Issues
In Focus.
Outlook for Labor Productivity Growth
The AEO2004 reference case economic forecast
is a projection of possible economic growth, from the short term
to the longer term, in a consistent framework that stresses demand
factors in the short term and supply factors in the long term [33].
Productivity is perhaps the most important concept for the determination
of employment, inflation, and supply of output in the long term.
Productivity is a measure of economic efficiency that shows how
effectively economic inputs are converted into output.
Advances in productivitythat is, the ability
to produce more with the same or less inputare a significant
source of increased potential national income. The U.S. economy
has been able to produce more goods and services over time, not
only by requiring a proportional increase of labor time but also
by making production more efficient. To illustrate the importance
of productivity improvements, on the eve of the American Revolution,
U.S. gross domestic product (GDP) per capita stood at approximately
$765 (in 1992 dollars) [34]. Incomes rose dramatically over
the next two centuries, propelled upward by the Industrial Revolution,
and by 2002 GDP per capita had grown to $30,000 (1992 dollars).
Productivity improvements played a major role in the increase in
per capita GDP growth.
Productivity is measured by comparing the amount
of goods and services produced with the inputs used in production:
- Labor productivityoutput per hour of all personsis
the ratio of the output of goods and services to the labor hours
devoted to the production of that output; it is the most commonly
used productivity measure. Labor is an easily identified input
to virtually every production process. For the U.S. business sector,
labor cost represents about two-thirds of the value of output
produced. Increases in labor productivity allow for comparable
gains in profits and/or compensation without putting upward pressures
on output prices. When labor productivity grows, the economy is
able to produce more with the same number of workers.
- Multifactor productivity reflects output per unit of
some combined set of inputs. A change in multifactor productivity
reflects the change in output that cannot be accounted for by
the change in combined inputs. As a result, multifactor productivity
measures reflect the joint effects of many factors, including
new technologies, economies of scale, managerial skill, and changes
in the organization of production.
The U.S. Department of Labor, Bureau of Labor Statistics
(BLS), is responsible for developing official productivity statistics
for the United States. BLS publishes four sets of productivity measures
for major sectors and subsectors of the U.S. economy:
- Quarterly and annual output per hour and unit labor costs for
the U.S. private business, private nonfarm business, and manufacturing
sectors. These are the productivity statistics most often cited
by the national media.
- Annual measures for output per hour and unit labor costs for
3-, 4-, 5-, and 6-digit North American Industry Classification
System (NAICS) industries in the United States, with complete
coverage in manufacturing and in retail trade, as well as some
coverage in other sectors.
- Multifactor productivity indexes for the private business,
private nonfarm business, and manufacturing sectors of the economy.
- Multifactor productivity indexes for 2- and 3-digit
Standard Industrial Classification (SIC) manufacturing industries,
such as the railroad transportation industry, the air transportation
industry, and the utility and natural gas industry. These include
indexes for total manufacturing and for 20 2-digit SIC manufacturing
industries on an annual basis, which compare real value-added
output measures to aggregate measures of input: labor, capital,
energy, non-energy materials, and purchased business services
[35].
In the AEO2004 reference case, productivity
growth in the nonfarm business sector is projected to average 2.25
percent annually from 2002 to 2025. The low and high macroeconomic
growth cases project average annual growth of 1.82 percent and 2.65
percent, respectively. As discussed below, the range of productivity
growth covered by the three cases is within the range of historical
experience as well as what is projected for the future by various
experts in the productivity field. Figure 8 shows 5-year average
annual growth rates for the three cases.
