Reflections on the Explanation of Heterogeneous Firm Capability
| Nicolai Foss |
Fritz Machlup famously argued that economists should not care about the specificities (e.g., internal organization) of individual firms, as this was unlikely to bring substantial additional insight in the market outcomes that were the real objects of interests for economists (here). Thus, for the purposes of price theory, firms within an industry could essentially be taken to be homogenous. Machlup’s view has been reflected in much of the micro-economics of the firm, not just in the standard Marshallian approach, but also in later contract theoretic and transaction cost approaches. While contract theory and transaction cost insights are surely capable of contributing to the understanding of firm heterogeneity, explaining such heterogeneity per se has never been a central explanatory task of these approaches. However, while the Machlup view was still holding sway among economists (well into the 1990s), dissenting economists and management scholars highlighted that heterogeneity among firms could be understood in terms of differential capability—an idea that helped them to explain firm boundaries (see much of the work of O&M blogger Richard Langlois), competitive heterogeneity in a population of firms (evolutionary economics), and competitive advantage (the resource-based view in strategy.
However, while management research has done much to advance the notion of intra-industry heterogeneity, it may have been less forthcoming with respect to theorizing the antecedents of such heterogeneity. Most work on such antecedents has highlighted cognitive a variables, such as managerial cognition and absorptive capacity, and variables related to skill levels and the efficiency of routines. Surprisingly, virtually no work in management research has linked differential capability to organizational design (e.g., the structures of communication, delegation, and incentives) or even to the human capital characteristics of firms’ workforces.
In most uses of the construct, “firm capability” refer to the efficiency level at which productive activities are carried out, or simply, “firm-level productivity.” Recent work in empirical economics (summarized here) documents not only massive but also persistent productivity differences between firms in several countries, even within narrowly defined industries. Industrial organization and labor market economists have examined many different antecedents of such differences, including human capital characteristics, firm-specific sunk costs, and differential access to technology spill-overs, usually being able to make use of register data, the traditionally preferred data sources of academic economists.
In contrast, excepting single-firm insider econometrics studies, most research on firm-level productivity has not considered organizational design as a factor explicitly contributing to firm-level productivity. The reason arguably has to do with the fact that very little publicly available register data contains information on design variables, coupled with economists’ traditional skepticism of survey-data. However, work by particularly Nicholas Bloom and John Van Reenen (summarized here) has, however, successfully demonstrated the explanatory success of rigorous interview-based surveys, particularly when such work can be linked to field-experiments. This line of research has documented massive productivity differences across firms and public organizations that are to a high extent explainable in terms of the efficiency of the applied management practices as well as ownership patterns.
While this line of research is highly promising, so far it has only identified a subset of the firm-specific determinants of capability/firm-level sources of productivity differences. Thus, it is mainly focused on human resource management practices, to the neglect of such traditional organizational design variables as delegation, departmentalization, specialization, and formal communication structures, many of which are theorized in the economics of organization. A further problem is that it tends to approach practices in a partial way, where changing one organizational practices directly leads to productivity improvements. Given this, it is less clear why such practices are not just changed. However, much theory in fact suggests that elements of organizational design cluster in bundles of discrete practices that are linked to each other in terms of, more or less binding, relations of complementarity. Given this, practices only have a significant performance effect when they are implemented (changed) as a discrete bundle, which is likely to be risky and costly. Moreover, the way in which such practices cluster may reflect not just complementarity, but also a high degree of firm-specificity in the sense of transaction cost economics. These characteristics contribute to the understanding of why differences in performance tend to be sustained within industries.