TY - JOUR
AU - White,, Rod
AB - Abstract What determines the emergence of a winner between competing technologies? We examine competition between flat panel display technologies, with the purpose of understanding of how liquid crystal display was able to surpass plasma display panel technology despite the initial lack of a convincing technological or cost advantage, and in the absence of network externalities. We propose an explanation whereby the relative availability of pathways of suitable adjacent applications markets provides differential opportunities for technologies to increase their scope and scale of application incrementally, effecting the speed of development as well as the cost effectiveness of the end products. Our findings suggest that these sets of adjacent application markets available can strongly influence which technologies emerge as winners and which are eventually abandoned. 1. Introduction Researchers have often observed that the degree to which technologies are adopted and spread is not always consistent with their apparent relative advantages vis-à-vis alternative technologies. There are stages in development and diffusion of technologies where it is difficult if not impossible to determine which technologies will be adopted and which will not (Anderson and Tushman, 1990; Jenkins and Floyd, 2001), regardless of the relevant benefits of the competing technologies. However, research examining this phenomenon has uncovered several mechanisms. Positive network externalities, where each additional unit of the technology adds to the value of the existing units, has been given as one reason why seemingly inferior technologies become dominant (e.g., Katz and Shapiro, 1985, 1986, 1992, 1994). The installed base of existing units makes investment into compatible units relatively more attractive than alternative, noncompatible technologies. Thus, when one technology begins to exhibit such externalities it can rapidly emerge as a winner and reduce the momentum of its competitors. The competition between VHS and Beta videotape systems is a well-known example (Cusumano et al., 1992). While network externalities occur in many industries, this explanation leaves the question of why certain technologies win in cases where positive network externalities do not exist. Meanwhile, small historical events have been found to have large impacts on the selection and eventual dominance of one technology over another under a variety of other circumstances (Arthur, 1989, 1996; Rycroft and Kash, 2002). Technologies enjoying higher returns gather more investment, encouraging faster development; however, when two alternative technologies have similar levels of returns, accidents of history can result in one starting to realize positive returns before the other, increasing the likelihood it will later dominate. Geroski (2000) discusses the idea that technologies need to be legitimized to disperse, and where competing technologies exist, they may achieve differing levels of legitimization, however, the outcome of the process is uncertain. Geroski and Arthur’s points are interesting and valuable in explaining competition between technologies, but they have limited ability to predict future outcomes. Accidents of history are just that, and returns cannot be accurately predicted in advance because they rely not only on the nature of the technology, which is changing, but also the nature of markets, which in many cases do not yet exist despite latent needs. Much of the innovation literature looks at different solutions for a specific technological problem or product, later resulting in the product taking on a dominant design (Abernathy and Utterback, 1978). However, a technology may be adopted for a number of distinct market applications, some of which were not considered by its inventors and developers. This matters because how a technology performs in one application can have spillover effects into adjacent application markets. Application markets have been described in evolutionary terms as niches where speciation of preexisting technology may take place (Levinthal, 1998; Adner and Levinthal, 2002; Cattani, 2006). Each application niche has different selection criteria, in other words, technological and market requirements, while exhibiting different market sizes and levels of munificence. Speciation in one niche may facilitate entry into other proximate, or adjacent niches. This process has implications for competition between technologies, which we build upon. In this article, we introduce an additional explanation for why seemingly inferior technologies can become dominant: the availability of adjacent application market opportunities. Zook and Allen (2003) proposed the concept of “adjacency” as a marketing concept that firms could use to grow into markets which were adjacent to their existing businesses. Zook and Allen’s concept considered adjacency in terms of geography, customer groups, and steps in the value chain. We put forward the concept of application market adjacency as a technology level concept, which considers development of technology through entry into a succession of adjacent application markets. These application markets are adjacent in the sense that the technological requirements to enter a new adjacent market build on the requirements for previous markets, and the additional technological challenges required for entry into this new adjacent market are relatively modest. We introduce the mechanisms by which the opportunities in adjacent application market spaces promote investment in, and development of technologies, through widening the scope of application and through increases in the scale of usage of the technology. We further consider how the existence, or lack of adjacent application market opportunities differentially affect the development of alternative technologies. We develop our explanation based upon examination of competition between liquid crystal display (LCD) and plasma display panel (PDP) technologies in the flat panel display (FPD) industry. After many years of development, LCD eventually dominated, and firms exited or stopped investing in PDP, despite technological drawbacks associated with LCD and several decades of debates over the relative merits of the two technologies. As we discuss later, network externalities did not play a role in this case. We suggest that the set of market opportunities faced by each technology played a large role in determining which technology became dominant. The remainder of this article is organized as follows. We first introduce the methods used in this article and then present historical evidence around the development and application of the two FPD technologies, ending with the competition to become the dominant display for use in the TV set industry. We next examine the patterns laid out in the historical examination to develop and explain our theory. Lastly, we propose implications for theory and managerial practice. 2. Research methods In this study, we develop grounded theory (Strauss and Corbin, 1998) through historical case study analysis (Lawrence, 1984; Eisenhardt, 2002; Yin, 2014) of FPD technologies, specifically focusing on the question: how did LCD become the dominant flat panel TV display? By considering this question, we seek to develop theory about technological competition in the absence of market externalities. Our approach first captured potentially relevant data, and then proceeded through an iterative cycle of organizing and interpreting the data for patterns, considering existing literature, and then returning to the data (Eisenhardt, 1989, 2002). Over several cycles, this allowed us to build an understanding based upon the historical record and positioned with regard to existing literature. FPD technology offers an advantageous context in which to study technological competition. First, network effects do not play a role here. The availability of software to be viewed does not depend on display technology type. For example, TV broadcasting is agnostic to the actual display technology used. Aside from PDP, LCD, and cathode ray tube (CRT), other displays, such as field emission display, organic light emitting diode, and plasma addressed liquid crystal have been used in TV applications. Standards for other applications, such as computers, are based upon screen resolution, and do not depend on the display technology selected. Displays are produced in accordance with these standards, meaning the end user could watch the same content regardless of display technology used, be it LCD, PDP, or some other technology. Second, many observers did not see LCD as an obvious winner for the largest application market for displays: TV sets. In 2002, for example, Panasonic suggested that PDPs were the obvious large screen display for living room TV sets, and that LCDs would be restricted to smaller applications (Nikkei Business Publications, 2002). Several years later, Sony’s CEO Nobuyuki Idei stated that he could not predict which display technology would become dominant because of the advantages and disadvantages of each (Jojima, 2006: 170). Finally, study of competition between flat panel technologies is facilitated by a large body of accessible archival data. Most of the data sources used were in Japanese; Japanese sources had the advantage of being nearby major players in FPD industry. Japanese firms were the first to commercialize both LCD and PDP on a meaningful scale. One of the coauthors is fluent in Japanese and played a key data collection role. 2.1 Data sources We gathered extensive archival data gathered from three major sources. Data were gathered from Fuji-Chimera’s annual industry reports for the years 1998–2007, from Nikkei FPD for the years 1990–2008, and from Sangyo Times for 1990 and the years 1992–2008. The existence of these multiple sources allowed opportunity for triangulation, but was also advantageous because each source had somewhat different coverage of the industry. Fuji-Chimera, for example, included data on applications of the different technologies, whereas Nikkei FPD included interview information, and Sangyo Times presented data on production facilities of the firms in the industry. We verified suitability of these data sources with Japanese researchers familiar with this industry. We augmented this archival data with interviews of industry participants, and attendance of industry trade shows including Society of Information Display’s Display Week and the FPD International show that takes place every year in Yokohama, Japan. Data from these sources were used to develop a set of databases (Yin, 2014) on events, quotes of importance, development and investment in production facilities, production volumes and other data, in searchable, traceable formats. 3. History of development of LCD and PDP 3.1 Mechanisms of the two technologies and their implications Liquid crystals are a substance that is a liquid but becomes a crystal when electrical current is run through it. The function of liquid crystals in displays is to act as a shutter that blocks light. In a simple watch or calculator display, for example, a segment representing part of a number to be displayed is turned on to make it become dark. It can be read with light reflected off of the back of the display, or with a backlight. Liquid crystals can also be made into a matrix of dots (sometimes called pixels, depending on the application) instead of segments. Color LCDs are made up a matrix of pixels, which in turn are made out of different colored subpixels, facilitated by color filters. Light from a backlighting unit is either blocked or allowed to flow through each subpixel. Polarizers, films, and optical filters are used to improve the image quality (Castellano, 1992; den Boer, 2005). A number of different liquid crystal technologies have been developed over the years. Some of these were not commercially successful (e.g., Ferroelectric LCD), and others were limited to narrow, specific applications (e.g., high-temperature polysilicon in viewfinders and projectors). Further, some of the earlier types, such as segment displays still have uses today. However, for the purposes of this article, LCD can be considered as single technology. Modern LCD is primarily amorphous silicon (a-Si), thin film transistor (TFT) LCD technology, which is referred to as a-Si LCD, TFT-LCD, or often Active Matrix LCD (AM-LCD), although technically speaking AM-LCD also includes high- and low-temperature polysilicon varieties as well. This basic design has several implications about qualities of LCD. First, the crystal turning on and off is a physical process, which takes time (this is measured as “response time”). When the crystal does not change state as fast as moving video, this causes a blurred image. Second, the design achieves displays by blocking light from the back from passing through; however, the ability to block light is imperfect, resulting in limited contrast and difficulty displaying deep blacks due to leaking from backlights. Furthermore, the alignment of pixel the area shut on and off to a surrounding aperture can lead to limited viewing angles. The electrical connections to each pixel and subpixel are extremely small, and have to be deposited on a glass substrate, which requires a clean room environment and is sensitive to contamination (rendering pixels that are permanently in on or off states). This has been one limiting factor in making larger LCDs in the past, and is also the primary reason why LCD facilities require such large investments. The fundamental design of LCDs has some strengths as well as these drawbacks. The strengths are that it is suited to making small displays and can achieve high pixel densities. LCD can also be very energy efficient (energy consumption of the backlight being the main electrical draw). The complexity of LCD technology also means that many different knowledge sets are required to effectively develop it, including, for example, organic chemistry, optics, physics, electrical and electronic engineering (den Boer, 2005). Development, therefore, required cooperation between experts in these different fields, and across firms specializing in different technological inputs, such as glass substrates, chemicals, filters, polarizers, etc. PDPs work in a very different way from LCDs. An electrical charge is run through a subpixel, energizing a plasma gas contained in a pocket inside the display. The plasma gas excites a colored phosphorous dot painted on the inside of the display, and the color becomes visible. The approach uses a significant amount of electricity, and for some time researchers had difficulty producing sufficiently bright displays. It has the advantages of virtually no response time and wide viewing angle. On the other hand, it has proven challenging to make very high pixel densities because of the difficulty of making and filling very small pockets (Weber et al., 2008). In Table 1, we provide a summary of the relative advantages and disadvantages of the two technologies. As we explain below, we believe that LCD’s triumph over PDP was not necessarily due to its being an obviously “superior” display technology for all applications. Table 1. Comparison of known technological advantages and drawbacks due to mechanisms employed by LCD and PDP (early 1990s) Technology Advantages Drawbacks LCD Low power consumption High pixel density possible Reliability/long life Brightness Difficult to make large displaysa Blurry motiona Narrow viewing anglea Costlya and complex production High plant investment required Difficult to make pure blacksa Low production yieldsa PDP Broad viewing angle Fast response time (no blur) Relatively easy to make large displays Simpler fabrication More modest investment requirements High power consumption Very high pixel density difficult Burn ina Technology Advantages Drawbacks LCD Low power consumption High pixel density possible Reliability/long life Brightness Difficult to make large displaysa Blurry motiona Narrow viewing anglea Costlya and complex production High plant investment required Difficult to make pure blacksa Low production yieldsa PDP Broad viewing angle Fast response time (no blur) Relatively easy to make large displays Simpler fabrication More modest investment requirements High power consumption Very high pixel density difficult Burn ina a With technological advancement, these drawbacks were dramatically reduced or eliminated. Therefore, the degrees of these drawbacks varied over time. Open in new tab Table 1. Comparison of known technological advantages and drawbacks due to mechanisms employed by LCD and PDP (early 1990s) Technology Advantages Drawbacks LCD Low power consumption High pixel density possible Reliability/long life Brightness Difficult to make large displaysa Blurry motiona Narrow viewing anglea Costlya and complex production High plant investment required Difficult to make pure blacksa Low production yieldsa PDP Broad viewing angle Fast response time (no blur) Relatively easy to make large displays Simpler fabrication More modest investment requirements High power consumption Very high pixel density difficult Burn ina Technology Advantages Drawbacks LCD Low power consumption High pixel density possible Reliability/long life Brightness Difficult to make large displaysa Blurry motiona Narrow viewing anglea Costlya and complex production High plant investment required Difficult to make pure blacksa Low production yieldsa PDP Broad viewing angle Fast response time (no blur) Relatively easy to make large displays Simpler fabrication More modest investment requirements High power consumption Very high pixel density difficult Burn ina a With technological advancement, these drawbacks were dramatically reduced or eliminated. Therefore, the degrees of these drawbacks varied over time. Open in new tab 3.2 Early development of LCD and PDP We next examine the history of development of LCD and PDP to gain an understanding of how they became competing technologies, eventually converging on the TV set application. Both PDP and LCD experienced a long lag between early discoveries and initial commercialization, and there was another lag between initial commercialization and when the technologies competed with one another more directly in the TV set application. Original discoveries of liquid crystals and plasma both predate broadcast television. Bell Labs developed a working plasma display in 1927; however, it lacked a suitable way to drive the display. In 1936, Marconi received the first LCD patent, which cited TV as a possible application (Castellano, 2005). In the 1960s, interest in both LCD and PDP returned. Researchers at the University of Illinois made breakthroughs on plasma technology in the 1960s. In the late 1960s and early 1970s companies working on PDP included AT&T, Burroughs, Control-Data, Fujitsu, Hitachi, IBM, Mitsubishi, NEC, Oki, Owens-Illinois, Sony, Texas Instruments, and Thomson (Weber et al., 2008). Japanese national broadcaster Nihon Hoso Kyoku (NHK) began research into plasma as a next generation TV set display (Weber et al., 2008). RCA’s 1968 announcement that it had developed a flat screen LCD display was a major milestone in LCD’s development (Murtha et al., 2001). RCA initially declared that wall hanging TV sets would soon come to market, but ceased LCD development in 1974 after concluding that LCDs TVs were not feasible (Murtha et al., 2001). From 1960 to 1971 Japanese companies including Citizen, Hitachi, Hoshiden, NEC, Nanox, Panasonic, Sanyo, Seiko, and Sharp began developing the technology (Sangyo Times, 1994, 1995; Numagami, 1999). 3.3 Early LCD and PDP applications Many Japanese observers consider 1973 the beginning of the LCD age because the first consumer LCD products hit the market in this year (Sangyo Times, 1994). Seiko and Citizen introduced the first commercially available LCD watches, and Sharp introduced its landmark LCD calculator, the EL-805 (Sangyo Times, 1994; Numagami, 1999; Murtha et al., 2001). In both watches and calculators, products incorporating a LCD quickly gained market share. The type of LCDs used in these applications were simple but had advantages in weight and power consumption over existing display technologies such as light emitting diode. These applications provided ready markets for LCD producers to sell into. Both were large volume consumer markets. Use in other consumer electronics and automotive display applications soon followed. Computer displays were the primary application for PDPs in the 1970s. IBM, Owens-Corning, and several other firms entered the market (Murtha et al., 2001; Weber et al., 2008; Bitzer et al., 2010). These PDPs were monochrome but had high resolution. However, computers were rare and had very limited graphics capabilities. Most computers used text only displays. Given the size and needs of the computer market at the time, early markets for PDP provided limited sales opportunities. Lacking a major application, interest in PDP technology waned. Over the period 1977–1987, most Japanese and US-based firms working on PDP discontinued their efforts (Nihon Hoso Kyoku, 2003; Bitzer et al., 2010).1 At the end of the 1970s, PDP lacked a significant market to support investments in product development or production. However, LCD had expanded its scope into a series of adjacent application markets with technical requirements LCD was able to meet even at its relatively basic level of development. Some of these markets, such as calculators and watches were large, and allowed LCD production scale to increase, supporting investments in production and development. 3.4 LCD and PDP in the 1980s As research and design on LCD moved forward, efforts to develop larger matrix screens took center stage. AM-LCD and passive matrix (PM-LCD) were two different ways of turning on and off pixels on an LCD screen. Both were capable of displaying color and had low enough power requirements that they could run on batteries. PM-LCD suffered from poor image quality (washed-out look, low contrast, and visible lines running through the display); however, it was inexpensive to produce. AM-LCD provided dramatically better image quality than PM-LCD but was difficult and expensive to manufacture. In the mid-1980s, a number of Japanese firms including Fujitsu, Sharp, Toshiba, Tottori Sanyo, began producing PM-LCDs, primarily aimed at portable computer and standalone word processors applications (Sangyo Times, 1992, 1994). Many firms began R&D activities related to AM-LCD technology in the early 1980s, but the technology proved difficult to master and took a long time to mature (Sangyo Times, 1990, 1992, 1994, 1995, 1997; Castellano, 2005). AM-LCD mass production did not begin until later. In 1985, LCD production by the five largest Japanese producers was 350 million units, valued at $477 million (Castellano, 2005). The world LCD market for that year was valued at $876 million (Castellano, 2005). Numagami (1999) notes that the LCD market grew by three times over the period 1981–1988, and that the composition changed from being almost all clock and watch applications in 1981 to computer and word processor applications taking nearly 50% share by 1988. Although interest in PDP had waned, a number of firms developed color PDP in the 1980s. Plasma display technology was adopted in some small volume commercial applications, such as banking displays. Monochrome PDP screens were used in “luggable” portable computers starting around 1986, marking the technology’s entry into a new application (Sangyo Times, 1990; Nikkei Business Publications, 1999; Menzawa et al., 2008). In 1989, Matsushita Electronics Corporation, the largest PDP supplier, had a US$150 million PDP business (Weber et al., 2008: 73). Plasma screens had high resolution but did not offer color and required an AC power source, limiting their attractiveness for portable computing. At the end of the 1980s, the scope of PDP application remained narrow, and scale was also limited with only one application market. Meanwhile, LCD had grown its scope substantially from applications markets with very basic display requirements to those with higher technological hurdles including larger screen sizes and more pixels. The relatively large market sizes of these applications resulted in significant increases in production scale. This combination of increasing scope opportunities in new and promising application markets further increased the velocity of investment in LCD R&D and production. 3.5 Early and mid-1990s The year 1990 was an important one for AM-LCD, as the first mass production facilities began to come online (Nikkei Business Publications, 1990). Initially, production yields were very low (2% for one of Sharp’s first products) due to defective transistors caused by contamination (Nikkei Business Publications, 2005). In addition to AM-LCD’s yield problem, investment requirements were high. Together, these translated into very expensive screens, typically costing over $1000 US each at the time (Murtha et al., 2001). PM-LCD technology continued to compete alongside AM-LCD for years because of its cost advantage. The use of PDP in portable computers peaked in 1991, just as AM-LCD laptops were being introduced by numerous players. In that same year, $758 million of LCD panels were sold to the portable computer market, including AM- and PM-LCD, making portable computing the single largest LCD application market by value (Castellano, 1992). While both LCD and PDP appeared to converge on portable computing, this application resulted in them taking very different paths. The ideal portable computer display at the time would have been thin, lightweight, rugged, large enough for word processing or spreadsheet applications, have reasonably good image quality, and be able to run on battery power. LCD fit these requirements better than PDP. Aside from cost, LCD’s other weaknesses were relatively unimportant in the portable computing market. With slow computer processors, motion video was not important in laptops, so LCD’s slow response time was not a major problem. Also, because laptop displays were primarily viewed by the individual user, wide viewing angles were not needed. AM-LCD had the potential to fit the requirements of the laptop application better than other technologies but realizing this potential required investment to improve production quality and yield. Arguably, the munificence of portable computer display market made it an ideal one to strengthen AM-LCD’s potential for other applications. Initial buyers of laptops were primarily business users, who were not very price sensitive. Furthermore, the market was growing strongly, providing ongoing opportunities for scale increases. Later improvements in processing power and operating system software made display quality more important, further emphasizing the benefits of AM-LCD over alternatives. Experience and additional investment in AM-LCD development paid off by leading to yield and quality improvements, thereby decreasing the unit costs. These improvements and increases in possible display size were facilitated by learning and somewhat unintentional knowledge sharing, across the LCD producers, input providers, and production equipment manufacturers (Akabane, 2014). As production costs decreased, AM-LCD was able to enter new application markets. One of the largest of these applications was desktop computer displays, where LCD competed with low-cost, high-quality CRT displays. LCDs started to be used in desktop applications around 1997 (Castellano, 2005). As display panel prices fell, LCDs gradually replaced CRT computer monitors, but it took years. Monitors later became another very large market for LCD. LCD also found opportunities in a variety of other applications outside of the computing arena, including aviation displays, car navigation systems, portable games and pachinko games. During this period, PDP had access to far fewer adjacent applications markets than LCD. Large color PDPs were developed in this period and PDP had a clear advantage in terms of screen size. However, it also had drawbacks; notably high power consumption, which limited its portability. PDPs were used in specialized industrial monitors and public displays such as those found in airports and train stations. Neither of these applications provided the size of opportunity that LCD enjoyed in its larger markets. In the early to mid-1990s, the scope of LCD usage grew greatly as it entered a large number of adjacent application markets, increasing scope and scale of application. Of these, the application market for laptop computer displays was particularly pivotal, attracting strong investment in R&D and production facilities to meet its challenging quality requirements and substantial demand. Moreover, the scope of PDP application did not grow as much, and the limited scale of its adjacent applications markets did not result in the large investments seen in LCD technology. Table 2 summarizes the two technologies fit with major applications markets prior to TV sets. Table 2. Major LCD and PDP applications prior to TV sets Time period Tech Application market Market Characteristics Market requirements Technology fit with requirements Outcome 1971–1975 PDP Computer monitor Small market—computers rare Middle display size; single color acceptable; low resolution Limited; PDPs had relatively high-resolution graphics, but most computers only used text Failed to gain traction, firms exited application 1973– LCD Watches and calculators Large consumer market Low energy usage; low cost; rugged Very good Sales took off in these applications, led to adoption in applications with similar needs; more advanced matrix displays later adopted into calculators 1986–1991 PDP Portable computing display Small market; willing to pay premium Portability; reasonable image quality; screen size 8–12 inches; low energy consumption Limited; Plasma’s high energy consumption required AC power Initially successful in market, but as LCD entered, PDP sales in the application peaked, manufacturers exited 1986– LCD (PM) Word processor and portable computing display Medium size market; Word processor price sensitive Very good, able to run on battery power; PM-LCDs used in price-sensitive segments Although image quality not very good, low cost, portability, and ability to run on battery power spur investment and scale increase 1991– LCD (AM) Laptop display Initially small but rapidly growing market; willing to pay premium; Becomes LCD’s largest application market in 1991 Portability; better image quality; screen size 8–12 inches; low energy consumption Very good, able to run on battery power; AM-LCD in less price-sensitive segments until production costs fell Large market potential and technology fit attracted entry and investment. Cost reductions and quality improvement resulted in adoption in new market applications 1996– LCD (AM) Desktop monitor Very large market, but price sensitive (CRT a competitor) Screen size at least 12 inches; wide viewing angle; speed for graphics Improved after production costs fell and ability to produce larger screens increased As costs improved, LCD displaced the legacy CRT technology in this arena; provided a large demand market for LCD that also valued larger screens at lower prices; this helped segue to TV sets 1997– PDP Specialized monitor Small market; some segments price sensitive Mixed data display usages including computers, industrial operations Poor. Benefits of display flatness not enough to justify higher price than CRT. Small markets; LCDs eventually took over the market space 1998– PDP Public monitor Small market Large display size. Wall mountable. Ability to show data Good. Requirements for larger displays than available from LCD shielded area from competition Small market, provides opportunities Time period Tech Application market Market Characteristics Market requirements Technology fit with requirements Outcome 1971–1975 PDP Computer monitor Small market—computers rare Middle display size; single color acceptable; low