Innovation Strategies Of Asian Firms In The United States

J. Eng. Technol. Manage. 22 (2005) 255–273 www.elsevier.com/locate/jengtecman

Innovation strategies of Asian firms in the United States Jessie P.H. Poon *, Alan MacPherson Canada-US Trade Center, 105 Wilkeson Quad, University at Buffalo-SUNY, Buffalo, NY 14261, USA Available online 9 November 2005

Abstract This paper examines the relationship between Asian firms’ technological and non-technological strategies and innovation capability. Particular attention is focused on subsidiaries in the United States (US) with headquarter units in South Korea, Singapore, and Taiwan. Empirical evidence from a sample of 151 subsidiary plants and establishments suggests that Asian firms invest in the US to upgrade their their knowledge base with a view to supporting new product and market-based innovations. The results of an ordered probit regression model of innovation performance suggests that new product development and marketing capability make a significant contribution to increased US patents among Asian firms while applied research is only marginally significant in explaining firms’ innovation capability. The major sources of innovation capability are revolved around a tacit understanding of technology and products than more explicit forms of knowledge. Our empirical findings also suggest that stronger business performance is associated with new product development and marketing capability. # 2005 Elsevier B.V. All rights reserved. JEL classification: 031; 033; N65; L6 Keywords: Asian latecomers; Technology sourcing; Applied research; Marketing capability

1. Introduction A considerable body of research exists that attempts to explain the remarkable speed and level of technology development among firms from industrializing countries in Asia (Hobday, 1995; Mathews and Cho, 2000; Choung et al., 2000). Until the 1980s, the technological base of Asian firms from South Korea, Taiwan and Singapore has largely depended on transferred or borrowed

* Corresponding author. Tel.: +1 716 645 2722; fax: +1 716 645 2329. E-mail addresses: [email protected] (Jessie P.H. Poon), [email protected] (A. MacPherson). 0923-4748/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jengtecman.2005.09.001

256

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

technology from foreign firms. More recently, however, there has been a shift in strategy among Asian firms toward a more knowledge resource-based model that allows them to acquire and apply knowledge and technology for advantageous rent-generation. This paper examines the innovation strategies driving Asian firms’ investment in the United States (US)-based on the resource-based model of knowledge. Specifically, we show that Asian firms’ competitive advantage revolves around their ability to capture rents through a series of technological and marketing capabilities that collectively constitute the firms’ knowledge resource base. Because Asian firms are late adopters of technology internationally, the sample of firms examined here tends to undertake incremental to moderate innovations than radical innovations. Furthermore, the firms’ incremental to moderate innovations are largely centered on market opportunities that allow them to differentiate their products and applications from their competitors. Bearing this in mind, the major research questions undertaken here will focus on identifying Asian firms’ strategic capabilities in innovation, the sources of their innovation capability, and, the relationship between the firms’ innovation capability and business performance. In the next section, we first discuss the knowledge resource-based theory and its relevance for conceptualizing the innovation strategies of Asian firms in the US. This theory is particularly pertinent since it directly links innovation capability to firm performance and thereby sustained competitive advantage. Next, innovation capability is examined in terms of the sources of knowledge acquisition. A series of hypotheses are built around the strategies as well as sources that shape innovation capability. This is followed by a series of models that are constructed to test these hypotheses including ordered probit and loglinear models. The results and findings are elaborated next, and the paper is concluded with a discussion of research implications. 2. Literature review and research hypotheses 2.1. Knowledge-based resource theory The knowledge resource-based theory of the firm (Grant, 1996; Spender, 1996) has its inception in the resource-based theory of the firm, the latter of which posits that a firm’s competitive advantage depends on its ability to accumulate and deploy assets (tangible and nontangible) such that the resource yields high returns over long periods of time (Wernerfelt, 1984; Barney, 1991). The knowledge resource-based view is distinguished from the resource-based view through its emphasis on knowledge as a strategic factor of production over more traditional factors, such as land, labor or capital. A fundamental tenet of the theory is that firms are inherently heterogenous because they possess distinctive knowledge competencies. Through resource selection and accumulation, firms exploit market imperfections in knowledge, thereby creating asymmetry in information, which in turn creates knowledge-based transaction costs. Firms that possess the capability to create, supply and commercialize knowledge resources face lower transaction costs while building barriers to the acquisition or imitation of key knowledge inputs at the same time. Firm heterogeneity arises from the contextually sticky nature of knowledge which enhances firm-specific advantages (Dro¨ge et al., 2003). By focusing on the nature and characteristics of knowledge resources which underscore firm heterogeneity, the theory directly links a firm’s strategy to its performance. The usefulness of applying the knowledge resource-based theory to Asian firms’ innovation strategies in the US may be described as follows. Since knowledge resource heterogeneity is central to competitive advantage, a firm that builds its strategy around a resource or capability not

