Productivity is a measure of output, and the most common use of productivity measures is in gauging economic performance at the national level. Statistics on productivity are collected routinely by the U.S. Bureau of Labor Statistics (BLS) and their publication every quarter usually brings coverage in the business press. BLS measures "labor productivity" based on dollar output per hour of labor; the agency also publishes a more complex measure known as "multifactor productivity" which takes other inputs into account. Productivity is also measured at the level of the enterprise in output of physical product by a worker. When the worker's pay is directly based on number of pieces produced, that type of work is known as "piece-work": pay is tied to the item ("collars sewn," for instance, rather than time spent).


In economic theory (echoed in popular opinion), labor compensation is determined by productivity. In theory a person can only earn a fixed amount by labor because the labor must be compensated by the sale of the product made, and all things being equal, competition will keep the prices competitive. This translates to an essentially stagnant economy unless, in some way, the cost of the production process can be lowered. One way to lower costs is to increase output while keeping the input the same. Thus if a worker can increase his or her production from 8 items an hour to 12 items an hour while still being paid $9 an hour, the labor costs of the items will decrease from $1.125 an item to $0.75 an item. (Economics has been called the "dreary science" because it delights in such things—but to go on '¦) The converse of such an improvement in productivity is that the price of the item could be held steady and the laborer could be paid more. In this instance the worker's pay could be increased to $13.50 an hour ($1.125 times 12, not times 8). For this reason, it is a fundamental assumption of economics that wages in a genuinely free market can only increase if productivity increases.

Productivity can only increase if 1) the worker's skills increase, 2) the worker's effort increases, 3) the quality of the material processed increases, 4) the worker's tooling is better, and 5) the work-process itself is improved by better arrangements of workers, work-flow, etc. Increases in skill require time and experience, increased effort requires incentives, and the remaining factors are produced by improvements in technology.

Wages, of course, can also increase as a consequence of social force. Thus workers can unionize and impose their will. Higher costs are then imposed on the public. Similarly, government can enforce a wage level with similar consequence, the minimum wage being an example. These situations, of course, no longer represent a genuine "free market"—which, to be sure, has never existed and never shall.

Throughout the period of modern industrial history, productivity has been rising steadily as a consequence of all of the factors enumerated above, namely education in general and the invention and deployment of technology which itself is based on knowledge and energy. Arguably modern civilization rests on the discovery of fossil fuels and their exploitation which have enabled humanity to have leisure to learn and power to burn.


Labor Productivity

Government data on productivity are calculated by measuring and/or estimating the output of different sectors of the economy in dollars and the hours worked. The output divided by the hours produces the base of a productivity measure. But because the economy has its ups and downs as well as its seasonal swings, BLS does not publish the raw numbers but, instead, produces an index number. At present the base year of this index is 1992. This means that the values measured in that year are taken as 100. Other years are expressed as deviations from 1992. In 2000, for instance, manufacturing output per hour was 138.3, meaning that it had improved 38.3 percent over 1992. Productivity data are seasonally adjusted and adjusted dollar values are used to eliminate the influence of inflation.

The two major categories used are Manufacturing and Business as a whole. The most precise are data for manufacturing because, in that sector, the U.S. Census Bureau collects hourly compensation data separately from other employment data. In both categories, productivity is up substantially over against 1992 and in recent years as well. Manufacturing productivity (output per hour) stood at 138.3 in 2000 and at 171.2 in 2005, having increased 23.8 points since 2000. In Business as a whole, the productivity index in 2000 was 120.3; it increased to 136.7 by 2005, increasing 13.6 points.

The significantly higher growth rate in manufacturing productivity reflects the fact that tooling acts as a "multiplier" of human labor. Much more machinery is used in manufacturing than in any other sector. High output per hour is also experienced in highly automated activities like utilities and where large sums of money are transferred as in wholesale trade and in the financial sectors.

