Steven Klepper

Time Paths in the Diffusion of Product Innovations

This study attempts to measure and analyse the diffusion of product innovations. Diffusion is defined as the spread in the number of producers engaged in manufacturing a new product. Thus, the term refers to the net entry rate in the market for a new product. We trace the history of diffusion for 46 new products and examine the inter-relations among diffusion, other aspects of technological change, price, output, and certain attributes of the relevant markets. To explain the 46 product histories, we construct a theory of the development of industries for new products. Our theory combines elements of traditional, neoclassical models with what Nelson and Winter (I974) have termed an evolutionary theory. A novel feature is that the historical sequence, or time path, of events is viewed as a critical determinant of the ultimate structure of new product markets. Thus the time path of events determines not only the course traversed in reaching the end result but the ultimate market structure itself. The paper is organised in four sections. In Section I we present our theory. In Section II we construct a series of alternative theories of the development of industries for new products based on approaches to be found in received literature. The evidence from the 46 new product histories is examined in Section III. Finally, a brief summary of principal findings follows in Section IV.

A Reprise of Size and R&D

Numerous studies have shown that, within industries, the propensity to perform R&D and the amount of R&D conducted by performers are closely related to the size of the firm, while R&D productivity declines with firm size. These findings have been widely interpreted to indicate that there is no advantage to large firm size in conducting R&D. The authors show how a simple model based on the idea of R&D cost spreading can explain the prior findings about the R&D-firm size relationship, as well as additional features of the R&D-firm size relationship, implying an advantage to large size in R&D. Copyright 1996 by Royal Economic Society.

Firm Size and the Nature of Innovation within Industries: The Case of Process and Product R&D

The effect of firm size on the allocation of R&D effort between process and product innovation is examined. It is hypothesized that, relative to product innovations, process innovations are less saleable in disembodied form and spawn less growth. This implies that the returns to process R&D will depend more on the firms output at the time it conducts its R&D than the returns to product R&D. Incorporating this distinction in a simple model, the authors derive and test predictions about how the fraction of R&D devoted to process innovation varies with firm size within industries. Copyright 1996 by MIT Press.

The Evolution of New Industries and the Determinants of Market Structure

Several empirical regularities concerning firm growth rates and industry firm-size distributions have been developed by studying primarily mature industries. The primary purpose of this article is to bring together and extend empirical regularities on the evolution of new industries and to use these regularities to gain further insight into the forces governing industry evolution. To explain these regularities, a model is constructed which emphasizes how factors governing the early evolution of industries may shape their market structure at maturity. It stresses how chance events and exogenous factors that influence the number of potential entrants to the industry, the growth rate of incumbents, and the ease of imitation of industry leaders will influence the ultimate number and size distribution of firms in the industry.

Disagreements, Spinoffs, and the Evolution of Detroit as the Capital of the U.S. Automobile Industry

The agglomeration of the automobile industry around Detroit, Michigan is explained using a theory in which disagreements lead employees of incumbent firms to found spinoffs in the same industry. Predictions of the theory concerning entry and firm survival are tested using data on the origin, location, and years of production of every entrant into the industry from 1895 to 1966. The geographic concentration of the industry is attributed to four early successful entrants and the many successful spinoffs they spawned in the Detroit area and not to conventional agglomeration economies benefiting co-located firms, as featured in modern theories of agglomeration. Implications of the findings regarding firm strategy are discussed.

Firm Survival and the Evolution of Oligopoly

After their commercial introduction, the number of producers of autos, tires, televisions, and penicillin initially grew and then experienced a sharp decline or shakeout. Guided by an evolutionary model of entry and exit, firm survival patterns in the four products are examined to determine whether there were common forces governing their distinctive evolution. Predictions concerning the effects of pre- and post-entry experience and the timing of entry on firm survival are tested. The findings are used to reflect on why industries experience shakeouts and evolve to be oligopolies.

The Making of an Oligopoly: Firm Survival and Technological Change in the Evolution of the U.S. Tire Industry

The number of producers in the U.S. tire industry grew for 25 years and then declined sharply, and the industry evolved to be an oligopoly. The role of technological change in shaping the industrys market structure is explored. A model of industry evolution featuring technological change is used to derive predictions that are tested using a novel data set on firm entry, exit, size, location, distribution networks, and technological choices prior to the shakeout of producers. Consistent with the model, earlier-entering and larger firms survived longer, principally because of the influence of age and size on technological change.