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

257

possessed currently by other firms (e.g. competitors) is more likely to create sustained competitive advantage. Asian firms are late internationalizers and possess considerable technological deficits compared to their western or even Japanese counterparts. To correct this deficit, accumulating innovation capabilities that contributes to distinctive knowledge competencies involves focusing on a specific set of strategies that reflect and add to their knowledge resource strength. These strategies are centered on incremental to moderate than radical innovations, and with applied research than basic scientific research. Such a strategy is distinguished from that of early internationalizers (e.g. European firms) in say the biotechnology industry. Here, Shan and Song (1997) have shown that inward foreign direct investment in the US is associated with the unequivocal leadership of the US in biotechnology innovations and a major motivation for location of foreign facilities in the US lies in the sourcing of scientific knowledge in the industry. Our aforementioned proposition follows from the nature of knowledge itself. In Nightingale’s (1998), cognitive model of knowledge, technology is seen as an artificial function which involves adapting means to reach a preconceived end. For example developing wireless technology in cellular phones to enable conversations between two spatially separated and mobile individuals. Applied research is usually tied to solving technological problems and opportunities and revolves around a design or blueprint and that carry lower levels of uncertainty. On the other hand, basic science and research is distinguished from applied research in that there is a level of abstraction involved that is not readily compressed into information and solutions. While the goal of scientific research is to establish cause and effect, to explain, determine and predict, it is not always possible for applied research on technology to extrapolate scientific patterns into the future. Nightingale (1998) observes that this is due to the fact that scientific knowledge and therefore basic scientific research involves forming an abstract correspondence between tacit understanding, patterns in nature and mathematical patterns. Hence, while applied knowledge and technology might make it is easier to develop a fuel-efficient car, it is more difficult to disentangle and acquire knowledge on fuel cell research. Furthermore, basic research is also more uncertain and demands longer development times with no guarantee of market returns. Given the recency of Asian firms’ entry to the industry of innovations, and their lack of a first mover advantage in basic scientific research compared to European, American or even Japanese firms, they are more likely to develop knowledge resources based on the ‘know-what’ and ‘know-how’ than the ‘know-why’ of technology. The importance of applied research and knowledge, not just the acquisition of knowledge further implies that innovations are largely directed towards the creation of market value in the marketplace (Demarest, 1997; Claycomb et al., 2002). That is to say, knowledge gains value only when it may be deployed to exploit market opportunities. This leads to the following hypothesis: H1a. Applied research is positively related to Asian firms’ innovation capability. As late entrants to innovations, Asian firms are able to equip themselves with the latest technical knowledge competence and to set new industry standards and performance. Such ‘‘free rider’’ advantages allow firms to benefit from information spillovers as a consequence of locating and operating in technology-rich environments such as those found in the U.S, while avoiding high R&D costs as well as the costs of educating customers about the product (Cho et al., 1998). However, positive imitation requires some level of technology capacity. Most successful imitations are characterized by relatively high levels of technological competence

258

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

that potentially turn the firm into an innovator (Schewe, 1998), or to use Rothwell and Gardiner’s (1988) description, ‘‘reinnovators’’. As a reinnovator, the imitator needs to transform the firm’s technology and production capacity into products that may be distinguished from its competitors whether in terms of cost, design or performance. Imitatorsturned-innovators do not simply reconstruct a technology or product but improve upon, modify or transform existing ones. To do so, the innovating firm must also develop considerable marketing capability that identifies market niches or new markets. Hence, innovations involve not only technological but non-technological activities such as marketing capability as well. The gathering of market information includes not only matching the product to customer or user needs, but also sourcing for relevant technologies, and, differentiating its technology or product from its competitors. Salomo et al. (2003) have pointed out that market, as opposed to customer, orientation must include this latter element of competition. In other words, learning from the market is a chief source of innovations and the knowledge resource-based theory suggests that marketing capability contributes to a firm’s knowledge capability by ensuring a high probability of commercial success of the products (Weerawardena, 2003). Hence we hypothesize that marketing capability will also be positively related to Asian firms’ innovation capability. H1b. Marketing capability is positively related to Asian firms’ innovation capability. As reinnovators, we expect that Asian firms are largely engaged in incremental to moderate innovations and product development than radical innovations. Garcia and Calantone (2002) have provided useful definitions of radical, moderate and incremental innovations. An incremental innovation is typically associated with more mature markets and with improving an existing product or design, or, in extending a product’s life. Any discontinuity introduced is often confined to the customer or firm level. Moderate innovations, on the other hand, will result in market or technological discontinuities at the level of the industry or world market. Finally, radical innovations tend to introduce products or technologies that cause both market and technological discontinuities at the industry and world market levels. Using their definition, Asian firms are rarely involved in radical innovations as might be illustrated by the impact of say the worldwide web on world markets and industrial transformation. Rather, their level of innovativeness is confined to more moderate or incremental forms.1 Engaging in incremental and moderate innovations reinforce previous arguments that Asian firms are more focused on market opportunities than in basic research generation. As will be seen below, this has resulted in a higher number of Asian firms reporting relatively successful introductions of new products to the US market than US patents. The difference between the two lies in the degree of technological input to the development of new products with firms reporting longer development times, greater uncertainties associated with commercializing patents, and higher R&D intensities in securing US patents. Securing US patents also requires spending substantial resources in preempting competitors from patenting the product first. Indeed, one difficulty confronting firms that engage in research to secure US patents is that they can find

1 For example, technology on scanners first originated from the Japanese, but it was the Taiwanese firms that improved upon the technology and then successfully commercialized and dominated the scanner industry in the 1990s (pesonal interview with a Taiwanese firm that manufactures scanners, 2003).

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

259

themselves in a situation where another firm has patented the product first. This leads to the final sub-hypothesis, that is: H1c. New product development is positively related to Asian firms’ innovation capability. The knowledge resource-based theory directly links a firm’s resource capability to its performance, hence specific knowledge forms and thereby innovative capability are expected to have a positive effect on firm performance. This positive relationship stems from the reasoning that knowledge is and innovations are a strategic resource that enables the firm to establish a position of competitive advantage through superior returns. However, empirical evidence on the aforementioned relationship is mixed with Teece (1996) finding no relationship between innovations and market share, and Dro¨ge et al. (2003) finding a positive relationship between applied knowledge and financial performance. Despite the mixed evidence but consistent with the knowledge resource-based theory, we expect that knowledge resource selection among Asian firms through strategies of applied research, marketing capability and new product development are likely to have a positive effect on firm performance. H2a. Applied research is positively related to firm performance. H2b. Marketing capability is positively related to firm performance. H2c. New product development is positively related to firm performance. 2.2. Sources of knowledge Since the seminal works of Polanyi (1967) and Nelson and Winter (1982) suggesting that knowledge may be tacit or more explicit and codified, the nature of knowledge has been extensively analyzed and its complexity documented (Spender, 1996; Grant, 1996; Kogut and Zander, 1993; Nonaka, 1991; Teece, 1977). Despite the increasingly sophisticated conceptualization of knowledge forms, we have found the original formulation of tacit and codified knowledge useful for operationalizing our research framework. While it has become accepted that tacit and codified knowledge are not mutually exclusive but may be expressed along a continuum, the cognitive dimension of tacit knowledge as elaborated by Nightingale (1998) indicates that the ‘meaning’ of scientific and technical information can only be understood by a group of practitioners and researchers who are well-versed in the subject or familiar with the language that describes the information. For this reason, tacit knowledge does not only consist of embodied knowledge but is also embedded in social networks. Social knowledge contains higher tacit content because its major mechanism of transfer is rooted in individuals, groups or organizations. Geographical proximity and interpersonal interaction help facilitate its transmission (Edmonson et al., 2003). On the other hand, explicit knowledge is more codifiable in the form of manuals (e.g. suppliers’ manuals), publications (e.g. scientific or technical journals) or hardware. This is because codified forms of knowledge are organized around procedures, properties, facts or axiomatic propositions, and are more readily transferable, taught and imitated. The nature and makeup of innovation-based and scientific knowledge indicates that the major sources of tacit knowledge for Asian firms in the US are likely to be harnessed from skilled or knowledgeable individuals, such as personnel involved in R&D, engineering consultants, and customers or users who have considerable knowledge accumulated from their technological,

