What explains the variation in the rate at which organizations learn? Many researchers have speculated about factors explaining organizational learning curves and con- tributing to the variation observed in organizational learning rates. For example, Hayes and Wheelwright (1984) discussed the following factors as facilitators of organizational learning: individual learning, better selecting and training of new members, improved methods, better equipment and substitution of materials and/or capital for labor, incentives, and leadership. Joskow and Rozanski (1979) identified several factors as contributors to the productivity gains associated with experience: routinization of tasks, learning by management that leads to more efficient produc- tion control, learning by engineers who redesign the equipment and improve routing and material handling, and learning by suppliers who are able to provide a speedier and more reliable flow of material. Dutton et al. (1984) noted that productivity gains derived from improvements in capital goods, labor skills, materials, engineering, and managerial expertise. Lieberman (1987) indicated that the productivity gains stemmed from a variety of underlying sources, including improvements in capital equipment, improvements in product and process designs, and improved organiza- tional and individual skills. In our interviews with managers at aerospace and truck plants regarding their perceptions of the most important determinants of organiza- tional learning curves, the managers emphasized: improvements in the performance of individual workers; improvements in the technology, tooling, and layout; improvements in the organization’s structure; and better understanding of who in the organization is good at what (Argote, 1993).
Thus, many factors have been hypothesized to contribute to the productivity gains associated with increasing experience. The factors can be grouped into three general categories: increased proficiency of individuals, including managers, engineers and direct production workers; improvements in the organization’s tech- nology; and improvements in its structure, routines and methods of coordination.
Studies assessing the contribution of various factors to organizational learning rate aim to open up the “black box” of organizational learning curves by identifying the factors that drive the productivity gains associated with experience. Organizational performance does not improve automatically with experience. The variation observed in organizational learning rates clearly indicates that productivity improve- ments are not guaranteed to occur as experience increases. A goal of much research on organizational learning rates is to identify the specific factors that lead to produc- tivity improvements.
Working in this tradition, Lieberman (1984) found that investment in Research and Development appeared to accelerate the rate of learning among firms in the chemical processing industry. Similarly, Sinclair, Klepper, and Cohen (2000) found that Research and Development appeared to drive the productivity improvements observed in a chemical firm.
Hayes and Clark (1986) investigated the effect of a variety of factors on the productivity of factories. Hayes and Clark found that reducing the work-in-process inventory, reducing the number of rejects, and decreasing the number of engineering change orders improved productivity. The researchers also found that investment in capital had a positive effect on productivity but cautioned about the importance of managing the introduction of new technology appropriately and adapting it to the organizational context. Surprisingly, investment in training showed a consistently negative relationship to productivity in the Hayes and Clark (1986) study. This latter finding, of course, does not imply that training necessarily hurts productivity but rather suggests that many training programs may be counterproductive or are used as a corrective device once productivity problems have appeared.
Galbraith (1990) examined factors explaining the productivity of “recipient” organizations after attempts to transfer advanced manufacturing technology to them had been made. Galbraith (1990) found that the time it took the recipient site to reach the level of productivity achieved by the donor before the transfer was greater when the organizations were geographically far apart and when the technology was complex. Conversely, the time to recover was faster when co-production occurred at the donor site, when an engineering team was relocated from the donor to the recipi- ent site for more than one month, and when individuals involved had a financial stake in the success of the transfer. Similar to the Hayes and Clark (1986) result, training was associated with greater productivity loss. The measure of training did not reflect the appropriateness or quality of the training program.