resolution Limited; PDPs had relatively high-resolution graphics, but most computers only used text Failed to gain traction, firms exited application 1973– LCD Watches and calculators Large consumer market Low energy usage; low cost; rugged Very good Sales took off in these applications, led to adoption in applications with similar needs; more advanced matrix displays later adopted into calculators 1986–1991 PDP Portable computing display Small market; willing to pay premium Portability; reasonable image quality; screen size 8–12 inches; low energy consumption Limited; Plasma’s high energy consumption required AC power Initially successful in market, but as LCD entered, PDP sales in the application peaked, manufacturers exited 1986– LCD (PM) Word processor and portable computing display Medium size market; Word processor price sensitive Very good, able to run on battery power; PM-LCDs used in price-sensitive segments Although image quality not very good, low cost, portability, and ability to run on battery power spur investment and scale increase 1991– LCD (AM) Laptop display Initially small but rapidly growing market; willing to pay premium; Becomes LCD’s largest application market in 1991 Portability; better image quality; screen size 8–12 inches; low energy consumption Very good, able to run on battery power; AM-LCD in less price-sensitive segments until production costs fell Large market potential and technology fit attracted entry and investment. Cost reductions and quality improvement resulted in adoption in new market applications 1996– LCD (AM) Desktop monitor Very large market, but price sensitive (CRT a competitor) Screen size at least 12 inches; wide viewing angle; speed for graphics Improved after production costs fell and ability to produce larger screens increased As costs improved, LCD displaced the legacy CRT technology in this arena; provided a large demand market for LCD that also valued larger screens at lower prices; this helped segue to TV sets 1997– PDP Specialized monitor Small market; some segments price sensitive Mixed data display usages including computers, industrial operations Poor. Benefits of display flatness not enough to justify higher price than CRT. Small markets; LCDs eventually took over the market space 1998– PDP Public monitor Small market Large display size. Wall mountable. Ability to show data Good. Requirements for larger displays than available from LCD shielded area from competition Small market, provides opportunities Open in new tab Table 2. Major LCD and PDP applications prior to TV sets Time period Tech Application market Market Characteristics Market requirements Technology fit with requirements Outcome 1971–1975 PDP Computer monitor Small market—computers rare Middle display size; single color acceptable; low resolution Limited; PDPs had relatively high-resolution graphics, but most computers only used text Failed to gain traction, firms exited application 1973– LCD Watches and calculators Large consumer market Low energy usage; low cost; rugged Very good Sales took off in these applications, led to adoption in applications with similar needs; more advanced matrix displays later adopted into calculators 1986–1991 PDP Portable computing display Small market; willing to pay premium Portability; reasonable image quality; screen size 8–12 inches; low energy consumption Limited; Plasma’s high energy consumption required AC power Initially successful in market, but as LCD entered, PDP sales in the application peaked, manufacturers exited 1986– LCD (PM) Word processor and portable computing display Medium size market; Word processor price sensitive Very good, able to run on battery power; PM-LCDs used in price-sensitive segments Although image quality not very good, low cost, portability, and ability to run on battery power spur investment and scale increase 1991– LCD (AM) Laptop display Initially small but rapidly growing market; willing to pay premium; Becomes LCD’s largest application market in 1991 Portability; better image quality; screen size 8–12 inches; low energy consumption Very good, able to run on battery power; AM-LCD in less price-sensitive segments until production costs fell Large market potential and technology fit attracted entry and investment. Cost reductions and quality improvement resulted in adoption in new market applications 1996– LCD (AM) Desktop monitor Very large market, but price sensitive (CRT a competitor) Screen size at least 12 inches; wide viewing angle; speed for graphics Improved after production costs fell and ability to produce larger screens increased As costs improved, LCD displaced the legacy CRT technology in this arena; provided a large demand market for LCD that also valued larger screens at lower prices; this helped segue to TV sets 1997– PDP Specialized monitor Small market; some segments price sensitive Mixed data display usages including computers, industrial operations Poor. Benefits of display flatness not enough to justify higher price than CRT. Small markets; LCDs eventually took over the market space 1998– PDP Public monitor Small market Large display size. Wall mountable. Ability to show data Good. Requirements for larger displays than available from LCD shielded area from competition Small market, provides opportunities Time period Tech Application market Market Characteristics Market requirements Technology fit with requirements Outcome 1971–1975 PDP Computer monitor Small market—computers rare Middle display size; single color acceptable; low resolution Limited; PDPs had relatively high-resolution graphics, but most computers only used text Failed to gain traction, firms exited application 1973– LCD Watches and calculators Large consumer market Low energy usage; low cost; rugged Very good Sales took off in these applications, led to adoption in applications with similar needs; more advanced matrix displays later adopted into calculators 1986–1991 PDP Portable computing display Small market; willing to pay premium Portability; reasonable image quality; screen size 8–12 inches; low energy consumption Limited; Plasma’s high energy consumption required AC power Initially successful in market, but as LCD entered, PDP sales in the application peaked, manufacturers exited 1986– LCD (PM) Word processor and portable computing display Medium size market; Word processor price sensitive Very good, able to run on battery power; PM-LCDs used in price-sensitive segments Although image quality not very good, low cost, portability, and ability to run on battery power spur investment and scale increase 1991– LCD (AM) Laptop display Initially small but rapidly growing market; willing to pay premium; Becomes LCD’s largest application market in 1991 Portability; better image quality; screen size 8–12 inches; low energy consumption Very good, able to run on battery power; AM-LCD in less price-sensitive segments until production costs fell Large market potential and technology fit attracted entry and investment. Cost reductions and quality improvement resulted in adoption in new market applications 1996– LCD (AM) Desktop monitor Very large market, but price sensitive (CRT a competitor) Screen size at least 12 inches; wide viewing angle; speed for graphics Improved after production costs fell and ability to produce larger screens increased As costs improved, LCD displaced the legacy CRT technology in this arena; provided a large demand market for LCD that also valued larger screens at lower prices; this helped segue to TV sets 1997– PDP Specialized monitor Small market; some segments price sensitive Mixed data display usages including computers, industrial operations Poor. Benefits of display flatness not enough to justify higher price than CRT. Small markets; LCDs eventually took over the market space 1998– PDP Public monitor Small market Large display size. Wall mountable. Ability to show data Good. Requirements for larger displays than available from LCD shielded area from competition Small market, provides opportunities Open in new tab 3.6 Late 1990s and early 2000s: competition between LCD and PDP in flat panel TVs A comparison of the attributes of the different display technologies with TV set requirements suggest that PDP fit best. It had fast response time, wide viewing angle, good color attributes, was relatively easy to make in large sizes, and had modest investment requirements compared with alternative display technologies. Indeed, during the 1990s many TV set manufacturers considered it the best large TV technology, as did image quality gurus.2 By comparison, the fundamental characteristics of LCD do not appear to fit the TV set requirements well in terms of size, image quality, speed and viewing angle. However, LCD’s ongoing increases in application scope and scale of adoption supported strong investment in production and R&D. Over time, ongoing R&D investments led to amelioration of LCD shortcomings vis-à-vis the TV set application (such as response speed and contrast). Next, we examine LCD’s progress relative to PDP along several important technological attributes. Historically, LCD suffered from a narrow viewing angle; however, LCD was later improved to the point where viewing angles of LCD and PDP were not significantly different. Table 3 tracks the progress of the “state-of-the-art” displays (note: mass-market displays lagged behind these). Viewing angles were not an issue for PDP, but the specifications generally show it to have been in the area of 160°. Around the year 2000, cutting edge LCD panels approached the viewing angle of PDPs. Table 3. Specifications state-of-the-art production displays, by display technology and year Display technology Measure 1992 1994 1996 1998 2000 2002 2004 LCD Viewing angle (degrees) 45 50 60 140 150 170 180 Response time (ms) 50 30 22 19 14 6 6 Max size (diagonal inches) 10.2 15 15 23 28 40 55 PDP Max size (diagonal inches) 21 42 50 60 61 80 Display technology Measure 1992 1994 1996 1998 2000 2002 2004 LCD Viewing angle (degrees) 45 50 60 140 150 170 180 Response time (ms) 50 30 22 19 14 6 6 Max size (diagonal inches) 10.2 15 15 23 28 40 55 PDP Max size (diagonal inches) 21 42 50 60 61 80 Sources: Compiled by authors from Fuji-Chimera, Nikkei Business Publications. Data for two missing years were estimated by authors. Open in new tab Table 3. Specifications state-of-the-art production displays, by display technology and year Display technology Measure 1992 1994 1996 1998 2000 2002 2004 LCD Viewing angle (degrees) 45 50 60 140 150 170 180 Response time (ms) 50 30 22 19 14 6 6 Max size (diagonal inches) 10.2 15 15 23 28 40 55 PDP Max size (diagonal inches) 21 42 50 60 61 80 Display technology Measure 1992 1994 1996 1998 2000 2002 