260

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

product or/and market experiences. On the other hand, the principle sources of codified knowledge may be found in published sources or reverse engineering of existing products and prototypes. There is reason to expect that Asian firms are more inclined to acquire tacit knowledge than codified knowledge that allow them to augment their technological capabilities as this potentially enables them to realize reinnovations through the internalization of more complex technological knowledge. H3a. Tacit knowledge has a more positive effect on applied research than codified knowledge. H3b. Tacit knowledge has a more positive effect on marketing capability than codified knowledge. H3c. Tacit knowledge has a more positive effect on new product development than codified knowledge. Finally, the sources of knowledge and thereby innovation capability are examined in terms of the age of the firms. Edmonson et al. (2003) have shown that late adopters to the same technology that relies on codified knowledge tend to perform better because they are able to learn from early adopters. Extant literature and earlier discussion indicate that new product development is intimately related to market knowledge competence (Li and Cavusgil, 1998). Older firms and early entrants to the US have a market lead advantage because they have time to establish product legitimacy among their customers and to gather market information about their products. Knowledge transfers between producers and customers are relatively well-established in the literature (Wikstrom, 1996; Cummings and Teng, 2003; Ritter and Walter, 2003; Salomo et al., 2003). Older firms are expected to use their embedded customer networks to create information asymmetries that allow them to exploit first mover advantages or to erect barriers against their competitors. Younger firms, on the other hand, suffer from the liability of newness (Stinchcombe, 1965) and are more inclined to resort to more imitative and codified forms of knowledge than tacit knowledge for new product development. The effect of age on sources of knowledge acquisition is examined in the context of the following hypotheses. H4a. Tacit knowledge has a positive effect on older firms’ innovation capability. H4b. Codified knowledge has a positive effect on younger firms’ innovation capability. The research framework presented above is summarized in Fig. 1. 3. Research design, constructs, and model 3.1. Telephone survey To gather relevant data for this study, we surveyed manufacturing firms in the US with national origins in Taiwan, South Korea and Singapore. To identify the population, multiple sources were consulted: 1. Singapore: Databases from the Agency for Science, Technology and Research (formerly, National Science and Technology Board), IE Enterprise (formerly Trade and Development Board), and newspaper archives from the country’s leading newspapers, the Business Times and the Straits Times.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

261

Fig. 1. Research framework. Note: See Appendix A for details of the variables.

2. South Korea: Firm directories from the Korean Chamber of Commerce and Industry in the US and the Korean Bank of Investment. 3. Taiwan: Firm directories from the Taipei Economic and Cultural Office in New York which lists Taiwanese companies in the US, and Taiwan’s Ministry of Commerce and Economics. The above was supplemented with Chinese Business in America (2002–2003) published by Caravel Inc. This directory, however, includes firms from PRC China as well as Asian American companies. The websites of all Chinese firms were examined to identify only those with national origins from Singapore and Taiwan. Besides investigating each firm’s website to determine the firm’s national origin and suitability for our study, telephone calls were also made to ensure consistency in the population. Other firms that did not fit included parent companies in the US that were established by Asians who may or may not have investment in the three countries. This group of businesses represents startups or investment from individuals from the three countries who have either completed their education in the US and/or worked in US companies for a few years before establishing their own companies. However, not all US headquartered Asian companies were excluded from the database. The headquarters (HQ) of Formosa Plastics, for instance, is in the US. But Formosa Plastics is clearly a Taiwanese company with national origin in Taiwan. It relocated its HQ from Taiwan to coordinate its six regional centers in the US. Finally, Asian subsidiaries in the US that are purely sales units were also excluded from the study. From company websites and telephone clarifications, it was possible to identify target populations for each of the three countries, that is, 210 for Taiwan, 56 for Singapore and 113 for South Korea. In the second stage of the fieldwork, pilot surveys were conducted with 15 firms in order to fine-tune the questionnaire. After refining the questionnaire, it was further determined that telephone surveys would communicate the questions better than a mail survey. This was because of language problems associated with Korean managers in particular. Despite careful translations of the survey instrument into Chinese and Korean, pre-testing yielded better and more usable

262

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

survey results with personal communication. All 383 firms were contacted and a total of 151 responses were collected over 3 months from February to April of 2003 resulting in response rates of 35, 44 and 48% for Taiwan, South Korea and Singapore, respectively. Those who did not respond included individuals who refused to participate in the survey citing company policies, or, who were out of the country despite three or more attempts to contact them over the 3 months. In the final stage of the fieldwork, post survey interviews were conducted with eight firms to clarify the results and responses. Three-quarters of the firms are in three major industries, namely computers, semiconductors and electronics. The remaining firms are in chemicals, transportation, telecommunications/IT, plastics and medical instruments or pharmaceuticals. Firms in low technology sectors such as textiles were excluded from the survey. The distribution of firm sizes in terms of worldwide sales is presented in Table 1. The table shows that Korean companies are generally larger, with 41.1% reporting sales of over US$ 1 billion. This is because the Korean industrial structure tends to be dominated by large business conglomerates (chaebols) which are diversified but integrated corporations that are involved in a number of product lines (Jung and Kwon, 2002). Singaporean and Taiwanese firms tend to be smaller with the majority of firms reporting worldwide sales of less than US$ 50 million. While large government-linked corporations also dominate in the Singapore economy, it is the IT sector and thereby smaller firms that have led the country’s internationalization process in the US, supported by government-based institutional infrastructure such as the Agency for Science, Technology and Research. 3.2. Dependent and independent variables The major dependent variable in this study pertains to innovation capability. Two measures are used to evaluate this namely, the ability of firms to increase the number of US patents, and the introduction of new products to the US market since they established operations in the country. Between the two measures, the ability to introduce US patents is expected to correspond to greater firm innovativeness and development of knowledge resources since patent-making tends to reflect a solution to a scientific puzzle and involves inventions that may or may not turn into successful products in the market. Introducing new products to the US market also reflect a firm’s innovation capability, but because Asian firms rely on more imitative and incremental product innovations, new product introduction here is largely associated with the adaptation and enhancement of existing products, production or delivery Table 1 Distribution of firm size Size (worldwide sales US$)

Distribution (%) All

<50 million 50–250 million 251–500 million 501 million–1 billion >1 billion Total

Korea

Taiwan

59.0 21.3 3.3 4.1 12.3

26.5 20.6 5.9 5.9 41.1

84.2 5.3 5.3 5.2 0.0

68.1 26.1 1.5 2.9 1.4

100.0

100.0

100.0

100.0

Based on sample of 151. Source: authors’ survey.

Singapore

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

263

systems (Song and Montoya-Weiss, 1998). A likert scale of 1–7 was used to rank these measures with 1 being ‘‘did not increase at all’’ and 7 being ‘‘increased very significantly’’ (see Appendix A for a list of the variables). Independent variables, on the other hand, consist of two major sets of constructs. First, six items describing innovation strategies were identified including technological activities, such as basic science, applied research, product design and development, technology/engineering process development, quality control of materials, and non-technological activities like marketing capability. For each item, firms were asked to estimate whether the degree of technology development had been significant after establishing their operations in the US with 1 being ‘‘not significant at all’’ and 7 being ‘‘critically significant’’. The Cronbach a-coefficient is 0.77, indicating reasonable internal consistency (see Appendix B for a summary of inter-item correlations). A second set of construct deals with the sources that contribute to Asian firms’ capacity to develop technology after location in the US. These items are designed to elicit responses on tacit and codified knowledge. Tacit knowledge includes items that measure the share of firms’ R&D personnel as well as the importance of US-based engineering/consulting services, strategic alliances and customer relationships. Items measuring codified knowledge included blueprints, manuals, publications and imitation of US products and prototypes. The Cronbach a-coefficient is very close to the recommended threshold of 0.7 at 0.69. In examining the strategies and knowledge sources underscoring innovation capability, two variables were also introduced to control for national differences in firm size and industry effects (see previous section and Table 1). Firm performance is measured by the firm’s annual sales growth over the past 10 years. Firms that registered sales growth of 5% and below are classified as ‘‘low firm performance’’, while those above 5% as ‘‘high firm performance’’. This classification was established based on sales growth’s median and it is consistent with firms’ perception of ‘‘low’’ and ‘‘high’’ performance (pilot interviews, 2002). Finally, we also report the major reasons for Asian firms’ location in the US with each item assigned a scale of 1–7, 1 being ‘‘not unimportant at all’’ and 7 being ‘‘critically important’’. These items provide an important context to the hypotheses being tested. The eight locational items are listed in Appendix A. 3.3. Model All of the hypotheses outlined above except for H2a, H2b and H2c are tested using the ordered probit model (OPM). This is because the constructs and their measures comprise ordinal data. Ordered-response models are useful when the response variable is of an indexed nature, that is to say, the dependent categories increase sequentially (e.g. from not important to very important). Underlying the indexing is a latent but continuous descriptor of the response. The structural model takes the form of: y
i ¼ xi bi þ ei

(1)

where xi is the vector of explanatory variables, bi the column vector of parameters to be estimated with the first element being the intercept, y
i the latent variable and ei is the random error term which is assumed to follow a normal distribution. OPM is derived from a measurement model where the latent variable, which ranges from 1 to 1, is mapped to an observable variable y such that the extreme interval categories w0 = 1

264

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

and wj = 1. That is to say, y is related to y
i according to the following measurement model: 8 1 ðnot important at allÞ if ’0 ¼ 1  y
i < ’1 > > > >  > > > >  <  yi ¼ > >  > > > >  > > : 7 ðcritically importantÞ if ’6 ¼ 1  y
i < ’7 Given the increasing order of the categories, a positive sign for b is interpreted as having greater significance on innovation capability. Hypotheses H2a, H2b and H2c will be tested using a loglinear model. 4. Analysis and results Table 2 reports the factors that influence Asian firms’ decision to establish operations and facilities in the US. The table is based on the analysis of covariance (anacova) and controls for industry and firm size. The F-statistics are based on Type III errors and indicate that significant to marginal differences exist between the three countries for three factors, namely, strategic alliance and proximity to customers as well as to develop distribution channels, respectively. Specifically, the anacova reveals that smaller firms such as those from Singapore and Taiwan place a greater emphasis on strategic alliance and customer proximity more than larger firms from Korea. Taiwanese firms also rank the development of distribution channels much more highly (mean = 5.7) over both Korea and Singapore which have similar means of around 4.4. While Korean firms gave a lower score to proximity to customers, their mean score is nonetheless still quite high at 5.5 while the mean scores of the other two countries are over 6.0. Despite some national differences, the results indicate that factors that are related to market orientation and customers constitute the strongest reason for Asian investment in the US with overall mean for all three countries at 6.0. Firms assigned neutral to moderately important scores to three locational factors, namely, proximity to competitors, developing new products and prototypes, and improving product Table 2 Analysis of covariance: locational factors in the USa Variable

