Publication: RAND Journal of Economics
Publication Date: 22-DEC-06
Delivery: Immediate Online Access
Author: Franco, April Mitchell ; Filson, Darren
Company: International Business Machines Corp.

Article Excerpt
In many industries, one important method of diffusion is through employee mobility: many of the entering firms are started by employees from incumbent firms using some of their former employer's technological know-how. This article explores the effect of incorporating this mechanism in a general industry framework by allowing employees to imitate their employers' know-how. The equilibrium is Pareto optimal because the employees "pay" for the possibility of learning their employers' know-how. The model's implications are consistent with data from the rigid disk drive industry. These implications concern the effects of know-how on firm formation and survival.

1. Introduction

The transmission of knowledge between firms often occurs through employee mobility. New firms are often spin-outs, firms started by a former employee of an incumbent firm. This has been observed in the automobile and construction industries, as well as among advertising agencies and law firms (Garvin, 1983; Phillips, 2003). Spin-outs are the most important type of entrant in many high-tech industries, where technological know-how is often embodied in human capital. Braun and Macdonald (1982) and Christensen (1993) document the importance of spin-outs in the semiconductor and rigid disk drive industries, two generic examples of high-tech industries.

Existing models of industry evolution, such as Gort and Klepper (1982) and Jovanovic and MacDonald (1994a, 1994b), do not specify the mechanism through which knowledge diffuses. Here, we introduce a model that specifies a mechanism: employees can attempt to imitate their employers' know-how and use it to start their own firms. Each period, each agent can either work outside the industry, work as a researcher in the industry, or run a firm. In equilibrium, agents with sufficiently high know-how run firms. Other agents can work in the industry to attempt to improve their know-how. Over time, as firms innovate and researchers imitate, the distribution of know-how improves and the critical level of know-how required to run a firm rises.

The model yields several testable hypotheses. First, more technologically advanced firms produce spin-outs. Because the critical level of know-how required to run a firm continues to increase as the industry evolves, a researcher who learns from a laggard may not acquire sufficient know-how to run a firm. Second, a firm's probability of surviving is increasing in its technological know-how. This result and the learning mechanism yield a third hypothesis: a spin-out's probability of surviving is increasing in its parent's know-how. These hypotheses are tested on the rigid disk drive industry. There is support for the model's results on spin-out generation and firm survival. We find that spin-out survival is more closely related to some forms of parent know-how than others.

We explain several other facts about the rigid disk drive industry using a simulation of the model. For example, Lerner (1997) establishes that in the 1970s and 1980s, while the industry was expanding, profits were low and rose as the industry matured, even though the price was steadily declining. During this time, technological laggards who were close behind the leaders tended to catch up to leaders, and laggards who were further behind were more likely to exit. Laggards occasionally surpassed leaders. The simulation matches these trends and explains how laggard "leapfrogging" can occur.

The model also yields an important welfare conclusion. In contrast to previous models of technological diffusion, despite the fact that technology spills over from incumbents to spin-outs, the competitive equilibrium is optimal. The employees "pay" for the possibility of imitating their employers' know-how, because imitation is locationally specific and the agents who benefit can be identified. In Jovanovic and MacDonald (1994a), imitation depends only on the distribution of know-how in the industry and not on the actions of the individuals. The lack of property rights creates an externality and a suboptimal equilibrium. The optimality result presented here is similar to that in Chari and Hopenhayn (1991). Unlike that model, here the arrival of new technology is endogenous and depends on actions undertaken by agents in the economy.

The model contributes to the literature on industry evolution and knowledge diffusion. The competitive framework we use is similar to Hopenhayn (1992) and Jovanovic and MacDonald (1994a). Higher know-how increases output, as in Lucas's span-of-control (1978). However, here agents can improve their productivity either by working as researchers and imitating or by running a firm and innovating. Researcher imitation has been analyzed by Pakes and Nitzan (1983), but they focus on monopoly and duopoly markets and on characterizing the optimal contract between the firm and the researcher. Our results are quite different. Instead of paying the researcher a higher salary in the following period to prevent him from becoming a competitor, the firm owner adjusts wages in the current period by taking into account the value of imitating her technology and allowing the employee to leave.

In recent work similar to ours, Klepper and Sleeper (2000) consider the case where employees may leave to start a firm in a differentiated-products market and test their model's implications using data from the laser industry. Their model builds on Hotelling-style models of firms locating along a line. In contrast, our model builds more directly on explicit models of industry evolution such as Jovanovic and MacDonald (1994a). Klepper and Sleeper (2000) find evidence that spinouts inherit know-how from parents, as do we. (1) They also find evidence that more successful firms spawn more spin-outs. However, they measure success using longevity. In contrast, we use a direct measure of know-how and find evidence that higher know-how firms are more likely to spawn spin-outs. Other firm characteristics such as size and sales growth have no explanatory power in determining whether the firm generates a spin-out in a given year once we control for the firm's know-how. In our data, while several spin-out parents are large established firms, others are small high-tech startups themselves.

