VERDOORN’S LAW (Social Science)

Productivity growth is the key to economic development. The continuous increase in labor productivity, defined either as the output per worker or the output per hour worked, is what allows human societies to experience a rise in their per capita income even in face of a growing population. Productivity growth is usually associated with technological innovations, but since the beginning of the Industrial Revolution in the eighteenth century, it became clear to social scientists that economic growth itself may foster productivity growth through the division of labor, as pointed out by Adam Smith in 1776. However, despite the direct evidence that productivity rises with the level of output in manufacturing, the concept of increasing returns was ignored through most of the nineteenth century in the mainstream economic literature because the hypotheses of decreasing marginal returns to capital and labor and constant returns to the scale of production became preponderant in neoclassical economic theory.

The importance of increasing returns for economic growth was revitalized only in the early twentieth century by Allyn A. Young (1928), who emphasized not only the reduction in the average cost of production brought by output growth in manufacturing but also the product diversification that characterizes an increase in the division of labor. Verdoorn’s law, an attempt to quantify this relationship, is named after the Dutch economist P. J. Verdoorn, who published a paper in 1949 in which he measured the impact of economic growth on labor-productivity growth in manufacturing for a group of countries in the late nineteenth century and early twentieth century.


In general terms, Verdoorn’s law implies the existence of a stable and positive causal relationship from the growth rate of output to the growth rate of productivity in manufacturing in the long run. More formally, let p and q represent the growth rates of labor productivity and output in manufacturing, measured in logarithmic terms. Verdoorn’s law was originally estimated as p = a + bq, where b is a positive parameter that measures the elasticity of labor productivity to output. The estimate of b, known as the "Verdoorn coefficient," in most empirical studies takes a value around 0.5. The intuitive meaning of this result is that an additional one percentage point increase in the growth rate of output leads to a half percentage point increase in productivity in manufacturing. Since p = q — n, where n is the growth rate of employment in manufacturing, Verdoorn’s law can also be estimated as n = —a + (1 — b) q.

The theoretical foundation of Verdoorn’s law is the existence of economies of scale in manufacturing, that is, the fact that the average cost of production falls with an increase in the amount of goods produced. The sources of economies of scale within a firm or industry are usually divided into two categories: static or dynamic. Static economies of scale come from the fact that most processes of production incur a fixed cost, that is, a cost that has to be paid no matter whether anything is produced. As a result, the higher the level of production, the lower the average fixed cost per unit produced and consequently the higher the economy of scale. It should be noted that static economies of scale are reversible because, if production is reduced, the average fixed cost rises. Dynamic economies of scale come from the productivity gains associated with innovations brought about by the increase in production. The intuition here is that the dynamic economies arise from learning by doing and as such are irreversible. Even if the level of production falls, the new knowledge acquired from experience does not vanish.

Verdoorn’s (1949) study of productivity growth was published in Italian and went unnoticed by the majority of the economic profession until Nicholas Kaldor (1966) drew attention to it. As summarized by Anthony P. Thirlwall (1983), Kaldor proposed that three growth laws characterized economic development: (1) the higher the growth rate of output in manufacturing, the higher the growth rate of gross domestic product (GDP); (2) the higher the growth rate of output in manufacturing, the higher the growth rate of labor productivity in manufacturing, as proposed by Verdoorn; and (3) the higher the growth rate of output in manufacturing, the higher the growth rate of labor productivity outside manufacturing.

In its broadest sense, Verdoorn’s law implies the possibility of endogenous or induced technical change, which forms the basis of theories of economic growth based on increasing returns. In neoclassical theories economic growth is usually explained from the supply side, and Verdoorn’s law tends to appear as a learning function that links the growth rate of labor or multifactor productivity to the growth rate of the stock of human and physical capital. In nonmainstream theories of Keynesian inspiration, economic growth is usually determined from the demand side, and therefore Verdoorn’s law tends to appear as the explanation of how demand problems can result in uneven development, that is, in permanent growth divergences across countries or regions.

The logic of demand-driven growth divergences is clear and intuitive. If productivity growth is a positive function of economic growth, then an initial increase in aggregate demand can set off a cumulative process along the lines proposed by Gunnar Myrdal (1957), in which productivity gains increase profits and wages, which in its turn leads to another round of demand expansion and so on. The main economic implication of Verdoorn’s law is therefore that growth may be self-reinforcing, especially if we bring international trade into the analysis. The basic idea here is that fast-growing economies may be able to maintain their international competitiveness because of the fast productivity growth induced by income growth itself. By analogy, slow-growing economies may not be able to break out from their situation because of the slow productivity growth induced by their very own poor growth performance.

Since its revitalization by Kaldor, Verdoorn’s law has been the object of many studies, and the empirical debate tends to revolve around four main issues. First, if productivity growth is exogenous across regions, Verdoorn’s law may be spurious because it may be productivity that drives output instead of the other way around. The solution for this possible reverse causation is to investigate whether Verdoorn’s law holds across regions that share the same technology. Second, it may also be the case that productivity growth depends on the level of capital per worker, so that Verdoorn’s law finds an influence of output growth on productivity growth because the former functions as a proxy of capital growth. The solution for this problem is to include the capital per worker as a separate explaining variable when estimating Verdoorn’s law. Third, Verdoorn’s law may actually be a misspecified labor demand function, that is, a demand function that ignores the impact of the real wage in the determination of employment. The obvious solution for this problem is to include the real wage as a separate explaining variable when estimating Verdoorn’s law. Finally, because firms do not adjust employment as fast as output in the face of demand fluctuations, productivity tends to increase during upswings and fall during downswings because of labor hoarding. If that is the case, Verdoorn’s law may actually be just a short-run phenomenon resulting from business fluctuations, known as Okun’s law in economics. The natural solution to this problem is to control for changes in the business cycle when estimating Verdoorn’s law.

As usually happens in economics, after decades of empirical studies there is evidence both in favor of and against Verdoorn’s law. From the many studies on North America and Europe surveyed by John McCombie, Maurizio Pugno, and Bruno Soro (2002), the balance seems to confirm Verdoorn’s law. When we consider developing economies, Vaishali Mamgain (1999) finds evidence in favor of Verdoorn’s law only in one of six newly industrializing countries of East Asia, whereas Thirlwall and Heather Wells (2003) find evidence in support of Verdoorn’s law for a sample of forty-five African countries. In Latin America, E. Luis Lemos Marinho, Claudio Andre Gondim Nogueira, and Antonio Lisboa Teles da Rosa (2002) find mixed evidence for Brazil. It will probably take some time until the increasing number of applied studies on developing economies tend one way or another.

Next post:

Previous post: