|
Article Excerpt
1. Introduction
The literature on industry evolution finds that the number of firms in an industry increases to a peak and then declines to a roughly steady-state number. This pattern has been attributed to competition and legitimation (Hannan and Freeman 1989), different cost structures of entrants and incumbents (Jovanovic and MacDonald 1994), and the technological prowess of incumbents relative to entrants (Klepper 2002). Recent research that has analyzed this pattern with more detailed data has noted a "sales takeoff" accompanying the increase in the number of firms in the industry. This has generally been attributed to the entry of firms with a lower (marginal) cost position. These firms shift the supply curve down and consequently cause the quantity of goods sold to increase (Bass 1980, Stoneman and Ireland 1983, Gort and Klepper 1982, Goldner and Tellis 1997). More recent work has argued that the sales takeoff is due not only to a shift in the supply curve, but also to an outward shift in the demand curve (Agarwal and Bayus 2002).
While these outcomes have been well documented (Gort and Klepper 1982, Agarwal and Gort 2002), the analysis of the causal mechanisms remains somewhat incomplete. Because data are usually aggregated at a high level in cross-industry studies focusing on industry-level dynamics, the current literature has not been able to identify one of the key causal mechanisms for a sales takeoff: entry or expansion by a dominant firm. Similarly, this industry-level focus, which assumes homogeneity in industry demand, has precluded study of the evolution of segments within an industry. (1)
This paper extends the industry evolution literature by developing a theory of the evolution of segmented industries, an area in which there is very little extant theory. A dominant firm expands within its existing segment until growth in that segment slackens. This causes the firm to selectively enter new segments in which it can exploit its technological capabilities but avoid cannibalization of its current products. The dominant firm acts as a Stackelberg leader, triggering a price decline, sales takeoff, and change in the pattern of entry and exit by fringe firms in that segment. This dynamic cycle repeats itself when growth in the new segment falls. To test the predictions of the theory, we adopt a novel empirical approach to explore dominant firm entry and its consequences for fringe firms. We study a single industry--the desktop laser printer industry--that is composed of several identifiable market segments. By studying the entry of the same dominant firm into different segments at different points in time, we are able to conduct multiple "experiments" while holding basic features of the industry and the players constant.
Our paper integrates economic and management theory on industry evolution to make three contributions to the literature. First, rather than examine dynamics at the aggregate industry level, our work examines how an industry evolves in different segments, and to what extent the theoretical predictions and empirical regularities found in the aforementioned literature inform the way in which segments develop in an industry. The evolution of multiple segments allows relationships between related segments to be considered through panel-data methods, and also allows us to extend theory to a more fine-grained level than this literature has previously considered.
Second, using segment-level microdata that have not been employed before in studies of this type, we shine new light on stylized facts from the industry evolution and dominant firm literatures. We show that the simultaneous or delayed correlation between entry and sales found in other studies is not due solely to the entry of fringe firms moving the supply curve or shifting demand. Rather, the entry of the dominant firm alters supply and demand leading to price declines and the sales takeoff. The strategic selection of segments by the dominant firm can have a significant effect on whether a sales takeoff occurs. With respect to the literature on the dominant firm, theoretical models typically begin with the formation of such a firm, and explore theoretically or empirically how such a firm defends its position by applying dynamic limit pricing to manage the tradeoff between high current profits and increased future competition (Kamien and Schwartz 1971, Caves et al. 1984). In this sense, the literature is static and usually considers the dominant firm solely in a defensive posture. Integrating these literatures with a comprehensive data set on the laser printer industry allows us to provide a much more complete picture of the sales takeoff, dominant firm behavior, and industry evolution.
Third, this paper explores the behavior and heterogeneity of fringe firms. We show that once the dominant firm enters, a swarm of fringe firms follows the dominant firm in, most likely in expectation of the sales takeoff. At the same time, fringe firms that precede the dominant firm into the segment tend to exit after dominant firm entry. This suggests that some fringe firms develop strategies, structures, and capabilities suitable for success in the presence of a dominant firm--perhaps entailing a low-cost structure and an ability to take advantage of positive externalities created by the dominant firm--while other fringe firms configure their resources to succeed in the selection environment that exists before entry of a dominant firm. It is the entry and exit decisions by small firms before, during, and after dominant firm entry that causes a large amount of "churn" in the segment, with both entry and exit rates increasing with arrival of the dominant firm. Many of the exiting firms move to new segments, which the dominant firm may subsequently enter, thus leading to a churning of the same firms once again.
