The growth of productivity—output per unit of input—is the fundamental determinant of the growth of a country’s material standard of living. The most commonly cited measures are output per worker and output per hour—measures of labor productivity. One cannot have sustained growth in output per person—the most general measure of a country’s material standard of living—without sustained growth in output per worker.

Increases in output per hour are the same thing as reductions in hours per unit of output. So, as labor productivity rose in the American car industry during the 1920s, it took fewer and fewer hours to assemble a Model T. The price of automobiles fell, and the real standard of living of Americans increased. This was reflected in the number of cars registered in the country, which rose from 6.7 million in 1919 to 23.1 million in 1929. As the result of productivity improvement, in other words, the number of households with access to automobile transportation more than tripled in the short span of a decade.

Recently, output per hour in the sectors of the economy producing computers and telecommunications equipment has soared. The prices of these goods have plummeted, and tens of millions of American households now have high-speed computers and cellular telephones, reflecting some of the more dramatic improvements in our standard of living in recent decades.

Productivity improvements can also take place in service sector industries, as they have recently in wholesale and retail trade and securities trading. Some of our greatest challenges and opportunities lie in the service sector. For example, if we can successfully use information technology to streamline the creation, storage, and retrieval of medical records, productivity in the health sector may rise substantially. This would mean that we could deliver more services with the resources currently deployed or the same services with fewer resources reemployed elsewhere. Either way, our standard of living would rise.

A final example: in 1790, the year of the first U.S. census, upward of 90 percent of the labor force worked in agriculture. In the year 2000, less than 1.4 percent of the labor force was so employed, still producing enough for the U.S. population to eat as well as substantial exports. Continuing improvements in labor productivity in agriculture made that possible.

If the demand for a product or service is price inelastic—that is, if a given percentage decrease in price results in a lower percentage increase in the quantity demanded—then rapid productivity improvement can result in workers having to leave the industry. The reason is that industry output, even if it has risen moderately, can now be produced with fewer workers. This eventually became true for grain farming, but not generally for computers, where the demand has been more price elastic. The relative price declines produced such a big increase in quantity demanded that industry employment has actually increased. But even in the case of grain farming, the falling food prices associated with the productivity improvement led automatically to increases in real income elsewhere. These increases eventually resulted in increased demand for other goods and services, leading to expansion of demand, employment, and output outside of agriculture.

Whether or not productivity improvement is associated with increasing or decreasing employment in the affected industries, and whether or not it is temporarily associated with rises in unemployment rates, such improvements are, in the long run, the basis for increases in our material well-being.

More Technical Points

In the United States, the Bureau of Labor Statistics calculates productivity measures for the private domestic economy and the private nonfarm economy, as well as for manufacturing, industries within manufacturing, and a few other subsectors. The private nonfarm economy accounts for about three-fourths of total GDP: it excludes agriculture, housing (which is entirely services and produced almost entirely by capital), and government. The private domestic economy includes agriculture. For subsectors of the economy, or for particular industries or firms, the measure of output is value added, not gross sales. The contribution to GDP (as well as gross domestic income) of any particular economic entity is gross receipts less purchased materials and contract services.

For example, if your bakery business buys flour and yeast, rents a shop and equipment, and pays for fuel, its contribution to GDP is not the sales price of the bread made, but the difference between gross revenues and purchased materials and services except hired labor. Your firm’s output is what you and your employees have added to the value of the materials and services purchased from other firms. You do not get credit for what the other firms did. Increasing labor productivity in your bakery means increasing value added per worker or per hour worked.

A second important measure of productivity is called either total factor productivity, a term many economists favor, or multifactor productivity (MFP), the term the Bureau of Labor Statistics uses; the terms are interchangeable. Their rate of growth is often called the residual. MFP can be most easily understood by comparing the calculation of its growth rate with the calculation of the growth rate of output per hour (labor productivity).

If we use capital letters for levels and lower-case letters for rates of growth, Y/N can stand for the level of labor productivity, where Y is real output and N is hours; yn, the growth rate of the numerator less the growth rate of the denominator, is the growth rate of labor productivity. This simply says that if output per hour is to grow, output (the numerator) has to rise faster than hours (the denominator).

Multifactor productivity, in turn, is calculated as the difference between the growth rate of real output (y) and a weighted average of the growth rates of capital services and hours, the weights corresponding to shares in national income. Thus, if capital services and hours grew at the same rate, there would be no difference between the growth rate of multifactor productivity and the growth of labor productivity.

For example, between 1929 and 1941 in the United States—in other words, during the Great Depression—neither hours nor capital services increased measurably, but real output rose 32 percent. Because the weighted average of the growth of inputs in this instance was effectively zero, all of the growth of output (and growth in output per hour) was due to growth in multifactor productivity, which can be interpreted as a crude measure of the rate of “technical change.” If output rises faster than the growth of inputs conventionally measured, then we can say that some recipes for turning inputs into output must have improved.