Country differences

F-value

Access skilled labor Collect market information Develop distribution channels Strategic alliance Be near competitors Be near customers Develop new products and prototypes

No difference No difference TW > SP, KR SP > TW > KR No difference SP, TW > KR No difference

1.00 0.66 2.52* 3.90** 0.38 3.20** 1.62

a

The results are controlled for firm size (total annual worldwide sales) and industrial types. F-values are based on Type III errors. * denotes 10% significance level. ** denotes 5% significance level.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

265

quality and performance. Developing new products and prototypes has the lowest overall mean with a nearly neutral score of 3.9. This is consistent with earlier observations of the paper that Asian firms are predominantly imitators-turned-reinnovators rather than radical innovators with firms attaching slightly greater importance to improvements in product quality, performance or even design. Asian firms are also interested in locating near their rivals or competitors in order to gather intelligence on their competitive strategies and to exploit potential information spillover benefits. Finally, it would seem that access to skilled labor in the US is relatively unimportant with firms reporting a low mean of 3.1. A major reason that was provided to us during post-survey interviews is that skilled personnel in the form of engineers are readily available back in the firms’ home countries where wages are also lower. In particular, Korean companies in the US tend to depend on skilled personnel from back home rather than workers from the US. This may be explained by the fact that Korean parent companies tend to exert a stronger control over their subsidiaries compared to the other two nationalities. What Table 2 indicates is that despite some differences, firms from the three countries are relatively similar in their main motivations to set up operations in the US with market-based opportunities being the most important. To test hypotheses H1a, H1b and H1c, we report the report the results of ordered probit regressions (Table 3). The likelihood ratio is highly significant at 1% which indicates a good model fit. Table 3 shows that innovation strategies based on new product development and marketing capability are positive at the 5% level but applied research is only positively significant at 10%. The results strongly support H1b and H1c, but it is less evident for H1a. Many firms in the post-survey interviews emphasize the importance of benchmarking their position against competitors in relevant market segments in the US. Typically this means that they have to improve their products constantly. This is particularly the case when the product or application is

Table 3 Ordered probit regression: types of innovation activities and innovation capability Variable

Parameter estimate b (dependent variable = US patents)

Sector Size Basic science Applied research New product development Technology/engineering process development Quality control of materials Marketing capability w1 w2 w3 w4 w5 w6 2 log L

0.006 0.186 0.082 0.151* 0.137** 0.020 0.058 0.159** 1.337*** 17.100*** 1.998*** 2.288*** 2.810*** 3.645*** 308.576

Likelihood ratio * ** ***

denotes 10% significance level. denotes 5% significance level. denotes 1% significance level.

25.183***

266

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

developed for a specialized market segment. The results also support existing evidence that market-derived innovations and R&D tend to go hand-in-hand together (Gupta et al., 1986; Moenaert et al., 1994). This market-pull approach (Weerawardena, 2003) suggests that firms are constantly scanning markets for opportunities to satisfy their customers’ needs. Other innovation activities such as technology and engineering process do not appear to have led to the appropriation of US patents. One reason for this may be due to transfer costs associated with process designs and process engineering. Teece (1977) has noted that while the transfer process of such technology tends to be less difficult, the absorption process, on the other hand typically involves the transfer of instruments or specifications that facilitate continuous technology flow. Absorbing the engineering process, however, requires considerable advisory resources and access to complementary assets to make a new process feasible. As expected, basic science and product design and development are also not significant. Table 3 supports hypotheses H1b and H1c that firms from industrializing countries that focus on new product development and marketing capability tend to reap better results. The data show that these types of activities have directly resulted in abilities to secure US patents. One Taiwanese company we spoke to confirmed that such a trajectory is consistent with the Japanese catch-up model, since most of Asian firms’ innovative activities are associated with more mature stages of the product cycle. This also suggests that market rather than technological competence is used to build barriers against competitors, or, to distinguish the innovation from other producers. To test H2a, H2b and H2c, that is the effect of the above three innovation strategies on firm performance, we use loglinear models. Loglinear models allow examination of the relationship without the need to specify a dependent variable. This is useful when the number of categories is small (two for firm performance, see Section 3.2) and when the directionality of influence is unclear as suggested by some authors (Schewe, 1996). This is because innovations in mature technology can also be driven by increased sales which are compatible with the stage of the product cycle. No distinction between dependent and independent variables is necessary in loglinear modeling, therefore the model only indicates association between variables. Table 4 compares firms’ performance to that of their sales growth, and, the importance of new product development, applied research and marketing capability in its R&D activities (independence model). An interaction term for each innovative type has also been added that examines the significance of the relationship between firm’s performance and the R&D activity type. If the interaction term is found to be significant, then hypotheses H2a, H2b and H2c may be supported. The results show that hypotheses H2b and H2c may be somewhat supported. However, this is not the case for applied research. That is to say, the interaction variable sales growth  applied research is not significant. While applied research may be desirable in helping the firms build knowledge resources, it does not necessarily result in favorable firm performance. Tables 5 and 6 report the results of the ordered probit regressions. The estimated regressions in Table 5 suggests that tacit knowledge in the form of R&D personnel is a significant source of knowledge for firms’ innovation activities in applied research and new product development while interactions with customers and the strengthening of such relationships locally help develop marketing capability. The experience constituted within R&D personnel is measured by the share of engineers and scientists in the company and many of these personnel possess contextspecific knowledge that is relevant to the firm’s product and technology. Firm size is also positive and significant for new product development. There is also evidence that tacit knowledge through external partners such as strategic alliance and engineering/consulting firms are significant sources of tacit knowledge. None of the variables measuring codified knowledge is significant. The results would seem to support H3a, H3b and H3c.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

267

Table 4 Loglinear model: innovation activities and firm performance Variables

x2

1. Sales growth and applied research Sector Sales growth Applied research Sales growth  applied research Sales growth  sector Wald statistic a

6.70*** 3.05 15.82*** 3.54 0.85 0.18

2. New product development (NPD) Sector Sales growth NPD Sales growth  NPD Sales growth  sector Wald statistic a

8.28*** 1.10 6.20** 5.26* 1.52 1.65

3. Sales growth and marketing capability (MC) Sector Sales growth MC Sales growth  MC Sales growth  sector Wald statistic a a * ** ***

14.25*** 2.34 28.00*** 5.12* 4.93* 2.03

Test for independence. denotes 10% significance level. denotes 5% significance level. denotes 1% significance level.