The article proceeds as follows: Section 2 describes the evolution of the rigid disk drive industry, Section 3 presents the model and theoretical results, Section 4 provides a simulation, Section 5 provides the empirical results, and Section 6 concludes. All formal proofs are in an Appendix on the authors' websites www.cgu.edu/pages/428.asp.

2. A brief history of the rigid disk drive industry

* We refer the interested reader to Christensen (1993,1997), Lerner (1997), and the Disk/Trend Report, an annual market research publication that focuses on the rigid disk drive industry, for more complete descriptions of the industry's history. The industry began in 1956 when IBM introduced the first rigid disk drive. Followers began entering in the 1960s and were of two main types. Captive producers, such as Burroughs, Control Data, and Univac, were vertically integrated computer manufacturers that produced drives for in-house use. Plug-compatible market (PCM) firms were independent drive producers that made drives that were plug-compatible with IBM's computers. PCM firms sold drives directly to users of IBM computers. Christensen (1993) reports that many of the early PCM firms were IBM spin-outs, including Century Data, Memorex, Pertec, and Storage Technology Corporation. When the minicomputer market began growing rapidly in the mid- 1970s, an original equipment market (OEM) emerged. OEM firms served as either primary or secondary sources of drives for computer manufacturers.

Innovation and imitation in the disk drive industry occurred at an extremely rapid rate from 1956 to 1997 and took several forms. First, several advances in technical features improved capacities and access times. Second, several improvements in design and manufacturing techniques improved costs and reliability. Third, several architectural innovations occurred: drives with smaller diameters were introduced beginning with 8" and 5.25p" drives in the late 1970s and continuing with 3.5", 2.5", and 1.8" drives. When first introduced, the new drives served new buyers: 8", 5.25", 3.5", 2.5", and 1.8" drives were first used in minicomputers, personal computers, portable computers, notebook computers, and smaller portable devices, respectively. In response to the profit opportunities generated from rapid technological change and market growth, net entry occurred. The number of firms continued to rise until the mid-1980s and then leveled off a short time before falling in the early 1990s (Lerner, 1997). The patterns for net entry and the number of firms are similar to those established in industries with new products by Gort and Klepper (1982).

* Spin-outs: importance and imitation. Focusing on U.S. disk drive firms in the period 1976-1989, Christensen (1993) shows that while spin-outs were not the only source of entry, they were definitely the most important source. Only 3 out of 28 non-spin-out entrants survived until 1989, but 16 out of 40 spin-outs survived. Spin-outs accounted for all but four of the startups that were successful at generating revenue and accounted for 99.4% of the total cumulative revenues generated by the startup group. By 1989, seven of the world OEM/PCM market's ten largest firms were spin-outs. Table 1 provides our updated list of spin-outs. Our data show that after 1989, only five spin-outs and two non-spin-outs entered. This implies that Christensen's detailed analysis describes the vast majority of the entrants.

Spin-outs were an important source of entry, but were they imitators? Did spin-outs imitate firm-specific know-how of their former employer, as in our model, or did they simply learn industry-specific know-how? (2) The Disk/Trend Report describes several examples of firm-specific technical know-how being imitated. For example, founders of Amcodyne and Areal Technology learned how to make high areal density drives from their parent firms; founders of Dastek used thin film head technology after learning from IBM; and the founders of Tecstor modelled their drives after those of their parent firms.

Know-how associated with entering new diameters early was also imitated. Table 2 lists early movers by diameter. Almost all of the firms listed are either spin-outs, parents, or both, with the exception of BASF, New World Computer, and Control Data. (3) Many of the firms are related to each other. Table 3 lists the spin-outs from early-mover parents along with whether the spin-out was an early mover and, if so, in which diameter. From Table 3, the probability that a randomly selected spin-out from an early-mover parent is an early mover is 5/15 = .33. Of the 177 firms that were not spin-outs from early-mover parents, only 12 were early movers, resulting in a probability of .068. The two probabilities differ substantially, and the difference is significant...



NOTE: All illustrations and photos have been removed from this article.



More articles from RAND Journal of Economics
Durable-goods oligopoly with secondary markets: the case of automobiles., 22-JUN-07
Nonlinear pricing in an oligopoly market: the case of specialty coffee, 22-JUN-07
Measuring consumer welfare in the CPU market: an application of the pure-characteristics demand model., 22-JUN-07

Looking for additional articles?
Click here to search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Click here to search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.