and services is determined by how much labor we have, how many hours workers work, the workers' skills and intensity of work, the amount of capital workers have with which to work, and changes in technology. Over time, real GDP in the U.S. has increased for all of these reasons. We have a larger population with a larger percentage working. In the last ten years, the average worker has been working longer hours. The workers have significantly larger amounts of capital and new ways of producing and organizing production have been put in place.

The productivity measures capture the effects of the increased capital, the increased experience and education of workers, and the new technology. If productivity increases faster than population growth, real GDP per person can increase and we can all enjoy higher standards of living.

Current Reactions

Productivity growth most often slows or disappears in recessions and indeed the growth in productivity throughout 2001 (1.9%) was significantly slower than in 2000 (3.3%). In fact, productivity actually fell in the first two quarters of 2001. The rather rapid increase in the fourth quarter of 2001 (7.3%) and the even faster rise in the first quarter of 2002 (8.6%) appeared to many as evidence that the economy was likely coming out of the recession. Productivity normally grows rapidly following a recession as businesses expand output without adding significant numbers of employees.

The second quarter 2002 increase in productivity (1.5%) is sharply lower than the previous two quarters, but still represents positive growth. The upward revision from the preliminary numbers due to increased output can be seen as further evidence that the recession that began in March of 2001 has ended.

Data Trends

Figure 1: Changes in Nonfarm Business Productivity (Percent Change in Output per Hour) Quarterly Changes 1998-2002

The latest productivity data indicate that businesses are continuing to adapt to the slowdown in growth in total output that occurred throughout 2001 by adjusting employment and the hours worked. However, now it appears that output is beginning to rise once again. Employment decreased during 2001 and businesses reduced the hours worked by each employee as well. (See the latest Unemployment case for a description of recent changes in employment.)

In the first quarter, output in the non-farm business sector increased slightly (.8 % at an annual rate) while hours of employees decreased (-.7%). Increases in output alone will increase productivity. The fact that .7% fewer hours were used to produce that output also increases the measured productivity. Thus, the change in productivity was .8% + .7% = 1.5%.

The changes in productivity during 2001 were less than the changes in the previous three years. The increase in productivity for all of 2001 (1.1 percent) is less than increases in 2000 (annual average of 2.9 percent), 1999 (2.4 percent) and 1998 (2.6 percent). This is the lowest rate of productivity growth since 1995, but for the last four quarters it is still higher than the average rate of change in productivity over the previous twenty years. The longer run trend in productivity over the past decade has allowed real GDP per capita to increase. It also means that wages for workers can increase and can do so without excessive upward pressures on prices.

Hourly compensation rose at an annual rate of 3.7 percent during the quarter. Unit labor costs are the costs of labor per unit of output. Thus the increase in unit labor costs is the percentage increase in hourly compensation minus the percentage increase in productivity. Or 3.7% minus 1.5% = 2.2%. %. (The numbers do not exactly add up to the reported number, 2.1%, again due to errors in rounding).

The increase in unit labor costs provides evidence that slight inflationary pressures may exist. The recent relatively high and increasing unemployment rates provide contrary evidence that inflationary pressures may not be a concern.

Historical Data Trends

Figure 2: Changes in Nonfarm Business Productivity (Percent Change in Output Per Hour) Annual Changes 1970-2001

From 1950 to 1973, productivity grew at an average annual rate of 2.8 percent. But from 1973 to 1995, growth in productivity slowed to an increase at an annual rate of 1.4 percent. From 1996 to 2000, productivity increased at an annual rate of 2.5 percent, almost equal to the 1050 to 1973 rate.

The slowdown, beginning in the 1970s, and the increases in the late 1990s are not fully understood. The analysis of the Council of Economic Advisers is that about .47 percent of the recent increases can be explained by the effects of more computers and software being used in many businesses. Dramatic changes in the production of computers themselves helps explain about another .23 percent. The quality of labor (increased education and more experienced workers) explains about .05 percent.

The rest is not understood. It may be due to cyclical pressures (that is, fewer workers were being added to employment rolls, but those who were working were producing more) and perhaps to the effects of lower business costs as a result of business use of the internet.

For the future, education and experience will not likely continue to make significant advances. The computer contribution to increases in productivity will probably drop. A consensus forecast is for a declining growth rate in productivity and therefore in real GDP growth rates.

  1950 -73 1973 - 95 1995 - 00 Future
Growth in hours worked 1.6% 1.7% 1.7% 1.2%
Productivity 2.8% 1.4% 2.5% 2.0%
Real GDP 4.2% 3.0% 4.0% 3.0%

(The numbers do not add in all cases due to rounding and the inclusion of slightly different measures in the productivity and hours calculations. However, increases in the growth in hours worked and increases in productivity will cause similar increases in the growth rates of real GDP.)

How The Data Are Calculated

Productivity data represent the amount of goods and services (in real terms) produced per hour of labor. They do not identify the separate contributions of labor, capital, and technology. Changes in productivity include the effects of all (except hours of work) possible influences on output – technology, ability, skills, and effort of labor, capacity utilization, managerial skills, and the amount of capital.

Other periodic announcements report multi-factor productivity indexes, which do measure the separate effects of hours of labor, education levels and experience of labor, amount of capital, and the effects of technology change.

  1. Assume that hours worked increase by 1 percent and that GDP increased by 5 percent. Also assume that the GDP implicit price deflator increased by 2 percent. Calculate the increase in productivity.
  2. Suppose the labor force increases by 1 percent and the hours per worker increase by 1 percent. Also assume that increases in the amount of capital increases output per worker by 2 percent. Prices increase by 3 percent. Increases in education and technology increases output per hour by 1 percent. What are the rates of increase in real GDP and GDP?
  3. An increase in government spending on research and development may enhance abilities to produce goods and in effect increase productivity. Is such spending wise?
  4. Why is productivity rising when the economy is not growing very rapidly?