To examine the final hypotheses, H4a and H4b, we performed similar regressions on firms’ innovation capability with the above sources of knowledge according to their age. We measured innovation capability in terms of firms’ ability to introduce new products to the US market since they established operations in the country. Three out of four firms reported moderate to important success in this dimension compared to only one out of four firms in US patents. As observed earlier, new product development here is largely associated with enhancing product value which can be rapidly commercialized on the market (firm interviews, 2003). Table 6 indicates that older firms2 develop new product capabilities through deployment of their R&D personnel, and through the development of strong local relationships with customers. As firms grow older and become more embedded in the US, skilled individuals have more opportunities and time to acquire tacit knowledge. Transfer of tacit knowledge is also enhanced when customers are local or geographically proximate (see Table 2), and when they are not end-users but are other companies. Customers and skilled personnel constitute major sources of product and technological experiences. Customers are also the primary source of new knowledge since they can be involved in problem-solving and they can help make sense of knowledge (Dougherty et al., 2000).

2

Older firms are defined as Asian firms with operations in the US for more than 5 years.

268

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

Table 5 Ordered probit regressions: sources of innovation knowledge Variable

Parameter estimate (Dep. var. = AR)

Parameter estimate (Dep. var. = NPD)

Parameter estimate (Dep. var. = MC)

Sector Size R&D personnel US engineering/consulting service US blueprints/manuals/publications Follow US products/prototypes Strategic alliance Local customer relationship w1 w2 w3 w4 w5 w6 2 log L

0.050** 0.045 0.272*** 0.005 0.015 0.062 0.129** 0.089 0.944 1.444** 1.678** 1.824*** 2.443*** 3.332*** 279.170

0.0256 0.162** 0.202** 0.191*** 0.039 0.030 0.022 0.016 0.566*** 0.839*** 1.203*** 1.443** 1.922 2.745 343.496

0.029 0.015 0.095 0.194*** 0.890 0.096 0.030 0.210** 0.560 0.014 0.517 0.841 1.308** 2.117*** 343.641

22.790***

27.271***

18.056**

Likelihood ratio

AR, applied research; NPD, new product development; MC, marketing capability. ** denotes 5% significance level. *** denotes 1% significance level.

Table 6 Ordered probit regressions: firm age, new product development and sources of technological knowledgea Variable

Sector Size R&D personnel US engineering/consulting service US blueprints/manuals/publications Follow US products/prototypes Strategic alliance Local customer relationship w1 w2 w3 w4 w5 w6 2 log L Likelihood ratio

Parameter estimate, b Old firms

Young firms

0.001 0.031 0.200** 0.071 0.087 0.027 0.043 0.301** 0.686 1.255** 1.503** 1.875*** 2.451*** 2.985*** 274.005

0.020 0.380 0.451** 0.564*** 0.211 0.391 0.573** 0.070 2.917  1.726 0.347 1.738 3.211* 49.413

20.474***

21.119***

Firms that have only been operating in the US for less than 6 years are classified as young firms (n = 33). Note: w3 is not estimated because there is no response associated with this scale. a Dependent variable, new product development. * denotes 10% significance level. ** denotes 5% significance level. *** denotes 1% significance level.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

269

Customer inputs therefore turn a company’s technology into a commercially successful product. Similarly, and contrary to expectation, younger firms also rely on tacit knowledge to develop their innovation capability, using R&D personnel, engineering and consulting companies and strategic alliances. Codified forms of knowledge, it would seem, have no effect on innovation capability even for younger firms. There is, however, one notable difference between younger and older firms. The emphasis on customer relationships for older firms suggests a more market-focused learning strategy compared to younger firms. Clearly for young firms, one way of overcoming environmental uncertainties and the liability of newness is to externalize risks through collaborative relationships to maximize knowledge with partners who already possess the needed knowledge investment. A major motive for entering into strategic alliances among Asian firms is that of technology partnering and cooperation. Such alliances allow firms to reduce risks and uncertainty that are found in R&D research. As noted by Hagedoorn (1993), strategic technology partnering is much more significant in applied research than other types of research such as basic research because the latter is of less relevance to interfirm cooperation since it is more or less a public good. Moreover, basic research often forms the core of a firm’s technological assets, hence there is less incentive to collaborate. In contrast, older firms tend to tap into sources that are more conducive to the internalization of knowledge assets. This is consistent with the customer–supplier relationship literature that points to customers as a major source of generation and transfer of both incremental and radical innovations (Roy et al., 2004). As observed earlier knowledge is relatively sticky, hence strong local relationships increase the quantity and scope of interactions that, in turn, eases the transmission of knowledge. One Korean auto supplier for example reported a preference for locating the company’s operations and research facilities in the South because of the absence of unionized labor. But it finally settled in Detroit because this facilitates interactions with its major customer, General Motors (GM). Proximity to, and a strong relationship with GM enhances the firm’s tacit understanding of GM’s specifications that are related to new car models, providing it with a lead time to develop products for GM ahead of its competitors. 5. Conclusion, discussion, and research limitations To develop firm-specific technology, a firm first needs to accumulate some basic technological know-how. Asian firms attain this know-how capability initially by forging backward and forward linkages with the affiliates of foreign MNCs in Asia. The problem with relying on imported technology, however, is that Asian firms are unlikely to acquire more advanced forms of knowledge since the latter constitutes the principle ownership advantage of MNCs’ operations abroad. Over time then, technological catch-up and narrowing involve firm strategies that enable the firms to directly learn, build and indigenize technologies through foreign location in technology rich environments such as the US. Such relocation allows Asian firms not only to imitate and internalize technologies from leading US competitors and rivals, but also to directly engage in knowledge transfers with sophisticated buyers through co-production with customers. Indeed the more complex the knowledge, the more important is co-production between the firms and their customers. To engage in technological and non-technological innovations, however, a firm also needs a core of capabilities. These capabilities can be enabling or constraining depending on the skills

