Innovating-by-Doing: Skill Innovation as a Source of Technological Advance

Article excerpt

Technological advance often involves a mix of discrete innovations in products, machines, tools, organization, and skills. An extensive literature has investigated innovation in products, machines, tools, and organization [e.g., Schmookler 1966; Chandler 1977]. A literature on innovation in skill also exists, but it is much smaller. Particularly notable in this literature on skill innovation is the notion of "learning-by-doing" by production workers: increases in skill that follow from direct experience with producing a particular good [e.g., Alchian 1963].

However, the literature on learning-by-doing invariably presents skill innovation by workers as minor and passive and as occurring after the introduction of the (really important) innovations in products, machines, tools, and organization. While early writings on learning-by-doing emphasized the development of increased skill by production workers, more recent writings have placed greater emphasis on improved products, machines, tools, and organization introduced by managers as a consequence of their learning-by-doing [Dutton, Thomas, and Butler 1984; Adler and Clark 1991]. This change in emphasis has further depreciated the role of increased worker skill in technical advance.

This paper claims that, contrary to the existing literature, skill innovation is not always minor and passive and does not always occur only after innovations in products, machines, tools, and organization. Indeed, I argue that some product and process innovations can be directly attributed to innovations in worker skills. I also argue that, in other cases, particular technical innovations could not have been achieved without some particular previous skill innovation.(1) As these new skills are often acquired while workers are actively engaged in production, I label this process "innovating-by-doing" by workers.

To show that skill innovation is not limited to particular industries or periods, I consider case studies from a variety of industries and periods. Further, to ease the drawing of a contrast between my claims and the existing literature, I have selected incidents of technical advance that have direct links with the already-existing literature on technological change.

The Rise of the American System

The United States rose to world economic power in the twentieth century on the foundation of mass production. In turn, the technological basis for mass production was laid almost a century earlier by the development of the so-called "American system of manufacture" in U.S. government armories. Developments in government armories during the early 1800s led to the first large-scale factory production of complex mechanical devices using interchangeable parts. Not only was the particular set of innovations that occurred within these armories "important in itself, but it has been hailed as marking the point at which America ceased to be a net borrower of technology from other nations and became a key initiator of technological change. . . . [T]he American system of manufactures represented a radically new direction for technological progress" [Pacey 1990, 146].

Because of its importance, I use the American system as a test of my claim that skill innovation had an active role in technological development. This is a particularly good test case for this claim because the American system is often presented as the first success in the widespread elimination of worker skill from the production process through the systematic use of specialized machines [e.g., Hounshell 1984].

Before the American system came to U.S. government armories in the early 1800s, gun production involved skilled craftsmen. Sometimes these craftsmen fashioned each gun individually. However, in government armories a group of skilled craftsmen worked collectively (via the division of labor) in the factory production of guns. For their day, the guns produced in these armories were highly complex mechanical instruments. …