Adler and Clark (1991) investigated the effect of two learning process variables, engineering activity (measured as the cumulative number of hours spent by direct personnel on product development changes, running experiments, or learning new specifications) and training activity (measured as the cumulative number of hours spent in training by direct personnel), on the productivity of two manufacturing departments. The researchers found that the two learning process variables had different effects in the two departments. In one department, training facilitated productivity, whereas engineering changes impaired it; in the other department, training impaired productivity, whereas engineering changes had a direct positive but (through their effect on training) an indirect negative effect on productivity. The researchers suggested that in the department where engineering changes had a nega- tive effect on productivity, the changes were made for product performance con- cerns whereas in the department where engineering changes had a positive effect they were motivated by manufacturability concerns. Training had a more positive effect in the capital-intensive department where it was less disruptive to production than in the labor-intensive department where the training was seen by management as lacking discipline.
These studies are an important first step at understanding variation observed in organizational learning rates. The studies described in this section are important because they analyze productivity in actual organizations and document performance differences across them. A challenge with the approach, however, is that the same independent or predictor variable may be implemented very differently in different organizations. For example, two firms may make the same number of engineering changes: one chooses the changes judiciously based on analyses of how to improve manufacturability of the product; the other makes the changes on an ad hoc basis.
The changes at the former firm are more likely to improve productivity than those at the latter. Yet both firms would have the same value for the engineering change variable because the firms made the same number of changes. Thus, measuring organizational phenomena in such an aggregate form can mask important differ- ences in the phenomena. These more aggregate studies of productivity are useful in suggesting that a variable may make an important contribution to productivity. The aggregate studies can be complemented by more fine-grained studies that specify the conditions under which the variable has a positive or negative effect on produc- tivity. These more fine-grained studies are described later in this monograph.
In our work, we have found that differences in organizations’ abilities to retain and transfer knowledge are major contributors to differences observed in organiza- tional learning rates. For example, a firm that is consistently better at retaining knowledge will typically have a faster productivity growth rate than one where knowledge is lost. Empirical results on the acquisition and loss of knowledge are discussed in chapters on Organizational Forgetting (Chap. 3) and Organizational Memory (Chap. 4). Furthermore, a firm that is better at learning from other organi- zations will generally have a faster rate of productivity growth than one less adept at learning from others. Jarmin (1994) also found that firms differed in the extent to which they benefited from knowledge acquired at other firms. Empirical findings on the transfer of knowledge are presented in Chap. 6.
Levels of Analysis
Learning occurs at different levels of analysis in organizations: individual, group, organizational, and interorganizational (Kozlowski, Chao, & Jensen, 2010). For example, Reagans et al. (2005) determined in their analysis of hospital surgical teams that learning occurred at the levels of individual team members, the team and the hospital organization. In a study of firms in contractual relationships, Kellogg (2011) found evidence of relationship-specific learning between parties in the contract.
Although learning occurs at different levels of analysis in organizations, the knowledge acquired from learning must be embedded in an organizational reposi- tory in order for organizational learning to occur. Learning generally occurs through individuals in organizations. For organizational learning to occur, the individual would have to embed the knowledge in a repository such as a database, routine or transactive memory system. By embedding the knowledge in a supra-individual routine, the knowledge would persist even if the member who acquired the knowl- edge left the organization and other members could access the knowledge.
Although the focus of this book is at the organizational level of analysis, research at the group as well as the organizational level of analysis is included. Groups are often the level at which learning occurs in organizations. Many of the processes that occur in organizations such as communication, coordination, and influence occur in groups so they can serve as microcosms of organizations for research.
Research on interorganizational learning (see Ingram, 2002, for a review) is included to the extent that the findings have implications for organizational learning. Learning at the level of the population or industry is beyond the scope of this book. Reviews of this literature can be found in Miner and Anderson (1999) and Miner and Haunschild (1995). Learning from alliance experience (Zollo & Reuer, 2010), acquisition experience (Haleblian & Finkelstein, 1999; Hayward, 2002), and contracting experience (Mayer & Argyres, 2004; Vanneste & Puranam, 2010) are also interesting research areas that are beyond the scope of this book.
Source: Argote Linda (2013), Organizational Learning: Creating, Retaining and Transferring Knowledge, Springer; 2nd ed. 2013 edition.