270

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

and knowledge of R&D personnel and the links between R&D, production and marketing. This implies that some activities will meet with more economic success than others. In studying industrializing Asian firms in the US, this paper identifies new product development and marketing capability as the most promising innovation activities for appropriation of US patents, but not basic science, engineering process technology or design capabilities. As pointed out by Nelson (1991), firms need to learn to get good at certain types of innovation, and, at the sources that allow them to take advantage of, rather than a more ad-hoc effort in R&D research. The results in this paper indicate that Asian firms focus on an innovation strategy of new product development and marketing capability for two major reasons. First, innovations tend to be associated with more mature stages of the product life cycle, hence incremental to moderate innovations are necessarily tied to market competency. A firm’s innovativeness is correlated to its ability not only to identify applications and products, but also to commercialize the applications and products. Hence, fewer Asian firms in this research reported successfully introducing US patents than new products on the market. Through complementary linkages between marketing capability and product development, Asian firms have been able to secure a higher level of economic success with firms experiencing higher sales growth for those engaged in these strategies. Second, applied knowledge and R&D tend to be poorly diffused reflecting the relative stickiness of firm-specific knowledge in this sort of innovation activity. Because most Asian firms are reinnovators, firm-specific R&D tends to be more difficult to imitate particularly if their rivals have little access to these assets. The relative difficulty of appropriability implies that Asian firms that rely on codified or more explicit or even imitative forms of knowledge resources are less likely to build sustained innovation capability. Clearly the latter strategy has enabled Asian firms in the past to acquire considerable technological know-what, but increasingly, knowledge knowhow contains a learning dimension that requires a high tacit understanding. This is true for both recent (young firms) and late (old firms) innovators, and it is also true for the three types of innovation activities that underlie the firms’ knowledge base, that is, new product development, applied research and marketing capability. Knowledge acquisition among Asian firms indicate that tacit knowledge transfers through deployment of skilled personnel in R&D and close interactions with customers to identify and generate relevant understanding of applications of new technology, remain principle sources of developing innovation capability. These are supplemented by external sources such as strategic alliances with firms that possess relevant technological assets or through consultation with engineering service specialists. Finally some residual issues remain. First, the survey could be strengthened with more reverse-coding format. Second, other non-technological innovations need to be explored besides marketing capability such as distribution channels. Indeed post-survey interviews indicated that distributors can also be a major source of innovations because they work closely with customers. Finally, it is unclear if the current innovation strategies of Asian firms are sustainable in the long run for building knowledge competencies in the absence of parallel developments in more fundamental sources of knowledge generation, specifically basic scientific research. Acknowledgements Support for this research was provided by the National Science Foundation (award number BCS-0137045). We thank the editor and four reviewers for their constructive comments which greatly improved this paper.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

271

Appendix A Construct

Measure

1. Dependent variable 1.1 Innovation capability: since investing or operating in the US, has the number of (a) new products and (b) US patents increased for the parent and US companies? (1, did not increase at all; 7, increased very significantly) 1.2 Firm performance

Introduction of US patents, introduction of new products to US

Average annual growth in sales

2. Independent variable 2.1 Innovation strategies: please estimate if the degree of technology development has been significant in the following categories after establishing your company in the US (1, not significant at all; 7, critically significant) 2.2 Sources of technological knowledge: how important are the following factors in contributing to your US company’s capacity to develop technology? (1, not important at all; 7, critically important) (a) Tacit knowledge

Basic science, applied research, product design/development, technology/engineering process development, quality control of materials, marketing capability

Share of R&D personnel (%), US engineering/ consulting services, strategic alliances, local relationships with customers US blueprints/manuals/publications, follow US products/prototypes

(b) Codified knowledge 3. Locational factors How important are the following reasons for locating your company and operations in the US? (1, not important at all; 7, critically important)

Access skilled labor, collect market information, develop distribution channels, strategic alliance, be near competitors, be near customers, develop new products and prototypes

Appendix B. Correlations (a) Innovation strategies and activities

BS AR NPD TPD QCM

BS

AR

NPD

TPD

QCM

MC

–

0.73824 –

0.41709 0.46855 –

0.47990 0.57434 0.53107 –

0.33588 0.35520 0.41971 0.56037 –

0.00945 0.00331 0.13537 0.16664 0.19332

(b) Sources of Innovation

EC BM FPP SI

EC

BM

FPP

SI

CUST

–

0.15541 –

0.34512 0.45929 –

0.51629 0.30225 0.30650 –

0.08767 0.31619 0.34324 0.23137

272

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

Appendix B. (Continued ) (c) Locational factors

LABOR MI DISTRIB COMPET CUST

LABOR

MI

DISTRIB

COMPET

CUST

NPP

–

0.10319 –

0.09925 0.42503 –

0.20189 0.46958 0.30721 –

0.01575 0.25121 0.34597 0.37106 –

0.19759 0.20938 0.19808 0.23808 0.26377

BS, basic scientific research; AR, applied research; NPD, new product development; TPD, technology process development; QCM, quality control of materials/product; MC, marketing capability and research; EC, engineering/ consulting services in US; BM, blueprints/manuals/publications from organizations in US; FPP, following product pototypes or engineering systems in US; SI, strategic alliance in the US; CUST, strong local relationships with customers or firms in US; LABOR, to take advantage of skilled labor; MI, to collect market information and intelligence; DISTRIB, to develop distribution channels; COMPET, to be near competitors; CUST, to be near customers; NPP, to develop new products and prototypes.