2004 LCD Viewing angle (degrees) 45 50 60 140 150 170 180 Response time (ms) 50 30 22 19 14 6 6 Max size (diagonal inches) 10.2 15 15 23 28 40 55 PDP Max size (diagonal inches) 21 42 50 60 61 80 Sources: Compiled by authors from Fuji-Chimera, Nikkei Business Publications. Data for two missing years were estimated by authors. Open in new tab Response speed, another limiting factor in LCD’s suitability for displaying moving video, also improved dramatically over time. Response speed was never considered an issue for PDP, and therefore specifications have not been reported. Humans have difficulty noticing improvements in response times below 10 ms; however, response times much slower than this result in perceptible blur (Artamonov, 2007). The data suggest that response speed ceased to be a technological disadvantage in state-of-the-art LCDs by 2002. Size was another hurdle LCD faced in the TV set applications. While LCD started out offering small screen TVs and PDP with large ones, the two technologies eventually began to converge on larger, living room TV set applications, as shown in Table 3. A Panasonic manager appearing in Nikkei FPD 2002 described the company’s viewpoint on the three TV set displays: PDP was for 50″–37″ TVs, CRT was for 36″–15″ TVs, and LCD was for 22″ and smaller TVs. However, this soon changed. An article in Sangyo Times (2004), for example, was titled “Clash in the Large Screen FPD TV Market: LCD versus PDP.”3 LCD did not match PDP in all the attributes salient for TV set applications during the period we observed. As PDP matured, it became possible to produce in ever-larger sizes. LCD did not suddenly become a good living room TV set display. However, it gradually entered the market as it became possible to fabricate LCDS in larger sizes and subsequently improve production efficiencies. LCD does not appear to have had an inherent cost advantage against PDP. PDP technology was suited for low-cost mass production, whereas LCD fabrication required large capital investments in plant and equipment and has high processing costs (Castellano, 1992). In agreement with this, an engineer quoted in Nikkei’s FPD series in 2007 stated: “plant and equipment investment required to produce PDP is about one half of what is needed for LCD; furthermore, the time it takes to produce PDPs is one fifth of what is needed for LCD panels.”4 Comparison of similar capacity LCD and PDP plant investments made in the same timeframe backs up the assertion that PDP investment requirements were lower (see Sangyo Times series). PDP does not require cleanroom environments as LCD does. Driver electronics for PDP are more expensive than those used in LCD, balancing off PDP’s advantage in fabrication to some degree. Although LCD had inherent attributes which made it relatively costly to produce, the speed at which costs fell over time varied greatly between LCD and PDP. This was due to differing investment levels which lead to differences in scale economies, as well as differences in the competitive dynamics of the industries producing LCD and PDP. We will return to these points later. Figure 1 below depicts the pricing of comparable LCD and PDP displays for television use during the 2004–2006 timeframe. These data suggest that LCD was able to catch up with PDP costs for larger panels over time, but it took years. Figure 1. Open in new tabDownload slide Prices of PDP and LCD TV modules with XGA resolution, 2004–2006 (Units: Japanese Yen). Source: Compiled by authors using data from Fuji-Chimera, 2006–2007. Figure 1. Open in new tabDownload slide Prices of PDP and LCD TV modules with XGA resolution, 2004–2006 (Units: Japanese Yen). Source: Compiled by authors using data from Fuji-Chimera, 2006–2007. Above, we have examined several kinds of technical attributes where LCD was at a disadvantage to PDP for use in TV sets, and we tracked how LCD improved enough to become a meaningful contender. Competition between the two technologies continued for some time. For years, PDP remained the quality standard according to experts (e.g., Katzmaier, 2011). Eventually, LCD narrowed the gap and improved to the point that few consumers could discern a quality difference. Pioneer, which had led Plasma TV in terms of image quality, announced it was exiting PDP production in 2008 (Pioneer, 2008). The company had not been successful in reducing costs at the same rate that market prices were falling and had also found that it was difficult to command premium prices even for the highest quality in the TV set market (The Mainichi Newspaper, 2009). As penetration rates in developed markets grew higher in the late 2000s, the TV market became a source of mounting losses for consumer electronics companies. New investment in LCD and PDP and TV set production was curtailed (with some exceptions, mainly in China), and some firms exited production altogether. Finally, in 2013 (after the end of our study period), Panasonic, which had been a particularly strong supporter of Plasma in the TV set area, announced it would exit PDP production. 4. Discussion 4.1 LCD and PDP applications markets The historical evidence suggests that LCD was able to grow both application scope and scale by entering many adjacent applications markets—some of which were very large and munificent (i.e., rich in resources)—while PDP experienced fewer such opportunities prior to TV sets. We conceptualize these developments in terms of a cycle of development and investment. Relevant application markets can be represented as rings in the cycle. The cycle begins with applications having relatively modest technological requirements and moves in succession to those applications for which the technological requirements are adjacent in the sense that they are similar in nature or level of technological difficulty. In LCD’s case, shown in Figure 2, segment LCD attributes fit well with the requirements of the earliest and simplest applications, watches and calculators. These markets were also large, promoting investments in production technology, cost reductions and further growth in production. The successes and revenue streams further supported R&D in LCD product technology. Following these early markets, adjacent markets for matrix forms of LCD technology developed, with the emergence of the portable computer and word processor markets as two of the larger adjacent markets around 1986. PM-LCDs had many image quality issues, but nonetheless were able to sell and attract additional investment, thereby again resulting in gains in scale and cost reduction. Around 1991, AM-LCD screens began to enter mass production, and were sold into the laptop computer market. This market was munificent, large, and demanding of higher display image quality levels than PM-LCD was able to provide. Figure 2. Open in new tabDownload slide LCD scale and scope cycle. Figure 2. Open in new tabDownload slide LCD scale and scope cycle. Continuing investments yielded successive breakthroughs in product and production technology. These allowed LCD to enter successive adjacent application markets such as desktop computer displays. Overall, the munificence of the adjacent markets promoted increased investments in production and scale, forming a virtuous cycle that allowed LCD to eventually become a serious contender for the Flat Panel TV application. In PDP’s case, shown in Figure 3, on the other hand, early applications resulted in dead ends, limiting the scope of application and the level of investment into product and production technology for PDP. However, PDP was still able to become a TV set technology because the technological hurdles were low enough that they could be bridged with relatively modest level of investment. After PDP was adopted as a TV set display technology, it was able to gain momentum and increase in scale. However, by that time, LCD had substantial momentum and was rapidly addressing the cost and image quality issues it faced. Figure 4 depicts the markets, measured in number of panels, for several major LCD applications, and all PDP applications. LCD had substantial volume in the laptop application, which was followed by rapid growth in use as a desktop monitor. The flat panel TV application did not materialize in volume until well after these other applications were developed markets. In the time frame around 2003–2004, both LCD TV and PDP show significant increases in volume, but growth was faster for LCD. Figure 3. Open in new tabDownload slide PDP scale and scope cycle. Figure 3. Open in new tabDownload slide PDP scale and scope cycle. Figure 4. Open in new tabDownload slide Market size of PDP and major LCD display application (Units: 100,000 panels). Source: Fuji-Chimera. Figure 4. Open in new tabDownload slide Market size of PDP and major LCD display application (Units: 100,000 panels). Source: Fuji-Chimera. 4.2 Access to adjacent markets as a driver of technological competitiveness We submit that LCD’s ability to compete with, and eventually surpass, PDP in the TV set market was made possible by its adoption and development in numerous applications, which were adjacent in their technological requirements. Some of these application markets were large and munificent. Above, we described the adjacent markets in terms of a cycle, which is helpful because it shows interconnection between the different levels of development and the mechanisms of continued investment through growth in scale and scope. However, the succession of adjacent markets can also be thought as being like a pathway consisting of different levels. In LCD, each successive step (an adjacent market) allowed the technology to be improved, but none of required improvements was so high as to be insurmountable on its own. Figures 5 and 6 depict the major adjacent market opportunities, entered by LCD and PDP, respectively, using this pathway metaphor. Figure 5. Open in new tabDownload slide Conceptual drawing of LCD adjacent market pathway. Figure 5. Open in new tabDownload slide Conceptual drawing of LCD adjacent market pathway. Figure 6. Open in new tabDownload slide Conceptual drawing of PDP adjacent market pathway. Figure 6. Open in new tabDownload slide Conceptual drawing of PDP adjacent market pathway. Early applications such as calculators and watches spurred continued investment in the technology. More challenging but munificent applications, in particular laptop displays required higher quality, larger displays, which in turn required bridging difficult manufacturing hurdles. Compared with the laptop market, desktop displays required broader viewing angles (although not as broad as needed for TV sets), lower costs, and larger sizes. Efforts to meet the needs of these markets