Based on data derived from the 1997 Economic Census, cited in Social Trends & Indicators USA, it took 4.4 people to produce $1 million in output in manufacturing, 5.7 people to produce $1 million in retail, 9 people to produce that volume in professional, scientific, and technical services, and 15.3 people to produce $1 million in health care. To produce the same dollar figure as output, the finance and insurance sector only needed 2.7 people, utilities 1.7 persons, and wholesale trade 1.4.

Multifactor Productivity (MP)

Labor productivity, of course, is a very rough measure because it only incorporates sales or revenues on the one hand and hours worked on the other. It is thus used as a stand-in, a kind of abbreviation, for more complex and very difficult calculations that take other and often intangible factors into account. One attempt to do so is the effort to measure multifactor productivity.

The BLS, in its press release on this subject, provides the following comment: "Multifactor productivity is designed to measure the joint influences of economic growth on technological change, efficiency improvements, returns to scale, reallocation of resources, and other factors, allowing for the effects of capital and labor. Multifactor productivity, therefore, differs from labor productivity (output per hour worked) measures that are published quarterly by BLS since it includes information on capital services and other data that are not available on a quarterly basis."

The MP index separately measures labor and capital inputs and then combines them based on the relative importance of each in a given sector to create a "composite" input. It similarly measures outputs per hour and outputs per unit of capital employed and also combines these. The index is then computed from the two composites.

The MP index is available back to 1987, has a base year of 2000 (index at 100), and is available to 2004. The index has increased 7.7 percent between 2000 and 2004. Multifactor productivity thus produces a more sobering picture of productivity by reflecting the role of capital which, indirectly, reflects the importance and costs of technology.

The data streams required to calculate MP are difficult to get and the index therefore difficult to replicate. Cause-and-effect relationships can only be inferred indirectly. For these reasons, MP is used primarily in academic analyses.


Labor productivity and compensation grow in tandem but not in precise coordination. In times of economic slow-down, inventories tend to be high but over time is cut and early layoffs take place. In times of up-turn, employers are slow to hire new labor until growth is well established. The overall growth rate of compensation lags that of productivity, in part explained by the "multi-factor" influence of technology which, ultimately, accounts for productivity.

This lag has been pronounced in the early years of the 21st century. In the 1992 to 2000 period, productivity increased just 4 points more than compensation based on the indices. But in the 2000 to 2005 period, productivity increased 9.1 points over compensation. A possible explanation of this divergence may be globalization. If functions heretofore counted into hours worked are off-shored, but output continues to be counted, fewer hours will be divided into dollars. Productivity will be going up precisely because the hours are expended overseas and have become invisible.


The small business, by its very definition ("small") will lack the scale effects often needed to justify high levels of automation. Similarly, unavoidable overhead functions will have less production to absorb their costs. Technological means of increasing productivity are, of course, also available to small business—and deployed by the alert business owner. This applies to rather esoteric areas as well as the more usual. The continued expansion in information technology (IT) and related specialties, for instance, such as computer-aided design and manufacturing (CAD/CAM), is bringing IT more and more within the "affordable range" of small business, as illustrated in several places throughout this volume. Small business, however, also has unique opportunities to achieve productivity through flexibility and creativity. It is very common in small businesses to have highly skilled and cross-trained employees who do "everything." Communications and decision-making are easier and often swifter. Small businesses tend to be innovators, not least in the novel use or invention of technology. Many of these traits are indirectly captured in multifactor productivity statistics even though they escape the simple calculation of sales divided by hours worked.


Magee, Monique, ed. Social Trends & Indicators USA: Work & Leisure. Thomson Gale, 2003

Parry, Thomas, and Phil Lacy. "Promoting Productivity and Workforce Effectiveness." Financial Executive. November 2000.

U.S. Bureau of Labor Statistics. "Multifactor Productivity Trends, 2003 and 2004." Press Release. 23 March 2006.

U.S. Bureau of Labor Statistics. "Productivity and Costs." Press Release. 7 March 2006.