References Barney, J., 1991. Firm resources and sustained competitive advantage. Journal of Management 17 (1), 99–120. Chinese Business in America, 2002–2003, Caravel Inc. Cho, D.S., Kim, D.J., Rhee, D.K., 1998. Latecomer strategies: evidence from the semiconductor industry in Japan and Korea. Organization Science 9 (4), 489–505. Choung, J., Hwang, H., Rim, M., 2000. Transition of latecomer firms from technology users to technology generators: Korean semiconductor firms. World Development 28 (5), 969–982. Claycomb, C., Dro¨ge, C., Germain, R., 2002. Applied product quality knowledge and performance: moderating effects of uncertainty. International Journal of Quality and Reliability Management 19 (6), 649–671. Cummings, J.L., Teng, B.S., 2003. Transferring R&D knowledge: the key factors affecting knowledge transfer success. Journal of Engineering and Technology Management 20 (1–2), 39–68. Demarest, M., 1997. Understanding knowledge management. Long Range Planning 30 (3), 37–52. Dougherty, D., Borrelli, L., Munir, K., O’Sullivan, A., 2000. Systems of organizational sensemaking for sustained product innovation. Journal of Engineering and Technology Management 17 (3–4), 321–355. Dro¨ge, C., Claycomb, C., Germain, R., 2003. Does knowledge mediate the effect of context on performance? Some initial evidence. Decision Science 34 (3), 541–568. Edmonson, A., Winslow, A.B., Bohmer, R.M.J., Pisano, G.P., 2003. Learning how and learning what: effects of tacit and codified knowledge on performance improvement following technology adoption. Decision Science 34 (2), 197–223. Garcia, R., Calantone, R., 2002. A critical look at technological innovation typology and innovativeness terminology: a literature review. Product Innovation Management 19 (2), 110–132. Grant, R.M., 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal 17 (4), 109–122. Gupta, A.K., Raj, S.P., Wilemon, D., 1986. A model for studying R&D-marketing interface in the new product development process. Journal of Marketing 50 (April), 7–17. Hagedoorn, J., 1993. Understanding the rationale of strategic technology partnering: interorganizational modes of cooperation and sectoral differences. Strategic Management Journal 14 (5), 371–385. Hobday, M., 1995. East Asian latecomer firms: learning the technology of electronics. World Development 23 (7), 1171– 1193. Jung, K., Kwon, S.Y., 2002. Ownership structure and earnings informativeness: evidence from Korea. The International Journal of Accounting 37 (3), 301025. Kogut, B., Zander, I., 1993. Knowledge of the firm and the evolutionary theory of the multinational corporation. Journal of International Business Studies 24 (4), 625–645. Li, T., Cavusgil, S.T., 1998. Decomposing the effects of market knowledge competence in new product export: a dimensionality analysis. European Journal of Marketing 34 (1–2), 57–79. Mathews, J.A., Cho, D.S., 2000. Tiger Technology: The Creation of a Semiconductor Industry in East Asia. Cambridge University Press, Cambridge, 389 pp.

J.P.H. Poon, A. MacPherson / J. Eng. Technol. Manage. 22 (2005) 255–273

273

Moenaert, R.K., Souder, W.E., DeMeyer, A., Deschoolmeester, D., 1994. R&D-marketing integration mechanisms, communication flows, and innovativeness. Journal of Product Innovation Management 11 (1), 31–45. Nelson, R.R., 1991. Why do firms differ, and how does it matter? Strategic Management Journal 12 (2), 61–74. Nelson, R.R., Winter, S.G., 1982. An Evolutionary Theory of Economic Change. Harvard University Press, Cambridge, Mass, 454 pp. Nightingale, P., 1998. A cognitive model of innovation. Research Policy 27 (7), 689–709. Nonaka, I., 1991. The knowledge-creating company. Harvard Business Review 69 (November–December), 96–104. Polanyi, M., 1967. The Tacit Dimension. Anchor Books, New York, 108pp. Ritter, T., Walter, A., 2003. Relationship-specific antecedents of customer involvement in new product development. International Journal of Technology Management 26 (5–6), 482–501. Rothwell, R., Gardiner, P., 1988. Reinnovation and robust designs: producer and user benefits. Journal of Marketing Management 3 (3), 372–387. Roy, S., Sivakumar, K., Wilkinson, I.F., 2004. Innovation-generation in supply chain relationships: a conceptual model and research propositions. Journal of the Academy of Marketing Science 32 (1), 67–79. Salomo, S., Steinhoff, F., Trommsdorff, V., 2003. Customer orientation in innovation projects and new product development success: the moderating effect of product innovativeness. International Journal of Technology Management 26 (5–6), 442–463. Schewe, G., 1996. Imitation as a strategic option for external acquisition of technology. Journal of Engineering and Technology Management 13 (1), 55–82. Shan, W., Song, J., 1997. Foreign direct investment and the sourcing of technological advantage: evidence from the biotechnology industry. Journal of International Business Studies 28 (2), 18–29. Song, M.S., Montoya-Weiss, M.M., 1998. Critical development activities for really new versus incremental products. Journal of Product Innovation Management 15 (2), 124–135. Spender, J.C., 1996. Making knowledge the basis of a dynamic theory of the firm. Strategic Management Journal 17 (4), 45–62. Stinchcombe, A.L., 1965. Social structures and organizations. In: March, J.G. (Ed.), Handbook of Organizations. Rand McNally, Chicago. Teece, D.J., 1977. Technology transfer by multinational firms: the resource cost of transferring technological know-how. The Economic Journal 87 (June), 242–261. Teece, D.J., 1996. Firm organization, industrial structure and technological innovation. Journal of Economic Behavior and Organization 31 (3), 193–224. Wernerfelt, B., 1984. A resource-view of the firm. Strategic Management Journal 5 (2), 171–180. Weerawardena, J., 2003. The role of marketing capability in innovation-based competitive strategy. Journal of Strategic Marketing 11 (1), 15–35.