Our paper proposes that corporate technological relatedness, or the degree to which business units within a corporation utilize similar technological knowledge, has both positive and negative effects on corporate R&D activities. On the one hand, business units that employ similar technological knowledge have better absorptive capacity to source knowledge from each other. On the other hand, a higher level of technological relatedness means that each business unit possesses fewer opportunities to gain new knowledge not known to other units, thus promoting path dependence to each other. Using a patent data analysis of 201 firms in R&D-intensive industries, we examine the effects of corporate technological relatedness on within-firm knowledge flow, boundary-spanning combinations of prior knowledge, and innovation impacts.
In strategic management theory, the resource-based view suggests that firms should avoid operating business portfolios that require vastly different kinds of resources and capabilities. According to this view, related diversifiers that exploit identical resources and capabilities across different businesses perform better than unrelated diversifiers unable to create synergistic effects. Thus, this view recommends that firms develop core competences to excel in a small number of aspects along value chain activities and join industries where they can exploit those competences. This theoretical argument has generated many empirical studies that vary with respect to the kinds of resources examined and the methodological approaches taken to measure relatedness among different business units.
Recent studies have suggested that achieving economies of scope based on technological knowledge is one of the most appropriate reasons for diversification (Miller, 2004 and Robins & Wiersema, 1995). The costs associated with additional application of technological knowledge within a corporate boundary are trivial. At the same time, transferring knowledge through a market mechanism is cumbersome, since devising contracts sophisticated enough to prevent opportunistic behavior by both knowledge seller and buyer is difficult. In line with this argument, recent empirical studies have demonstrated that firms creating synergy based on technological knowledge exhibit superior performance than those do not (Miller, 2006, Silverman, 1999 and Tanriverdi & Venkatraman, 2005). These studies imply that corporation with business units of similar technological expertise will perform better than those with business units that pursue different technological paths.
Although the similarity of technological expertise among a corporation's subunits may create synergy in R&D activities, excessive technological overlap may undermine a firm's R&D performance. Within the context of overseas R&D networks of multinational corporations, Song and Shin (2008) argued that R&D units have little incentive to source knowledge from each other when they have similar technological capabilities, thereby learning little from each other. On the other hand, innovations generated by combining diverse technological expertise tend to have a wider impact on ensuing technological evolution, than do those originating from a narrow range of technological expertise (Rosenkopf & Nerkar, 2001 and Yayavaram & Ahuja, 2008). These studies imply that corporations with both similar and diverse subunits, in terms of technological expertise, may exhibit higher R&D performance than do those with little technological variation among their subunits.
In this paper, we investigate how technological relatedness among corporate subunits, or the degree to which corporate subunits utilize similar technological knowledge, affect a firm's innovation activities and performance. We hypothesize a positive linear effect of technological relatedness among the subunits on knowledge flow within a corporate boundary due to superior absorptive capacity (Cohen & Levinthal, 1990) among the subunits. As technological relatedness facilitates knowledge flow among subunits, these subunits will have more opportunities to learn from each other. However, a high level of technological relatedness entails a negative effect on corporate R&D performance in that subunits are less likely to introduce new knowledge elements to each other. We argue that at an extremely high degree of technological relatedness, such negative effect prevails over the positive effect of increasing knowledge flow. We support this argument by empirically showing that technological relatedness has an inverted-U relationship with the frequencies of technological boundary-spanning combination of knowledge elements and with R&D performance.
Our paper uses patent data to measure within-firm knowledge flow and the impact of innovations created by firms. In addition, we employ Silverman's (1996) data that link patent classes and SIC industries to construct technological relatedness measures, and to assess boundary-spanning innovation. We employ a robust generalized linear model and negative binomial regression to investigate the effects of technological relatedness. Statistical findings based on 201 firms in R&D intensive industries support the argument that technological relatedness has both positive and negative effects on corporate R&D activities.