brought LCD closer to meeting the requirements of the TV display market. Also, in each of the major adjacent markets, scale increased significantly, offering strong rationale for additional investment while also pushing unit costs down and thereby further amplifying the potential of growth into cost-sensitive segments and price-sensitive adjacent markets. With this momentum, LCD developed greater investment, attention, and reputation. PDP, on the other hand, had few adjacent application markets. Consequently, it received much less investment than LCD. Over the period 1996–2002, the leading LCD patenting firm received over 2500 related patents, while the leading PDP patenting firm had less than 500 (KSR Ltd\., 2003). Despite the lower levels of investment, PDP was able to enter and compete in the flat panel TV industry because the jump from earlier technology stages of PDP was much less challenging than the jump from early LCD technology to the required level for TV sets. In spite of PDP’s technological advantages in the TV set application, it lost out to LCD overall. How might the technologies have developed had there been no adjacent market opportunities before TV sets? Although this relies on some conjecture, we can make the following observations. LCD was only able to provide a mature TV set display technology after about 30 years of continued investment in R&D and production since the first LCD calculator. The total amount of R&D is difficult to estimate. However, to provide some sense of scale, investment in PDP production facilities in Japan over the period 1990–1999 was a mere 6% of that in LCD production facilities (roughly $18 billion USD at 100 JPY/USD) over the same period. These high levels of investment were clearly supported by the different munificent applications markets that LCD was able to sell into along its development pathway. These were also necessary in order to achieve the image quality, cost competitiveness, and display size requirements for the TV set application. It is quite hard to imagine that Japanese managers would be willing to sustain such large investments without the ongoing revenue and profit streams of the different application markets. Meanwhile, despite comparably quite modest investments5 in R&D and production equipment, PDP was able to successfully develop and deploy large screen TV set displays prior to LCD’s entry. As we mentioned before, RCA initially announced that it was developing LCD as a flat panel TV set in 1968, but in 1974 concluded the technological barriers were too high. Interestingly, one of our informants, a Sharp engineer closely involved in the company’s LCD and LCD TV businesses, stated that RCA’s conclusion was correct and rational. If LCD had not enjoyed a good fit with these often munificent, mass-market applications, it is extremely unlikely that LCD would have been able to compete against PDP in the TV sets or any other large screen applications. Thus, the most interesting part of the story is not how LCD eventually defeated PDP in TV sets, but given its initial technological disadvantages, how it evolved down a pathway resulting in dominance as a TV display technology. Other things besides the availability of adjacent market pathways may also systematically affect the development of technology over time. In the case of this industry, the role of managerial agency and firm strategy, organizational backing, and impacts of the boom and bust crystal cycle appear to be relevant. Below, we consider each of these. 4.3 The role of managerial agency and firm strategy What role did managers and firm strategies play in the competition between LCD and PDP? From a very early time it was realized that FPD technology had strong potential in the TV market. However, there was uncertainty about which display technology would become dominant. Some firms made large bets on either LCD or PDP (e.g., Sharp, Panasonic, Pioneer), while others made smaller bets on both (e.g., Fujitsu, Hitachi, NEC), or avoided betting on either (e.g., Sony). Sharp was perhaps the most influential single firm in LCD’s development. Sharp’s managers consistently invested into LCD. The company also focused R&D resources away from semiconductors and into LCD. Furthermore, Sharp’s approach to LCD was consistent with the mechanism discussed here. In order to obtain access to further investments, Sharp managers first had to make sales of LCDs. This motivated managers to introduce LCD into successive adjacent applications as the technology matured enough to meet the applications requirements. Indeed, the first LCD calculator was a Sharp product. Sharp’s LCD Viewcam, a video camera incorporating a large LCD screen, was another product that pushed the envelope of LCD applications at the time. Sharp was also one of the first producers of laptop displays. Finally, Sharp developed its TV strategy around LCD technology, with its president Machida’s 1998 proclamation that Sharp’s TV sets would all be LCD TVs by 2005 (Sankei Shimbun, 1998). These pieces of evidence suggest that Sharp’s managers consistently took advantage of the opportunities present in adjacent market segments with the belief that LCD could eventually succeed in the TV market. Of course, this belief does not explain the existence of the adjacent application market opportunities themselves. Would these have been developed if there had not been a firm in the industry which, like Sharp, was focused developing each of these opportunities as it unfolded? We believe the answer is yes. Sharp may stand out as a firm that most consistently followed this approach, but the applications were known by engineers and managers in other firms. While Sharp may have been the single firm that most persistently searched out and entered these markets, it wasn’t the only player to pursue any of these opportunities. Sharp was consistently an early mover into new applications markets. It is, therefore, possible that the speed of LCD’s development might have been somewhat slower had Sharp not adopted this strategy. On a broader level, LCD and PDP players followed related diversification and vertical integration strategies in their electronics businesses. These strategies put them in a position to invest into new and developing application niches, from both offensive and defensive perspectives. In the offensive, diversification into adjacent applications markets offered the potential for new growth. In terms of defensive perspective, vertical integration increased exit barriers, and thereby gave firms incentive to develop alternative applications markets in response to decreases in market prices for display panels (related to the crystal cycle discussed later). Would LCD and PDP technologies have developed differently if these firms had not followed vertically integration with diversified product portfolios? We cannot know for certain. However, the mechanism of development of a technology through entry into a sequence of adjacent markets does not appear to rely heavily on corporate level strategy for its effectiveness. So long as there was sufficient communication between display producers and would-be users (e.g., watch makers, video camera makers, laptop manufacturers, etc.) about market opportunities and technological requirements, these adjacent market opportunities would have been identified and developed. Examination of the applications LCD eventually entered finds evidence of niche applications in markets LCD producers themselves did not integrate into, such as avionics, industrial control equipment, automotive displays, and pachinko game machines. 4.4 Differences in organizational backing for LCD and PDP efforts Development efforts for PDP and LCD differed in several important ways, in particular, the number of firms involved in each of the technologies, the degree to which development was proprietary or not, and government involvement in the development efforts. Below, we examine these differences in organizational backing and consider how these differences may have affected the competition between LCD and PDP. The data clearly show that more firms were involved in developing LCD than PDP. There are two possible explanations. One, the broad set of potential applications for LCD attracted a larger variety of firms, for example, IT firms (e.g., NEC, Fujitsu), watch making firms (e.g., Seiko, Citizen), consumer electronics firms (e.g., Sharp, Toshiba), and pure play display firms (e.g., Hoshiden). This is consistent with the tendency for Japanese electronics firms to adopt vertical integration strategies. At the same time, it fits with our argument that earlier adjacent market opportunities supported the required investment in LCD to allow it to compete with PDP in large screen applications. Another potential explanation is that PDP may have had higher barriers to entry than LCD, due to its proprietary nature and despite its lower capital requirements for plant and equipment. LCD production equipment was primarily made by equipment manufacturers who also supplied the semiconductor industry and was less proprietary in nature than PDP equipment. As LCD equipment grew more sophisticated, it incorporated solutions to many manufacturing issues, developed through coevolution between the equipment producers and industry participants (Jenkins and Floyd, 2001). In effect, much of the technological progress became accessible to everyone in the industry, and the availability of this equipment also reduced barriers to would-be entrants (Jovanovic and MacDonald, 1994). Having said that, Akabane (2014) argues that the fifth generation LCD equipment, which started to come online in 2002, was the first to incorporate all the major improvements required to significantly reduce technological entry barriers to larger panel production. By this time, however, the capital requirements had grown to represent a substantial entry barrier. By comparison, PDP equipment while more proprietary was simpler and less costly. Given these observations, the differential in entry barriers alone does not provide a convincing explanation for the difference in number of players in the two technologies. Both technologies received government support, although this does not appear to have had a substantial impact on development. Japan’s Ministry of International Trade and Industry (MITI) instigated cooperation between LCD-related firms in 1989, under the name Giant Electronics Technology Corporation; however, this was not successful in making breakthroughs in its mission to make large-scale AM-LCD manufacturing commercially viable (Stolpe, 2002). NHK, the Japanese national public broadcaster, sponsored a PDP-TV development group. Work was done inside of NHK, but the broadcaster also organized firms working on PDP. Beyond the fact that NHK was successful in publicizing PDP through public displays of the Nagano Winter Olympics, little is publicly known about the effort’s impact on technological development. Altogether, the impact of these different arrangements appears to have been limited. Japanese consortia have been found to be ineffective at technological development when they included firms that competed in resulting end product markets (Branstetter and Sakakibara, 2002). 