Total (multi) factor productivity and labor productivity are related to each other. Output per hour grows as the result of two conceptually distinct mechanisms. First, if the economy saves and invests more of its current output such that the physical capital stock rises more rapidly than the number of labor hours employed, output per hour should rise as the result of “capital deepening.” Capital deepening occurs when the ratio of physical capital to labor hours rises. The idea that this positively affects labor productivity is based on the intuitive proposition that ditch diggers move more cubic meters of earth if they are using backhoes than if they use only shovels. But output per hour can also rise through the discovery of new technologies or ways of organizing production. Such discoveries contribute to growth in our measures of multifactor productivity and enable output per hour to rise even in the absence of more capital accumulation (think about the Depression example).

To return to our example of the bakery, if your firm invests in more machines so that less hand labor per loaf is required, output (value added) per hour should go up. But multifactor productivity will not necessarily rise, because your combined input measure will rise by about the same amount as output. There is another potential source, however, of increases in output per hour. If you discover a way to rearrange your labor force and equipment so that production is more efficient, or discover a great new recipe for a loaf that is equally tasty but costs you less to bake, multifactor productivity in your firm may go up, increasing your output (value added) per hour even in the absence of any capital deepening.

The bottom line: If a country wants its standard of living to rise over the long run, its labor productivity has to go up. And for that to happen, it either has to save more or innovate.


About the Author

Alexander J. Field is the Michel and Mary Orradre Professor of Economics at Santa Clara University. He is the editor of Research in Economic History and the executive director of the Economic History Association.


Further Reading

 

Abramovitz, Moses. “Resource and Output Trends in the United States since 1870.” American Economic Review 46 (May 1956): 5–23. Important article, the first to document the rise in the value of the residual in the United States during the second quarter of the twentieth century.
Abramovitz, Moses, and Paul David. “American Macroeconomic Growth in the Era of Knowledge-Based Progress: The Long Run Perspective.” In Stanley Engerman and Robert Gallman, eds., The Cambridge Economic History of the United States. Vol. 3. Cambridge: Cambridge University Press, 2000. Pp. 1–92. Analysis, up through 1989, extending the idea of the shift from dominance of physical capital accumulation in the nineteenth century to knowledge-based growth in the twentieth.
Field, Alexander J. “The Most Technologically Progressive Decade of the Century.” American Economic Review 93 (September 2003): 1399–1413. Shows that the high value of the residual in the second quarter of the century was principally due to the high growth rate of MFP between 1929 and 1941.
Field, Alexander J. “Technical Change and U.S. Economic Growth: The Interwar Period and the 1990s.” In Paul Rhode and Gianni Toniolo, eds., Understanding the 1990s: The Economy in Historical Perspective. Cambridge: Cambridge University Press, 2005. Compares economic growth in the 1930s and the 1920s with that in the 1990s.
Gordon, Robert J. “Interpreting the ‘One Big Wave’ in U.S. Long Term Productivity Growth.” In Bart van Ark, Simon Kuipers, and Gerard Kuper, eds., Productivity, Technology, and Economic Growth. Boston: Kluwer, 2000. Pp. 19–66. Argues that high rates of MFP growth in the second and third quarters of the twentieth century may have been historically unique.
Jorgenson, Dale. “Information Technology and the U.S. Economy.” American Economic Review 91 (March 2001): 1–32. Optimistic interpretation of the effect of the IT revolution on U.S. productivity growth.
Kendrick, John. Productivity Trends in the United States. Princeton: Princeton University Press, 1961. Classic reference for anyone wishing to push analysis back before 1947. Detailed aggregate and sectoral estimates for the U.S. economy.
Lipsey, Richard J., and Kenneth Carlaw. “What Does Total Factor Productivity Measure?” International Productivity Monitor (Fall 2000): 23–28. Skeptical view of what inferences we can draw from measures of the residual.
Oliner, Steven D., and Daniel E. Sichel. “The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?” Journal of Economic Perspectives 14 (Fall 2000): 3–22. Analysis of contribution of the IT revolution to recent productivity growth by two Federal Reserve economists.
Solow, Robert J. “Technical Change and the Aggregate Production Function.” Review of Economics and Statistics 39 (August 1957): 312–320. Seminal article laying out the dynamics of the Solow growth model and providing a production function interpretation of growth accounting. Analyzes data from 1909 to 1949.

 

Web Sites

 

http://www.bls.gov. This is the Web site to visit for the latest U.S. productivity data, as well as historical data running back in some cases to 1947.
http://www.oecd.org/topicstatsportal/0,2647,en_2825_30453906_1_1_1_1_1,00.html. Provides productivity data for members of the Organisation for Economic Co-operation and Development (OECD).