We conduct our empirical examination of the theoretical predictions using data on the desktop laser printer industry from 1984 to 1996. Hewlett-Packard (HP) pioneered this industry in 1984, maintained at least 45% market share through 1996, and was widely perceived as the dominant player (de Figueiredo and Kyle 2005). We find empirical support for most of our theoretical predictions. In particular, we show that HP enters new segments when sales in its current segments stagnate or decline. The new segments that HP enters are related to HP's existing segments, in a manner suggesting that HP can exploit its existing technological capabilities in them (consistent with Klepper 2002). However, HP tends to avoid segments that are adjacent to its current segments in favor of segments further away, as long as such segments exist (consistent with Katz 1984). Furthermore, HP's entry is positively associated with a substantial price decline and a dramatic sales increase in a segment. Our evidence indicates that the dominant firm's entry triggers these changes in the segment. Finally, HP's entry precipitates a churning of fringe firms--increased entry rates following HP's entry, and increased exit rates for firms that entered the segment before HP entered. Although this study does not explore in detail the specific characteristics of fringe firms that encourage entry versus exit upon the entry of a dominant firm, we hope that it will encourage further work in this area. More broadly, we hope that this study will open up new avenues of research on the strategic dynamics in segmented industries and on dominant firm-fringe firm interactions.
2. Theoretical Development and Framework
2.1. Preliminaries: Definition of the Dominant Firm and Summary of Theory
In theoretical treatments of dominant firm behavior, dominance is defined by two characteristics: possession of a cost advantage and ability to price as a Stackelberg leader (Gaskins 1971, Caves et al. 1984). These characteristics imply that a dominant firm should be characterized by substantial market share. Empirical studies of dominant firms have identified dominance primarily by a firm's market share, with a share of 40% or 50% used as the typical threshold for dominance (Yamawaki 1985, White 1981). We follow this convention, assuming in our theory that a dominant firm behaves as a low-cost Stackelberg leader and empirically identifying HP as the sole dominant firm in the laser printer industry based on a market share threshold of 40%.
[FIGURE 1 OMITTED]
With this definition, Figure 1 outlines the basic dynamic theoretical framework developed to explain the timing and direction of dominant firm entry and the response of fringe firms in a multisegment market, and in particular, in the desktop laser printer industry. The next four subsections explore this framework in detail, but the basic outline is as follows: The dominant firm focuses on competing within its existing segment until it encounters reduced (expected) profitability in that segment, due to declining price or declining sales growth or both. This triggers the firm's entry into a new segment where it can better utilize its resources and investment. The selection of the new segment balances the tension between two basic forces: the firm's desire to exploit its innovative capabilities and its desire to avoid cannibalization of its current product portfolio. This causes a dominant firm to seek segments in which it can exploit its existing technological capabilities, but to skip neighboring segments to avoid cannibalization. When the dominant firm enters a new segment, it acts as a low-cost Stackelberg price leader, setting prices that are below those of the fringe firms already in the segment. This sparks a sales takeoff.
Fringe firms, in turn, respond to the entry of the dominant firm. A set of fringe firms whose resources and capabilities are well suited for competing in the shadow of the dominant firm enter the segment on the heels of the dominant firm. However, incumbent fringe firms tend to exit the segment, presumably because their resources and capabilities, although appropriate for the pre-dominant firm period, are not well suited for the selection environment in the post-dominant firm era. This entry and exit behavior by these different types of firms results in a churn in the segment. Many of the exiting fringe firms continue to compete in other segments of the industry, often entering new segments as they exit the existing one. Ultimately, this segment experiences sales growth decline, and the cycle begins again.
2.2. Moving the Giant: What Motivates the Firms to Move?
There are many motivating factors that may cause a dominant firm to seek new segments. We focus on three: (1) the profitability of the current segment, (2) the profitability of the potential segment, and (3) the costs the firm faces in moving from one segment to another (costs of growth). Assume that the dominant firm exists in a market segment, Segment 1. If the firm can expand costlessly, then the profitability of Segment 1 should have no effect on the firm's decision to enter Segment 2; rather, the firm moves into Segment 2 if the new segment is sufficiently attractive. For a firm that faces adjustment costs for expansion, notably an increasing marginal cost of growth in a given segment as well as a fixed cost of expanding into new segments (e.g., Klepper 1996, 2002), the calculus is different. In this case, expanding into Segment 2 occurs only if it is more attractive than growing in Segment 1; thus, the profitability of both segments drives the entry decision. The firm will enter Segment 2 only if the discounted expected profits from Segment 2 minus the expansion and growth costs are greater than the discounted expected profits from Segment 1 minus the growth costs. (2) Assuming that the dominant firm's costs do not change over time within a segment, it is straightforward to see that as quantity or price drops in Segment 1, ceteris paribus, the hurdle for entering Segment 2 declines. That is, demand slowdown in one segment creates opportunities to grow in new segments (e.g., freeing factory capacity that can be used to grow into Segment 2).
HYPOTHESIS 1 (H1). When there are growth costs, a market price or sales drop in a dominant firm's existing segment(s), ceteris paribus, makes it more likely to enter a new segment.
2.3. The Dominant Firm's Entry Decision: Which Segment?
How does a dominant firm choose which new segment to enter? Theoretical and empirical evidence on diversification indicates that a firm will enter new industries in which it can exploit its existing technological capabilities (Silverman 1999, Helfat...
|