Heritage and Agglomeration: The Akron Tyre Cluster Revisited*

We use new data on the location and background of entrants into the U.S. tire industry to analyze the factors that caused the industry to be so regionally concentrated around Akron, Ohio, a small city with no particular advantages for tire production. We analyze the states where firms entered and for the Ohio entrants the counties where they originated and entered, and we conduct various analyses of how proximity to other tire firms and to demanders affected the longevity of tire producers. We also examine how the heritage of the Ohio entrants influenced their longevity. Our findings suggest that the Akron tire cluster grew primarily through a process of organizational reproduction and heredity rather than through agglomeration economies, as has been commonly posited by scholars of the industry.

The nature, sources, and consequences of firm differences in the early history of the semiconductor industry

Four entrants into the early semiconductor industry—Sprague Electric, Motorola, Shockley Semiconductor Laboratories, and Fairchild Semiconductor—displayed remarkably different performance and behavior. Case studies of the firms demonstrate that the key differences stemmed from the firms’ technological goals and activities and their abilities to integrate R&D and manufacturing. These differences can in turn be related to the firms’ origins and their different conditions upon entry into the semiconductor industry, which had lasting effects due to constraints on change. While the cases offer limited prescriptions for management, they underscore the importance of technological diversity for an industry’s rate of technical advance and, in turn, public policies that support such diversity. Copyright

The tradeoff between firm size and diversity in the pursuit of technological progress

ConclusionOur analysis lends support to both sides of the debate concerning the optimal firm size for achieving technical advance. It provides a basis for why industries composed of many small firms will tend to exhibit greater diversity in the approaches to innovation pursued, and why greater diversity will contribute to more rapid technological change. It also provides a basis for why industries populated by larger firms will achieve a more rapid rate of technical advance on the approaches to innovation that are pursued. These arguments together suggest that a tradeoff exists between the appropriability advantage of large size and the advantages of diversity that accrue from numerous small firms. Others, suchas Nelson (1981), have also recognized a tradeoff between the diversity-inducing advantage of more competitive industry structures and advantages of large firm size, but not the particular tradeoff we have identified. Our analysis has been more appreciative than rigorous and, indeed, often explicity speculative. While we attempted to raise important questions, our framework requires more structuring before we can be confident about any of our conclusions. Even in its inchoate form, however, our analysis demonstrates that much needs to be done before the current debate about firm size can seriously inform policy. If we accept the plausibility of our basic framework, it focuses attention on a range of issues and questions. The fundamental premise of our analysis is that firm capabilities and perceptions differ within industries. This premise is not, however, widely reflected in analyses of industry behavior and performance, which typically take some representative firm as their starting point. Indeed, the analytic utility of our particular premise deserves scrutiny. Are differences in firm capabilities and perceptions as critical to explaining the industry patterns in innovative activity and performance as we suggest? Do these differences persist? Is our abstract characterization of these differences and their effects on innovative activity up to the task of providing a basis for policy?These intraindustry differences in capabilities and perceptions underpin the hypothesized relationship in our framework between the number of firms within an industry and the number of distinct technological activities pursued by the industry as a whole. Surely this hypothesis should be tested. To establish the relationship between numbers of firms and technological diversity, we also made two important assumptions, which themselves should be examined. First, we assumed that firms independently decide upon which approaches to innovation to pursue.This assumption precludes the clustering of firms around innovative activities due to imitation, a phenomenon highlighted by Nelson (1981) and Scott (1991). To the degree that innovative activities yield relatively fast, public results, the assumption may be suspect. While our evidence indirectly suggests that such clustering may not be critical for explaining innovative activity in a wide range of industries, more research would be helpful. Second, we assumed that the number of approaches to innovation pursued by firms is independent of their size, implying large and small firms will tend to pursue the same number of approaches. This assumption probably does not apply to the smallest firms within an industry, particularly to the extent that such firms are often not full line manufacturing firms. Does it apply, however, to the medium to large firms that account for the preponderance of R&D and economic activity inthe manufacturing sector? While our evidence again provides indirect support for this claim, more empirical and theoretical research is indicated.We also made other claims and assumptions that deserve further attention. For example, we argued that greater technological diversity stimulates technical advance and provides gross increments to social welfare. Assuming it exists, the mechanism linking diversity and technical advance has never been examined empirically and is not obvious. Our assumption that expected firm growth due to innovation is increamental played an important role in permitting usto hypothesize an appropriability advantage of large size. Again, both the assumption and its alleged effect on innovative activity are worth examining. Finally, we also need to test whether the relationship between R&D and firm size within industries depends upon appropriability conditions, particularly upon the extent to which firms can sell their innovations or grow rapidly due to innovation.Cohen and Klepper (1990) demonstrate that if firms can sell some fraction of their innovations in disembodied form or if growth due to innovation is unconditioned by existing output levels, then large firm size will confer less of an advantage and R&D effort should rise less than proportionally with firm size. In conclusion, this litany of reasonable but unsubstantiated assumptions and arguments should make clear that this paper is only a modest beginning of a daunting research agenda.