Estimates of Historical Productivity Growth
and Their Determinants
Productivity Growth up to 1995
For the period 1917-1927, labor productivity growth
averaged 3.8 percent per year, the highest rate for any comparable
10-year period for the U.S. economy [36]. That productivity
boom coincided with the adoption of the assembly line and the proliferation
of the automobile. Broadcast radio and the electric utility industry
saw strong development in the 1920s, and Lindbergh made his famous
transatlantic flight, which ushered in the age of aviation. Slow
productivity growth in the 1927-1948 period accompanied the Great
Depression and World War II. After the war, two factors combined
to boost productivity growth: first, output had dropped so far during
the Great Depression that simply returning to trend growth required
a period of faster economic growth; second, the economy benefited
from a wave of innovations, including the building of the interstate
highway system, the discovery of transistors, and the emergence
of commercial aviation. Between 1948 and 1973, annual labor productivity
growth averaged 2.8 percent.
Printer
Friendly Version
Component |
1948-1973 |
1973-1995 |
Difference |
Output
per hour |
2.9
|
1.4
|
-1.5
|
Contributions
from |
Capital
per hour |
0.8
|
0.7
|
-0.1
|
Information
technology |
0.1
|
0.4
|
0.3
|
Other
|
0.7
|
0.3
|
-0.4
|
Labor
quality |
0.2
|
0.2
|
0.0
|
Residual
(MFP) |
1.9
|
0.4
|
-1.5
|
R&D
|
0.2
|
0.2
|
0.0
|
|
Productivity growth began to slump again in the early
1970s. Higher oil prices undoubtedly played a role in slowing output
during the 1970s, but when oil prices returned to pre-1973 levels
during the 1980s (in real dollar terms), productivity continued
to sag. Other possible explanations include a slower rate of innovations,
slower growth of workers skills, and increased government
regulation.
Martin N. Baily has estimated the contributions to
nonfarm labor productivity (output per hour) coming from increases
in capital per hour worked and labor quality over the period 1948-1995
[37]. The unexplained residual, also termed multifactor
productivity (MFP), is defined as the difference between total productivity
growth and the contributions from these two factors. Neither capital
per hour nor labor quality explains the slowdown in labor productivity
in the 1973-1995 period, leaving the explanation or lack thereof
to the unexplained residual (Table 4). Interestingly,
although the contributions from capital per hour did not differ
by much between the pre-1973 and post-1973 periods, the contributions
from information technology capital rose in the later period, while
the contributions from other capital fell.
Information Technology and the Productivity Growth
of the Late 1990s
Printer Friendly Version
Component
|
Oliner
and
Sichel |
2001 Economic
Report of the President |
Jorgenson,
Ho, and Stiroh |
Output per hour |
1.2
|
1.4
|
0.9
|
Contributions
from |
Capital per
hour |
0.3 |
0.4
|
0.5
|
Information
technology |
0.6
|
0.6
|
0.4
|
Other
|
-0.3
|
-0.2
|
0.1
|
Labor
quality |
0.0
|
0.0
|
-0.1
|
Residual
(MFP) |
0.8
|
0.9
|
0.5
|
Computer
sector |
0.2
|
0.2
|
0.3
|
Other
|
0.3
|
0.7
|
0.2
|
|
Numerous studies have attempted to explain the increase
in labor productivity from the 1973-95 period to the post-1995 period.
The conclusions of Steven Oliner and Daniel Sichel, the 2001 Economic
Report of the President, and Dale Jorgenson, Mun Ho, and Kevin
Stiroh [38] were summarized by Baily (Table 5). Although
the three studies used slightly different data to support their
analyses, there are fundamental similarities in their conclusions.
As in Bailys analysis of the earlier time period, information
technology was the largest single identifiable factor contributing
to labor productivity growth after 1945. The boost to productivity
from information technology more than offset the drag on productivity
from other capital.
In each of the three studies, the majority of the
acceleration in labor productivity growth in the post-1995 period
was assigned to the residual (or MFP) effect: 0.8 percent to 0.9
percent of the estimated 1.2-percent and 1.4-percent increases in
labor productivity (nonfarm business sector) in the first two studies
and 0.5 percent of the estimated 0.9-percent increase in labor productivity
(business sector) in the third analysis. In the studies by Oliner
and Sichel and Jorgenson, Ho, and Stiroh, more than one-half of
the MFP effect was attributed to the computer sector. The 2001 Economic
Report of the President suggested, however, that most of the
increase came from outside the computer sector.