4.5 The impact of the LCD crystal cycle The AM-LCD industry has regularly experienced a boom and bust cycle, called the “crystal cycle” (Mathews, 2005). The cycle results in large swings in the availability and pricing of LCD panels. The cycle affected existing competitors and entry into the LCD business. At the bottom of the cycle, prices are low due to excess capacity. This encouraged existing LCD manufacturers to enter additional adjacent display applications. For example, during a bottom in 1995, where prices fell to one-half their level at the previous peak, firms started to make a push into desktop computer displays and other nonlaptop consumer applications (Nikkei Business Publications, 1998). With increasing demand, supply tightened again, but the overall level of production was higher than before. New firms also entered the LCD business during the downswings because entry barriers declined during these periods. There are several reasons why entry barriers fell. First, LCD production equipment manufacturers offered discounts, reducing initial investment costs. Second, existing LCD producers, eager to limit further financial damage, licensed technologies to other firms. They also set up cooperative arrangements including technology transfer, with new or would-be entrants. These factors helped Korean manufacturers to enter AM-LCD production in 1995, and Taiwanese manufacturers in 1999 (Nikkei Business Publications, 1999; Akabane, 2014). The mechanisms discussed here had positive impact on LCD in its competition with PDP. The scale increases from new generations of LCD production equipment reduced panel prices, increasing access into more price-sensitive adjacent application markets. Increases in the number of LCD manufacturers increased price competition for LCDs. Furthermore, the widespread knowledge of the severity of the cycle pressured firms to continually search for ways to reduce costs. However, these effects required the availability of additional opportunities for growth, principally in the form of untapped or underdeveloped adjacent application markets. 4.6 Toward a theory of technology competition without network externalities Above, we examined a number of alternative explanations for LCD’s defeat of PDP, including differences in organizational backing, the nature of competition, reputational differences, and the crystal cycle. Although a number of these mechanisms may have influenced the competition between LCD and PDP, it appears unlikely that these mechanisms would have had any impact if LCD had lacked access to the key applications markets discussed here. To summarize our theory, the existence of different sets of adjacent application markets for competing technologies can strongly impact how they are developed and the subsequent competitiveness of these technologies. Therefore, a technology facing an attractive set of adjacent application markets may end up beating a competing technology which has some initial fundamental design or technical advantages but lacks the right application markets for these advantages to be further developed or for scale economies to be realized. The attributes of the overall set of adjacent application opportunities are important. Not all of the applications need to be closely adjacent in terms of technological requirements, and not all need to be munificent. Those applications requiring larger leaps in technological development need to have high incentive, or a greater potential for profit to support these investments. On the other hand, firms may enter applications even if they are not large or highly profitable where only minor improvements in the technology are needed. A set of applications which includes many different market opportunities, adjacent in terms of their technological requirements, provides a pathway to development and scale increase, while limiting the need for large leaps in development or big “bets” on single applications. This makes investment decisions less risky for managers. Together with providing a pathway to developing a technology to meet the requirements of more demanding applications, these sets of application markets also provide increased scale. Scale brings experience, efficiency, and cost reduction, which further opens the technology for use in more price-sensitive applications and market niches. It also increases the technology’s cost effectiveness vis-à-vis alternative technologies. These scale impacts are not only at the firm level but also permeate the value chain. Given this, technologies facing attractive sets of application opportunities are likely to beat out those facing less attractive sets of application opportunities. This is true even if the basic technology has a fundamental disadvantage in technological performance or cost effectiveness, as was the case with PDP and LCD. However, it may not be possible to fully identify adjacent market opportunities in advance. Application niches are themselves evolving. Indeed, in many cases the niches are discovered as the markets and technology, including related supporting technologies, develop through a process of coevolution. This further complicates the identification and evaluation of such application markets. Above, we have introduced our theory of adjacent markets at the technology level. Applied to the firm level, our theory suggests potential avenues for growth and development involving diversification. Rumelt (1974) discusses different motivations for related diversification, including “exploitation of science-based research” (17). Adjacent markets represent one case of such exploitation of science-based resources of the firm, and as such provide a logic behind strategic investments into such resources. Indeed, Sharp’s investments into LCD applications markets over time suggest a strategy of related linked diversification based upon technological knowledge. Overall, our theory provides a logic that can be used to build upon the concept of related diversification in resource-based view of the firm (Rumelt, 1974). 4.7 Implications for managers Our findings have implications for managers making investments into competing technologies under uncertainty without network externalities. Firms should map out potential adjacent markets for the different technologies to better understand the opportunities each technology faces. To do this, managers should look for applications that are just a little more technologically challenging than the current state, and then continue to ask if there are additional opportunities that are only marginally more technologically challenging. This is not likely to be an easy or perfect process, as quite often the application market opportunities develop over time and may not be well understood in advance. For the application markets identified, the potential market size and growth should be estimated, and the key attributes required for success in these markets identified. Technologies that have access to pathways of adjacent applications market opportunities can be developed sequentially will have advantages against those without such a pathway. But not all pathways of adjacent applications market opportunities are the same. Those pathways having the initial market opportunities that do not require advanced versions of the technology will be more attractive than those starting with higher technological hurdles (Christensen, 2000). Where there are applications requiring a variety of different levels of the technology offer more opportunities to develop scale and scope at different levels of the technology’s development trajectory; and, relatively high technological hurdles will be more easily overcome in applications when they come along with sufficient munificence. Firms can increase their odds of investing in the winning technology in such cases by selecting technologies that have the most attractive set of potential adjacent market opportunities. Additionally, technologies facing many adjacent market applications provide opportunities to reduce investment risk because each of the adjacent markets represents a distinct decision and may provide its own payback to support the required investment. Note that in the case of competing technologies, firms still need to decide which to invest in. To fully take advantage of the perspective we propose, managers may need to shift their thinking from focusing on the impediments to development a technology faces to the market opportunities available for developing the technology. Furthermore, while we have discussed this in terms of competing technologies, this frame may also be helpful for considering the potential single technology as well. Indeed, we can conceive of situations where companies might enter into adjacent markets with the intention to polish and improve a technology before attempting to use it to disrupt other markets (Christensen, 2000). The structure of potential application markets also matters to development. Low levels of competition in an application market may retard innovation and development even if the market size is large. On the other hand, high levels of competition and entry into a technology put relentless pressure on cost and performance attributes, increasing the potential for it to outperform a competing technology that has lower levels of rivalry and entry. This presents an interesting paradox for managers because higher competition makes the technology more competitive in markets while it also reduces the potential for producers to realize advantageous profits from their investments in it. Entering or serving different applications markets, as we suggest, will encourage greater diversification of the firm. Accordingly, narrowly focused firms may find this approach unattractive. In the FPD industry, most firms were vertically integrated producers of at least some of the finished goods sold in application markets, while also selling panels to existing manufacturers serving other applications. Firms considering adopting the approach suggested here would not necessarily have to vertically integrate into the finished goods areas. However, in order to serve multiple applications, they would nonetheless have to diversify their product lineup and develop the application-specific knowledge to serve and sell to the adjacent market in question. It is also worth bearing in mind that how the market structure evolves is likely to impact firm performance. FPD manufacturers have suffered major losses in many circumstances. For LCD manufacturers, this was largely due to industry structure and the boom and bust crystal cycle (Mathews, 2005). Therefore, these kinds of technology investment decisions require careful consideration of changing industry economics. Following adjacencies helps to explain the emergence of the dominant technology, but does not necessarily result in higher profitability for the firm(s) pursuing these adjacencies. 5. Conclusion While engineers often debate the technological features, benefits, and drawbacks of technologies, our research suggests that more attention to identifying the potential set of future market opportunities is warranted. Given a pathway of adjacent applications market opportunities seemingly difficult to develop technologies can advance by making incremental improvements. Entering into new adjacent markets can increase the scale and scope of application, further fueling investment in R&D while also increasing economies of production. Paradoxically, technologies which appear to be very attractive on the basis of features will not flourish if they do not have access to market opportunities that allow them to be fully developed. Footnotes 1 Eggers (2014) concluded that PDP was essentially dead for a period in the late 1980s to the early 1990s, whereas LCD was going strong. 2 For example, Reiji Asakura of the Society of Picture Quality Engineers. 3 Translated by authors. 4 Translated by authors. 5 For example, an NHK TV documentary on PDP development describes Fujitsu’s PDP development efforts as being started by one engineer in secret in his spare time. References Abernathy W. J. , Utterback J. M. ( 1978 ), ‘ Patterns of innovation in industry ,’ Technology Review , 80 , 40 – 47 . WorldCat Adner R. , Levinthal D. A. ( 2002 ), ‘ The emergence of emerging technologies ,’ California Management Review , 45 ( 1 ), 50 – 66 . Google Scholar Crossref Search ADS WorldCat Akabane J. ( 2014 ), Higashi Ajia ekisho paneru sangyo no hatten . Keiso Shobo : Tokyo . Google Preview WorldCat COPAC Anderson P. , Tushman M. L. ( 1990 ), ‘ Technological discontinuities and dominant designs: a cyclical model of technological change ,’ Administrative Science Quarterly , 35 ( 4 ), 604 – 633 . Google Scholar Crossref Search ADS WorldCat Artamonov O. ( 2007 ), ‘Contemporary LCD Monitor Parameters: Objective and Subjective Analysis,’ http://www.xbitlabs.com/articles/monitors/display/lcd-parameters.html. Arthur W. B. ( 1989 ), ‘ Competing technologies, increasing returns, and lock-in by historical events ,’ The Economic Journal , 99 ( 394 ), 116 – 131 . Google Scholar Crossref Search ADS WorldCat Arthur W. B. ( 1996 ), ‘ Increasing returns and the new world of business ,’ Harvard Business Review , 74 ( 4 ), 100 – 109 . Google Scholar PubMed WorldCat Bitzer D. , Johnson R. L. , Webber L. ( 2010 ), ‘Plato@50: innovations in hardware: mission based developments led other places,’ Computer History Museum (video). June 3, 2010. Philip McKinney (Moderator), Donald Bitzer, Roger L Johnson, and Larry Weber (speakers). Branstetter L. G. , Sakakibara M. ( 2002 ), ‘ When do research consortia work well and why? Evidence from Japanese panel data ,’ American Economic Review , 92 ( 1 ), 143 – 159 . Castellano J. A. ( 1992 ), Handbook of Display Technology . Academic Press : San Diego, CA . Google Preview WorldCat COPAC Castellano J. A. ( 2005 ), Liquid Gold: The Story of Liquid Crystal Displays and the Creation of an Industry . World Scientific : Singapore . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Cattani G. ( 2006 ), ‘ Technological pre-adaption, speciation, and emergence of new technologies: how Corning invented and developed fiber optics ,’ Industrial and Corporate Change , 15 ( 2 ), 285 – 318 . Google Scholar Crossref Search ADS WorldCat Christensen C. M. ( 2000 ), The Innovator's Dilemma. Harper : New York, NY . Google Preview WorldCat COPAC Cusumano M. A. , Mylonadis Y. , Rosenbloom R. S. ( 1992 ), ‘ Strategic maneuvering and mass-market dynamics: the triumph of VHS over Beta ,’ Business History Review , 66 ( 1 ), 51 – 94 . Google Scholar Crossref Search ADS WorldCat den Boer W. ( 2005 ), Active Matrix Liquid Crystal Displays: Fundamentals and Applications. Newness : Burlington, MA . Google Preview WorldCat COPAC Eggers J. P. ( 2014 ), ‘ Competing technologies and industry evolution: the benefits of making mistakes in the flat panel display industry ,’ Strategic Management Journal , 35 ( 2 ), 159 – 178 . Google Scholar Crossref Search ADS WorldCat Eisenhardt K. M. ( 1989 ), ‘ Building theories from case study research ,’ Academy of Management Review , 14 ( 4 ), 532 – 550 . Google Scholar Crossref Search ADS WorldCat Eisenhardt K. M. ( 2002 ), ‘Building theories for case study research,’ in Huberman A. M. , Miles M. B. (eds), The Qualitative Researcher's Companion . Sage : Thousand Oaks, CA , pp. 5 – 36 . Google Preview WorldCat COPAC Fuji-Chimera ( 1998–2007 ), Ekisho Kanrenshijo no Genjo to Shorai Tenbo . Fuji-Chimera Research Institute : Tokyo . WorldCat COPAC Geroski P. A. ( 2000 ), ‘ Models of technology diffusion ,’ Research Policy , 29 ( 4–5 ), 603 – 625 . Google Scholar Crossref Search ADS WorldCat Jenkins M. , Floyd S. ( 2001 ), ‘ Trajectories in the evolution of technology: a multi-level study of competition in formula 1 racing ,’ Organization Studies , 22 ( 6 ), 945 – 969 . Google Scholar Crossref Search ADS WorldCat Jojima A. ( 2006 ), Sonii Byo . Yosensha : Tokyo . Google Preview WorldCat COPAC Jovanovic B. , MacDonald G. M. ( 1994 ), ‘ Competitive diffusion,’ The Journal of Political Economy , 102 ( 1 ), 24 – 52 . Google Scholar Crossref Search ADS WorldCat Katz M. L. , Shapiro C. ( 1985 ), ‘ Network externalities, competition, and compatibility ,’ American Economic Review , 75 , 424 – 440 . WorldCat Katz M. L. , Shapiro C. ( 1986 ), ‘ Technology adoption in the presence of network externalities ,’ Journal of Political Economy , 94 ( 4 ), 822 – 841 . Google Scholar Crossref Search ADS WorldCat Katz M. , Shapiro C. ( 1992 ), ‘ Product introduction with network externalities ,’ Journal of Industrial Economics , 40 ( 1 ), 55 – 83 . Google Scholar Crossref Search ADS WorldCat Katz M. , Shapiro C. ( 1994 ), ‘ Systems competition and network effects ,’ Journal of Economic Perspectives , 8 ( 2 ), 93 – 115 . Google Scholar Crossref Search ADS WorldCat Katzmaier D. ( 2011 ), ‘Will any 2011 TV best the 2008 Pioneer Kuro?,’ Cnet.com, February 22, 2011. http://news.cnet.com/8301-17938_105-20034816-1.html. KSR Ltd. ( 2003 ), Flat Panel Display. Patenttec : Tokyo . WorldCat COPAC Lawrence B. S. ( 1984 ), ‘ Historical perspective: using the past to study the present ,’ Academy of Management Review , 9 ( 2 ), 307 – 312 . Google Scholar Crossref Search ADS WorldCat Levinthal D. A. ( 1998 ), ‘ The slow pace of rapid technological change: gradualism and punctuation in technological change ,’ Industrial and Corporate Change , 7 ( 2 ), 217 – 247 . Google Scholar Crossref Search ADS WorldCat Mathews J. ( 2005 ), ‘ Strategy and the crystal cycle ,’ California Management Review , 47 ( 2 ), 6 – 32 . Google Scholar Crossref Search ADS WorldCat Menzawa J. , Ohnishi K. , Takahashi J. ( 2008 ), Toshiba. Shuppan Bunka Sha : Tokyo . Google Preview WorldCat COPAC Murtha T. , Lenway S. , Hart J. ( 2001 ), Managing New Industry Creation . Stanford University Press : Stanford, CA . Google Preview WorldCat COPAC Nihon Hoso Kyoku ( 2003 ), Project X Challengers: Purazuma Terebi Ai no Moji Kara Hajimatta. TV Documentary . Japan . WorldCat COPAC Nikkei Business Publications ( 1998 ), Flat panel display. Nikkei Business Publications . Tokyo . WorldCat COPAC Nikkei Business Publications ( 1999 ), Flat panel display. Nikkei Business Publications . Tokyo . WorldCat COPAC Nikkei Business Publications ( 2002 ), Flat panel display. Nikkei Business Publications . Tokyo . WorldCat COPAC Nikkei Business Publications ( 2005 ), Flat panel display. Nikkei Business Publications . Tokyo . WorldCat COPAC Numagami T. ( 1999 ), Ekisho Disupurei no Gijutsu Kakushinshi. Hakuto-Shobo : Tokyo . Google Preview WorldCat COPAC Pioneer C. ( 2008 ), ‘Press release: Paionia to Matsushita ga PDP jigyo no hokatsu teikei nit suite kihon doi,’ April 24, 2008. http://pioneer.jp/press/2008/0424-1.html. Rumelt R. P. ( 1974 ), Strategy, structure, and economic performance. Graduate School of Business Administration, Harvard University . Google Preview WorldCat COPAC Rycroft R. W. , Kash D. E. ( 2002 ), ‘ Path dependence in the innovation of complex technologies ,’ Technology Analysis and Strategic Management , 14 ( 1 ), 21 – 35 . Google Scholar Crossref Search ADS WorldCat Sangyo Times ( 1990, 1992–2007 ), Ekisho meka Keikaku Soran. Sangyo Times : Tokyo . WorldCat COPAC Sankei Shimbun ( 1998 ), “Shapu buraun-kan TV zenpai he, 7 nengo madeni subete ekishoka,” Sankei Shimbun, August 19, 1998. Stolpe M. ( 2002 ), ‘ Determinants of knowledge diffusion as evidenced in patent data: the case of liquid crystal display technology ,’ Research Policy , 31 ( 7 ), 1181 – 1198 . Google Scholar Crossref Search ADS WorldCat Strauss A. , Corbin J. ( 1998 ), ‘Grounded theory methodology,’ in Denzin N. K. , Lincoln Y. S. (eds), Strategies of Qualitative Inquiry . Sage : Thousand Oaks, CA , pp. 158 – 183 . Google Preview WorldCat COPAC The Mainichi Newspaper ( 2009 ), ‘Paionia, TV kanzen tettai shomosen sake, ikinokori hakaru,’ February 13, 2009. Retrieved from the Factiva Database August 12, 2013. Weber L. , Kurashige H. , Wani K. ( 2008 ), Shogeki! Purazuma Terebi ha Shakai wo Kaeru: Nihongata Kigyoka Supiritto Yume no Dai-hitto Shohinn wo Umu. Jitsugyo no Nihon Sha : Tokyo . Google Preview WorldCat COPAC Yin R. K. ( 2014 ), Case Study Research: Design and Methods , 5th edn. Sage : Thousand Oaks, CA . Google Preview WorldCat COPAC Zook C. , Allen J. ( 2003 ), ‘ Growth outside the core ,’ Harvard Business Review , 81 ( 12 ), 66 – 75 . Google Scholar PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press on behalf of Associazione ICC. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Market adjacency and competing technologies: evidence from the flat panel display industry
JF - Industrial and Corporate Change
DO - 10.1093/icc/dtz009
DA - 2019-12-01
UR - https://www.deepdyve.com/lp/oxford-university-press/market-adjacency-and-competing-technologies-evidence-from-the-flat-1OabXP3qJG
SP - 1429
VL - 28
IS - 6
DP - DeepDyve
ER -