Meyer, Baily, and others see the bunching of productivity-enhancing
innovations working in combination with a favorable U.S. economic
environment to boost productivity. In Bailys words, rapid
advances in computing power, software and communications capabilities
formed a set of powerful complementary innovations. An increasingly
deregulated U.S. economy created a highly competitive environment
that drove out inefficiencies, displaced low-productivity firms
with high-productivity ones, and forced the adoption of new innovations
in order to survive. While the new innovations were available globally,
the highly competitive environment may explain why U.S. productivity
rates benefited more from them than did other world economies. And
finally, globalization expanded markets and increased international
competition, further raising the productivity of U.S. firms.
More recently, Stiroh has found that the recent productivity
revival is broad-based, with nearly two-thirds of the 61 industries
in his analysis showing accelerating productivity gains [39].
Furthermore, Stiroh found that productivity growth was higher in
industries that either produced information technologies or used
them intensively. Thus, Stirohs industry analysis supports
the conclusion that information technology capital was a significant
contributor to the post-1995 productivity surge.
Future Outlook for Productivity Growth
The issue of productivity growth is very important
for the future economic growth of any nation. For the United States
this issue has given rise, understandably, to a significant amount
of empirical literature that has investigated the determinants of
productivity growth in the past and the future. The AEO2004
projections for productivity growth lie within the range of historical
experience and of the future expectations published by experts,
as described below.
Printer
Friendly Version
Source
|
Point estimate
|
Range |
Oliner
and Sichel (2002) |
|
2.0
to 2.8 |
Jorgenson,
Ho, and Stiroh (2002) |
2.25
|
1.3
to 3.0 |
Congressional
Budget Office (2002) |
2.2
|
|
2001
Economic Report of the President (2002) |
2.1
|
|
Baily
(2002) |
|
2.0
to 2.5 |
Gordon
(2002) |
|
2.0
to 2.2 |
Kiley
(2001) |
|
2.6
to 3.2 |
Martin
(2001) |
2.75
|
2.5
to 3.0 |
McKinsey
(2001) |
2.0
|
1.6
to 2.5 |
Roberts
(2001, updated) |
2.6
|
|
|
Most researchers who have studied the issue and prognosticated
about the future outlook have an expectation that annual labor productivity
growth will be above 2 percent for the next decade or so. Table
6 shows estimates from recent studies of projected growth in labor
productivity. The list represents most of the well-known researchers
in the productivity field. All the point estimates of future annual
labor productivity growth shown in Table 6 are 2.0 percent or higher,
and the estimated ranges fall between a low of 1.3 percent and a
high of 3.0 percent.
The key question in developing the AEO2004
reference case forecast was whether the recent surge in productivity
growth would continue. The majority view of the productivity experts
cited here is that strong growth in labor productivity will continue
for several more years. For example, the U.C. Berkeley economist
J. Bradford DeLong writes: Will this new, higher level of
productivity growth persist? The answer appears likely to be yes.
The most standard of simple applicable growth models . . . predicts
that the social return to information technology investment would
have to suddenly and discontinuously drop to zero for the upward
jump in productivity growth to reverse itself in the near future.
More sophisticated models that focus in more detail on the determinants
of investment spending or on the sources of increased total factor
productivity appear to strengthen, not weaken, forecasts of productivity
growth over the next decade [40].
Naysayers about the productivity revival include
Steven Roach and Robert Gordon. Roach believes that much of the
post-1995 productivity revival is a statistical illusion resulting
from the lack of a satisfactory measure of productivity in the white
collar services sector. Gordon argues that the role of information
technology has been overstated, and that other factors influencing
productivity growthsuch as the international and domestic
economic environment and fiscal and monetary policiesled to
the strong trend in recent years. Regardless of his views about
the role of technology in productivity growth, Gordons expectation
is that productivity will soon return to its trend growth rate of
2.25 percent [41].
Notes and Sources
Released: January 2004
|